PDA

View Full Version : LW 2018 Comments/Opinions



Pages : [1] 2 3

Sekhar
12-31-2017, 05:55 PM
Guys, I thought I'd open a thread for feedback from folks who are trying 2018. Let me start with: how does the metallic work? I'm trying to control tint with 100% metallic, and no matter what I do, it colors it with the "color" setting. Shouldn't PBR allow 100% metallic to render without any diffuse color?

Also, I'm not able to resize the object properties window.

I'm liking the changes in UI that I'm seeing so far!

gclayton
12-31-2017, 06:02 PM
Anyone installed the MacOS trial yet? I downloaded the trial but installer says it is damaged and will not run.

I tried downloading multiple times and it still will not run, so it looks like MacOS trial is corrupt.

jperk
12-31-2017, 06:04 PM
Is it possible to change viewport orbit from turntable to trackball in modeler?

Sekhar
12-31-2017, 06:08 PM
Yes, multiple people are reporting of issues with the file. Really sorry about, must be frustrating, hope they fix it quickly. I didn't have any issues on Windows.

BTW, WRT to my question above, here are my settings and render. Messing with Specular and Specular Tint isn't making a difference.

138998

Wickedpup
12-31-2017, 06:13 PM
Is it possible to change viewport orbit from turntable to trackball in modeler?
AFAIK it isn't something that has been implemented i LW2018 so the answer is no.

jperk
12-31-2017, 06:23 PM
AFAIK it isn't something that has been implemented i LW2018 so the answer is no.

thanks no biggie

Kryslin
12-31-2017, 06:24 PM
My first impressions...
Gonna take some getting used to. Some things have moved around. I can see tutorials on how to optimize your render options for speed.
FiberFX is both familiar, and yet, not. As I opened, gonna take some getting used to.

3D Kiwi
12-31-2017, 06:29 PM
First crash rendering two balls and a plane. Awesome!

RebelHill
12-31-2017, 06:31 PM
how does the metallic work? I'm trying to control tint with 100% metallic, and no matter what I do, it colors it with the "color" setting. Shouldn't PBR allow 100% metallic to render without any diffuse color?

Metallic does have no diffuse colour... but that's not the same as having no specular colour...

See here...


https://www.youtube.com/watch?v=QyrHEejaHnk

danielkaiser
12-31-2017, 06:33 PM
In Layout, I'm able to select objects, camera, and lights. But unable to move or scale (respectively) selected items, also can't move the timeline with the mouse.


Win 10

3D Kiwi
12-31-2017, 06:40 PM
Reflections from a background image do not show in the Principled shader until you switch the shading model from photoreal to cell and then back to photoreal???

Sekhar
12-31-2017, 06:55 PM
Metallic does have no diffuse colour... but that's not the same as having no specular colour...

See here...


https://www.youtube.com/watch?v=QyrHEejaHnk

Thanks, but as I said changing the Specular and Specular Tint isn't making any difference. My understanding of PBR was that the Metalling (AKA Specular, if you will) has no diffuse color but can be assigned a tint (which could be same as the diffuse color or different, but controllable regardless). E.g., a gold plate in a green ambient lighting should reflect green but have a gold tint. First of all, I see a Specular AND a Metallic setting here, which itself is confusing. But more importantly, once you set Metallic as 100%, setting Specular to 0% or 100% (or changing the Specular Tint %) has no real effect on the render that I see, which is what I'm trying to understand.

I see the same issue (if it is really an issue or something I'm missing here) in your video at 5:55, where nothing really changed when you reduced the Specular Tint to 0%.

- - - Updated - - -


In Layout, I'm able to select objects, camera, and lights. But unable to move or scale (respectively) selected items, also can't move the timeline with the mouse.


Win 10

I had the same issue till I realized I was using tablet whereas the input was set to mouse. If you're using mouse, I'd double check it's set to mouse in options.

Kryslin
12-31-2017, 06:59 PM
...And I think I have a very minor bug: Image reconstruction filter on the render display viewport appears to be stuck on classic, even though I've gone into render buffer panels and set "Mitchell" on the two default render buffers.

Never mind, it appears to have unstuck itself.

Aww167
12-31-2017, 07:03 PM
Metallic does have no diffuse colour... but that's not the same as having no specular colour...

See here...


https://www.youtube.com/watch?v=QyrHEejaHnk



Thanks for that very helpful simplification of the shader process. Probably saved me several weeks (at least) of testing things out by putting it all into a much more focused perspective which I find really easy to grasp. As always, your videos have a wealth of information which rewards repeated viewing.
A very happy new year to you, and everyone as well.

Sekhar
12-31-2017, 07:04 PM
I'm confused with the noise...looks like it's a lot more nuanced now in 2018 (and less brute force, so I guess that's good). But, I just can't get rid of noise in this simple setup with a ball and only an image environment (no light). The alpha map is pristine, so per the docs it's not the Alpha noise and likely Diffuse Indirect, but increasing camera samples (and reducing threshold) isn't helping. Looks like it's going to take a little time getting used to the workflow, hopefully not too much. :confused:

139002

RebelHill
12-31-2017, 07:05 PM
once you set Metallic as 100%, setting Specular to 0% or 100% (or changing the Specular Tint %) has no real effect on the render that I see, which is what I'm trying to understand.

Because metals ALWAYS have specular tint... that's one of the features of metals. Metals have no diffuse and specular colour, insulators have diffuse colour and sometimes some amount of specular colour. When you turn up metallic, the surface behaves like a metal, which is why the other controls dont do anything at that point.

Sekhar
12-31-2017, 07:10 PM
Because metals ALWAYS have specular tint... that's one of the features of metals. Metals have no diffuse and specular colour, insulators have diffuse colour and sometimes some amount of specular colour. When you turn up metallic, the surface behaves like a metal, which is why the other controls dont do anything at that point.

Ah, OK...I guess I need to get used to the terminology. I had watched some general PBR videos on YouTube for my education to prepare for this release and the terminology in LW isn't quite lining up with what I watched. Regardless, thanks a bunch for your video and prompt responses, you're an asset!

RebelHill
12-31-2017, 07:11 PM
Thanks for that very helpful simplification of the shader process. Probably saved me several weeks (at least) of testing things out by putting it all into a much more focused perspective which I find really easy to grasp. As always, your videos have a wealth of information which rewards repeated viewing.

Good, good.

And there's gonna be a lot more where that came from.

danielkaiser
12-31-2017, 07:21 PM
Thanks, but as I said changing the Specular and Specular Tint isn't making any difference. My understanding of PBR was that the Metalling (AKA Specular, if you will) has no diffuse color but can be assigned a tint (which could be same as the diffuse color or different, but controllable regardless). E.g., a gold plate in a green ambient lighting should reflect green but have a gold tint. First of all, I see a Specular AND a Metallic setting here, which itself is confusing. But more importantly, once you set Metallic as 100%, setting Specular to 0% or 100% (or changing the Specular Tint %) has no real effect on the render that I see, which is what I'm trying to understand.

I see the same issue (if it is really an issue or something I'm missing here) in your video at 5:55, where nothing really changed when you reduced the Specular Tint to 0%.

- - - Updated - - -



I had the same issue till I realized I was using tablet whereas the input was set to mouse. If you're using mouse, I'd double check it's set to mouse in options.

Yup set to mouse, very strange. This is happening in demo mode. I've installed and reinstalled.

Snosrap
12-31-2017, 08:02 PM
Reflections from a background image do not show in the Principled shader until you switch the shading model from photoreal to cell and then back to photoreal???

Getting the same here. Plus I wasn't getting any Fresnel even without an background image. I did notice Antti was using 2018.0.1 in his videos, so maybe there will be a fix coming our way soon. :)

Sekhar
12-31-2017, 08:20 PM
Getting the same here. Plus I wasn't getting any Fresnel even without an background image. I did notice Antti was using 2018.0.1 in his videos, so maybe there will be a fix coming our way soon. :)

What kind of reflections are you guys talking about? If it's background on a metallic surface, it worked right away for me? E.g., see the image I posted earlier here. If it's something else, please clarify so I can also test and confirm.

Also, I'd appreciate if you could throw some light on the noise question I posed earlier. Basically, are you able to get a noise free render with any object with default settings (just a plain diffuse surface) with only a background image (like an ImageWorld env)?

jwiede
12-31-2017, 08:24 PM
Just playing with the trial...

How do I add additional specular terms/layers (for coatings, etc. -- _besides_ clear-coat) on to the BRDF shader? Even with nodes, I'm still not seeing any easy way to layer specular terms onto an existing material.

dyls_E
12-31-2017, 09:38 PM
i have to admit, i'm struggling with it... Changing textures over from the shader tools to the new Principlied bsdf node is doing my head in... shader tools were not elegant, but i felt like i had the control i wanted and gave good predictable results.

Transparency doesn't seem to work right at all for it, and the only way i can get something to go to true transparency is to do a material mix, and add my transparency to the alpha there, then run a standard material with 0 trans, and 1 refraction. and that seems to mess with the edges too. i'm guessing it doesn't work because items aren't physically realistic if they use transparency that isn't like glass or something solid, so it isn't compatible with the Principlied bsdf... but they took away the alternatives, so i don't know what to use.

and i really dislike not having tabs in the camera/lights/output panel. i loved just being able to flick through and adjust everything in one place... now they are scattered again. (i haven't used the mac for this yet, but really hoping they removed the grouping of other windows, that always throws me at work on the macs)

I really want to like LW 2018, there seems to be a few hurdles but once i get over them i hope i can enjoy all the improvements.

Snosrap
12-31-2017, 09:46 PM
and i really dislike not having tabs in the camera/lights/output panel. i loved just being able to flick through and adjust everything in one place... now they are scattered again.

Yep - and I remember when they went to that a lot people gave them flack for it. :) I think this was done out of necessity though with the new rendering system and kind of makes sense on a workflow level. Also I think the Render Properties panel and Surface Editor are using QT, as they have a slightly different look and additional functionality.

Aww167
12-31-2017, 09:49 PM
Good, good.

And there's gonna be a lot more where that came from.

Well, I for one am certainly looking forward to that!

I have a suggestion for the subject of a forthcoming video, if that's not too presumptuous ?
Your point about metal only having specular tint and no diffuse colour has certainly dispelled some confusion regarding understanding how this PBR process works, but something (much?) still puzzles me. Maybe I'm overlooking the obvious or just over-thinking the whole thing, or need to better educate myself on the science (more than likely) but, ... as I see it, a substance is either metal or it isn't; I can't think of anything that's partly metal and partly non-metal, so if you're configuring a metallic shader it would make sense to me that the metallic option is either on or off? On that basis, how does percentage of metallness work in respect of a PBR metal shader?
If the substance is metal, that suggests always 100% on the metallic scale, but then as has been reported, the specular tint setting has no effect. Yet, you point out that metal always has specular tint - so somehow that setting has to play a part. Which leads me to think that some % of metallic should always be employed with a corresponding % of specular tint to obtain a correct result, (on a kind of energy-conservation principle approach), or am I way off??

We're told that the shader engine is designed to help replicate the physically accurate process of light interaction with surfaces, which leads me to think that there must be numerical principles involved toward obtaining a correct result. There has to be a rational approach of some kind to applying certain principles, one would think. I'm really just trying to avoid endless time spent fruitlessly trying out settings without fully understanding what I'm doing, which has been a feature of my experiences so far, so I guess the key is in understanding how far the actual science translates, or is represented by the way the slider settings actually work in practice? How far does such a rationalised approach help in understanding this whole process ? I suspect I'm just over-thinking it all, but the continued confusion is detracting from the potential for creative enjoyment of making art, or whatever it is I like to do..:)

dyls_E
12-31-2017, 10:01 PM
Yep - and I remember when they went to that a lot people gave them flack for it. :) haha yeh i remember someone absolutely offloading about how it messed with his work flow having everything tabbed.
maybe a future update could be simply giving a bit more control over how the windows look and act.

i feel a bit weird in the camera window not being able to select the render style, but the renders are looking really nice, so guessing they don't need those options any more.

Kryslin
12-31-2017, 10:12 PM
First off, Happy new year from the Midwest in the US.

Second, where did the node editor go in FiberFX? While I realize it's not need for shading the fibers, it was useful for connecting up weight maps to control things like Fiber length, selectively placing some tangle, and so on. Heck, with a little bit of node stringing, you can hook up a bias map, too. There is something taking place that winds up soaking my i7 at 100% for 15-20 seconds at a time just editing fibers. Kind of makes making changes painful. Note that this isn't common to FFX in LW2018; This occurred in FFX from 9.6 thru 2015.3 as well.

Still, despite the load/unload cycle, FFX renders blazingly fast.

Cageman
12-31-2017, 10:14 PM
Just playing with the trial...

How do I add additional specular terms/layers (for coatings, etc. -- _besides_ clear-coat) on to the BRDF shader? Even with nodes, I'm still not seeing any easy way to layer specular terms onto an existing material.

What are you trying to do? Remember that the BSDF material is PBR based... so, what you try to achive might not work with the material.

Can you show a screenshot of what you want to achive vs what you currently get out of LW?

jwiede
12-31-2017, 11:41 PM
What are you trying to do? Remember that the BSDF material is PBR based... so, what you try to achive might not work with the material.

Can you show a screenshot of what you want to achive vs what you currently get out of LW?

Off the top of my head: Think of something like a smear of oil on a metal or glass surface, or lacquer wearing off a wood surface, where the BDRF is the base material (metal/glass/wood), and the additional specular represents the spec of the oil/laquer.

It isn't a switch situation, because the underlying material's spec still applies, there's just another specular "above" it optically in places. Such coatings are generally too thin for modeled geo+volume stacking to be viable (nor is that usually efficient/feasible), yet in such situations you need the "stacked" specular terms for the material to look "correct", esp. with coatings that can produce iridescent effects like oils, or certain optical coatings.

Look at how coatings work in Thea or Maxwell, or how specular/reflectance layers work in Vray or C4D Physical to better understand the functionality I'm trying to replicate (Octane can do it as well, but off the top of my head I don't recall the setup needed there).

3D Kiwi
12-31-2017, 11:54 PM
Love that they have worked on network rendering. The new network render controller looks good. Im only in discovery mode so cant test it but was able to get it to see two nodes.

Would be great in the future if they added support for comping apps, AE, fusion etc.

Snosrap
01-01-2018, 12:12 AM
Opinions - the new renderer is great! Big smile on face.

Chris S. (Fez)
01-01-2018, 12:33 AM
Opinions - the new renderer is great! Big smile on face.

Yup. Love it. Much faster than Arnold in Max for interactive surfacing and rendering. Will stress test tomorrow.

2018.5 will hopefully add tone mapping and image filters directly to the VPR. None of the image filters are visible in VPR.

adk
01-01-2018, 03:29 AM
Where the heck is Render-Q ???

Marander
01-01-2018, 03:33 AM
Did some LW2018 trial tests and this is my first verdict (just my personal opinion)

+ the volumetrics render quality and ease-of-use is very good
+ the lighting is good
+ FiberFX looks good
+ displacements look good
+ some OpenGL improvements
+ the Layout view in Modeler seems to work surprisingly fast
+ some parts of the renderer look good (dialectric, subsurface)
+ VPR

- the rest is basically the same as 2015
- UI / UX is still a mess
- undo still doesn't work as it should
- the control for particles and volumetrics shapes is limited
- the OpenGL display is slow compared to C4D (comparing the same fbx scene)
- the OpenGL display looks very jaggy / aliased
- the OpenGL view for volumetrics looks horrible
- I don't see any speed improvements in Layout
- dialog boxes / tabs are still truncated and unreadable
- a character with morphs is handled very slow compared to C4D
- an old massive LW11.6 scene is not handled better in 2018
- conductor / reflective materials don't look good to me
- conductor / reflective channel settings are very limited
- Cel shading is very limited
- I already had a crash in Modeler and Layout but it seems more stable than 2015 (I guess because I've not added 3rd party plugins yet)
- the overall architecture seems outdated to me, definitely not what was promised in the early blog posts
- lots of other stuff missing

After trying some things and playing with some demo content I have to say that I can't really stand the UI and workflow in LW any more, it's just not comfortable or efficient to me after working some time in a modern environment.

No upgrade for me (at least for some time). Besides some simple (but limited) volumetrics there is nothing appealing to me currently. I'm very disappointed that almost nothing was done for a better user experience.

Asticles
01-01-2018, 03:40 AM
The render is really FAST! also with portal lights and denoise!

I will upgrade.

dyls_E
01-01-2018, 04:08 AM
139013
i am really impressed with how easy loading and texturing the megascans 3D assets are now... that used to take me about 10 mins per model, but now it is about 1min, and they look much better. but still haven't worked out how to make the flat plane models like leaves and grass to work now. i used to use a mix of orenNayer*,translucency and blin

Photogram
01-01-2018, 04:11 AM
On my side i can't use enable noise filtering. i check it do a render and when i go in the noise filter option my render is not there so i cannot tweak the noise reduction..

Also i don't find where the reconstructions filters are now... Gaussian, mitchell etc....

Overall the render engine is very good and fast!

Here is my first conversion from Lightwave 2015.3 to 2018

I can't get rid of the noise because the Noise filtering is not working...
Attached Thumbnails Attached Thumbnails

Markc
01-01-2018, 04:23 AM
Anyone installed the MacOS trial yet? I downloaded the trial but installer says it is damaged and will not run.

I tried downloading multiple times and it still will not run, so it looks like MacOS trial is corrupt.
I just purchased the upgrade (for Mac), and both the link in the email and the link in my account say the download is damaged.........:thumbsdow:mad:

avkills
01-01-2018, 04:38 AM
Just upgraded, install says it is damaged! WTF Newtek?

Photogram
01-01-2018, 04:46 AM
I can't get rid of the noise because the Noise filtering is not working...
Attached Thumbnails Attached Thumbnails

I finally found the settings in the render buffers you need to enable noise and then the noise filter begin to operate. The image reconstructions are there too.. :)

Markc
01-01-2018, 04:57 AM
Just upgraded, install says it is damaged! WTF Newtek?

If your on a mac, the solution I found was to use Stuffit Expander 15, which worked fine.
The default Archive Utility and Stuffit 16 say it is damaged.

The .zip must be sensitive to certain archive software, try any alternatives :)

tyrot
01-01-2018, 05:03 AM
we really need a serious test vs. octane ... on rendering. Rendering very insanely complex scenes with full HD under a minute in Octane spoiled me over the years i just got freaked out scene GI caching for animation!

3dworks
01-01-2018, 05:05 AM
If your on a mac, the solution I found was to use Stuffit Expander 15, which worked fine.
The default Archive Utility and Stuffit 16 say it is damaged.

The .zip must be sensitive to certain archive software, try any alternatives :)


i'm using keka, worked well.

pauland
01-01-2018, 05:25 AM
Quite shocked at the Mac unzip debacle.
Even more shocked that if you look at the "try" tab on the lightwave3d page, there's a link to two intro videos for layout and modeller. That would ordinarily be good, but the links are to videos made in 2012! Yes, they are five years old and as blurred as can be.

If I visit a site and the product details hark back more than a couple of years, I just assume it's a dead product.

Newtek are getting real lazy - a new release with a download that doesn't work and promotional material of poor quality and five years old. That won't inspire confidence for new or old customers.

[ Keka worked for me too ]

Anttij77
01-01-2018, 05:51 AM
Getting the same here. Plus I wasn't getting any Fresnel even without an background image. I did notice Antti was using 2018.0.1 in his videos, so maybe there will be a fix coming our way soon. :)

In the "Shading Model" tab, check to see if "Glossy Reflections" is ON. I believe it is OFF when importing an object made pre-2018, because in previous versions indirect reflections were not tied to specularity and did not match.

next_n00b
01-01-2018, 06:12 AM
Orbiting millions of polygons in modeler 2018 is crazy fast on Radeon Pro WX compared to 2015.3. It is like 10x speed.

gamedesign1
01-01-2018, 07:40 AM
Because I use Octane I am finding it really difficult to find a good enough reason to upgrade :( It's shame because I was really looking forward to the new version. But the improvements that have been added just don't benefit me at all. At the moment I cannot justify upgrading to it just to get the cheaper upgrade price. If it had a lot of modeler improvements, or if the CC had been fixed then I may have considered it. It does seem great though for those who use the native renderer, so I do hope they all enjoy it :)

gamedesign1
01-01-2018, 07:50 AM
Quite shocked at the Mac unzip debacle.
Even more shocked that if you look at the "try" tab on the lightwave3d page, there's a link to two intro videos for layout and modeller. That would ordinarily be good, but the links are to videos made in 2012! Yes, they are five years old and as blurred as can be.

If I visit a site and the product details hark back more than a couple of years, I just assume it's a dead product.

Newtek are getting real lazy - a new release with a download that doesn't work and promotional material of poor quality and five years old. That won't inspire confidence for new or old customers.

[ Keka worked for me too ]

Yeah I thought that was bad about the videos too. It does come across as unprofessional. Strange.

BeeVee
01-01-2018, 07:56 AM
For leaves, just try PBSDF and add Thin in the Surface Editor... ;)

B

Verlon
01-01-2018, 09:20 AM
How does the Mac download not work on a default MacOS install? That one seems like a really silly thing to slip past testing, especially in the age of virtual machines. Just have a clean install of supported OS version to test it on.

Bill Carey
01-01-2018, 09:32 AM
So far my first impression is lots to learn, more options all good news. Performance with VPR, abysmal. Tried turning off gi and a few other options but it still is a magnitude slower, even using new objects and shading system.

DBMiller
01-01-2018, 09:35 AM
Anyone installed the MacOS trial yet? I downloaded the trial but installer says it is damaged and will not run.

I tried downloading multiple times and it still will not run, so it looks like MacOS trial is corrupt.

I had the same thing happen. I tried Stuffit Expander instead of the Mac "Archive Utility" expander and it worked fine.

tyrot
01-01-2018, 09:35 AM
Because I use Octane I am finding it really difficult to find a good enough reason to upgrade :( It's shame because I was really looking forward to the new version. But the improvements that have been added just don't benefit me at all. At the moment I cannot justify upgrading to it just to get the cheaper upgrade price. If it had a lot of modeler improvements, or if the CC had been fixed then I may have considered it. It does seem great though for those who use the native renderer, so I do hope they all enjoy it :)

For an Octane user i also have big doubts about it.. :( Yes new geo engine is handling more polies . new shaders are more than cool!.. I got it but in the world of Redshift or Octane .. or other hybrid renderers what this new renderer can offer. May be i am missing something.. I hope somebody who earns his/her life from Octane LW combo can answer me.. and i am not ready to bury my precious 32 bit plugins :(

Sekhar
01-01-2018, 10:04 AM
Poser Pro seems to be working fine (tried with my 2012 version). Only, the Poser Pro Fusion installer adds the plugin under bin/plugins, and it used to be picked up (e.g., in 2015.3), but 2018 doesn't seem to look at that folder anymore. So, moving the PzForLW plugin folder to support/plugins will make it work.

I'm having difficulties with Vue xStream however.

Nicolas Jordan
01-01-2018, 10:08 AM
I haven't had much time to play around with 2018 yet but I'm bummed that Final Gather GI is gone now. I always used Final Gather rather than Monte Carlo for GI since I could always get faster results with it that were almost as good as Monte Carlo. I will have to spend some time with Interpolated Monte Carlo to see if I can get it to render at a decent enough speed for the work I do.

next_n00b
01-01-2018, 10:38 AM
Volumetrics still does not work with instances. Is this something that is very hard to implement?

Snosrap
01-01-2018, 10:50 AM
Where the heck is Render-Q ???

It's there for me.

tyrot
01-01-2018, 10:56 AM
I haven't had much time to play around with 2018 yet but I'm bummed that Final Gather GI is gone now. I always used Final Gather rather than Monte Carlo for GI since I could always get faster results with it that were almost as good as Monte Carlo. I will have to spend some time with Interpolated Monte Carlo to see if I can get it to render at a decent enough speed for the work I do.

please share your findings as you render out scenes...

prometheus
01-01-2018, 10:56 AM
Volumetrics still does not work with instances. Is this something that is very hard to implement?

Well..what do you intend to instance with volumetric items? I do not see any difference as to just cloning the volumetric items..as I understand it, instances is volumetric items..so there would be no point.

Gar
01-01-2018, 10:58 AM
Nothing looks right on 4K under windows 10. The -f command for increasing the font size that worked in 2015.3, screws up the interface. I can't even get It look right, eyes hurtings as its all to small. Can't believe no interface scaling..

Snosrap
01-01-2018, 11:23 AM
In the "Shading Model" tab, check to see if "Glossy Reflections" is ON. I believe it is OFF when importing an object made pre-2018, because in previous versions indirect reflections were not tied to specularity and did not match.

Thanks Antti! Lots to learn. :) Kudos to you and Mark for your work.

erikals
01-01-2018, 11:36 AM
Nicolas, note, in the Benchmark test the new 2018 renders almost Twice as fast as 2015 >
http://forums.newtek.com/showthread.php?133251-11-5-s-BenchmarkMarbles-lws-share-your-machine-s-render-time-here&p=1530277#post1530277

Dan Ritchie
01-01-2018, 11:43 AM
So far my first impression is lots to learn, more options all good news. Performance with VPR, abysmal. Tried turning off gi and a few other options but it still is a magnitude slower, even using new objects and shading system.

Is that because there's no draft mode?

- - - Updated - - -

Shadow maps are gone. I am so glad to see artifacts like those go.

prometheus
01-01-2018, 11:43 AM
Anyone gotten dpont Rman collection procedural textures to work with the volumetrics? I installed the node textures for 64 bit, they show up in the list in node edtior, but the refuse to be added when I click on them for the volumetric item.
Sunsky work as textured environment, but the sk_sunlight isn´t working..it´s there in the list of options for lights, but it refuses to change from distant to sk_sunlight.
pitty those textures doen´t work ..love the gardner clouds and many more in there.

gclayton
01-01-2018, 11:52 AM
Just wondering why there is no documentation delivered with the LW Trial download for new users of LW?

There should also be some sample content made available for the new user to get acquainted with LW!

Ok, I stand corrected. Clicking Help on the menu does open online documentation. I was looking for some delivered docs to be used offline.

erikals
01-01-2018, 11:53 AM
i think the documentation is online now. someone will have to verify.

gclayton
01-01-2018, 11:55 AM
I really don't see any improvement in the GUI for LW like I was expecting.

One of the biggest complaints that I have is that some of the UI can't be resized! For example, in Layout the Object Properties window still cannot be resized and you can only see 2 letters on each tab!!!

I took a quick look at the SDK and I don't see any more options,other than XPanels, to build GUI. I was hoping after seeing some of the comments, that there would be better integration with Qt available to developers!

next_n00b
01-01-2018, 11:59 AM
Well..what do you intend to instance with volumetric items? I do not see any difference as to just cloning the volumetric items..as I understand it, instances is volumetric items..so there would be no point.

Well, I would like to create earth clouds all around the globe. As far as I can see, only nulls can have volumetrics, so the best option would be instancing tens of thousands of nulls on sphere.

Right now I am impressed by Layout primitive type and displacement on it... sort of micropoly displacement with nodes. No more millions of polys for decent asteroid.

next_n00b
01-01-2018, 12:02 PM
Anyone gotten dpont Rman collection procedural textures to work with the volumetrics? I installed the node textures for 64 bit, they show up in the list in node edtior, but the refuse to be added when I click on them for the volumetric item.
Sunsky work as textured environment, but the sk_sunlight isn´t working..it´s there in the list of options for lights, but it refuses to change from distant to sk_sunlight.
pitty those textures doen´t work ..love the gardner clouds and many more in there.


The same is with IFW2 nodes.. they install but cannot be added into the node editor... As far I understand, plugins must be recompiled with new SDK.

Sekhar
01-01-2018, 12:02 PM
Just wondering why there is no documentation delivered with the LW Trial download for new users of LW?

There should also be some sample content made available for the new user to get acquainted with LW!

Docs are online at https://docs.lightwave3d.com/display/LW2018. Also, you can download the content with trial, the link will be on your products page (under the trial that you registered).

goakes
01-01-2018, 12:55 PM
Pick Surface tool not working. 2018 docs show it as: ( This tool needs no initial selection. As you hover over your geometry it will show the surface names and clicking will bring up the Surface Editor with the chosen surface.) This tool is very helpful and shows up in Edit plugins under modeling tool. Pick Surf core_tools.p . Does it work for anyone?

Never mind. Found it.

Dan Ritchie
01-01-2018, 12:58 PM
Are volumetrics completely unaffected by lights? I have a red light in my scene, and no red in my volumetrics.

Snosrap
01-01-2018, 01:00 PM
I really don't see any improvement in the GUI for LW like I was expecting.

One of the biggest complaints that I have is that some of the UI can't be resized! For example, in Layout the Object Properties window still cannot be resized and you can only see 2 letters on each tab!!!

I took a quick look at the SDK and I don't see any more options,other than XPanels, to build GUI. I was hoping after seeing some of the comments, that there would be better integration with Qt available to developers!

It looks like the Surface Editor and the Render Properties panels were built with QT. It's probably just a matter of time before they refactor the others or just use QT on new systems they bring into play like they did here.

Sean Martin
01-01-2018, 01:03 PM
am i missing something, was expecting these 2 to look almost the same?

the one on the left should look almost the same as the one on the right, 2018 new openGL feature. Can someone please test from the sample content pbr, rock or stone I think its called.

from the new docs
OpenGL Improvements

With the PBRGLSL Shading Method, display in OpenGL matches VPR as closely as possible. The screenshot above shows the Principled BSDF material in OpenGL and VPR side by side. Apart from the shadows evident in VPR, the shading matches closely.

Sekhar
01-01-2018, 01:09 PM
am i missing something, was expecting these 2 to look almost the same?

Which two? I see only one image? If you meant the two within the image, the one on the left is a textured shaded solid and the other is a VPR, so it seems right to me?

Dan Ritchie
01-01-2018, 01:17 PM
Are volumetrics completely unaffected by lights? I have a red light in my scene, and no red in my volumetrics.

Ok, I found the "volumetric intensity" under light settings.

jpsent
01-01-2018, 01:21 PM
Just upgraded a few minutes ago and it's nice to see that 2018 can use all 88 threads I have at my disposal. 2015 only allowed 44 threads max. My 14 minute renders have suddenly gone under 2 minutes...

Sekhar
01-01-2018, 01:30 PM
BTW, did they end the charter program? I've been seeing the products as "UPGRADE -CHARTER," but for 2018 I only see "UPGRADE." Anyone else on charter who also upgraded to 2018?

grabiller
01-01-2018, 01:34 PM
[Reconstruction Filters]
I'm not sure this is a good idea to have reconstruction filters choice per buffer while having filter radius per camera. Normally each reconstruction filter needs a different filter radius. With this new setup, one will end up using totally wrong filter radius for some reconstruction filters choosen on some buffers.
Yet I'm not sure this is a good idea to have different reconstruction filters per buffer.

(otherwise I ike LW2018 so far)

grabiller
01-01-2018, 01:46 PM
[Noise Reduction Filter]
To avoid bad surprises and disappointments I advise those who want to test the Noise Reduction Filter to start with default values: Filter Size 3 and Tolerance 0.03.
The default Tolerance 0.1 is way too high and you will end up with a blurry mess.

Airwaves
01-01-2018, 02:14 PM
Quick question from a noob to most things lightwave. I am trying the scene content that was provided and so far everything looks grainy. I just cannot figure out if it is the lighting used or noise like you talk about. Just not sure why things look grainy. I wish I had a better description to give. What setting can I turn off to see what it looks like without the grainy or is this the new PBR?

grabiller
01-01-2018, 02:20 PM
Quick question from a noob to most things lightwave. I am trying the scene content that was provided and so far everything looks grainy. I just cannot figure out if it is the lighting used or noise like you talk about. Just not sure why things look grainy. I wish I had a better description to give. What setting can I turn off to see what it looks like without the grainy or is this the new PBR?

It is explained here:
https://docs.lightwave3d.com/display/LW2018/Removing+Noise+workflow

paulhart
01-01-2018, 02:26 PM
Sorry for the brain *art, but I had forgotten that I had to "Import" the menu branches. It had been awhile since I needed to do any of this??
SOLVED
[Maybe I am the slow one here?? But after install 2018, I attempted to add a menu for RHiggit, using the RHiggit2Menu.cfg file saved off from prior installation, but nothing changes. Same for the 3ps_menu_for+_Layout.cfg for the 3DPowers tools. How to add new menu or plugins. Yes I looked at the documentation but I must not be as on topic as others, as no one has mentioned any difficulties getting old plugins installed. I re-installed LWCad without difficulty, including new menu tabs across the top, so it must work, somehow???]

Marander
01-01-2018, 02:44 PM
I really don't see any improvement in the GUI for LW like I was expecting.

One of the biggest complaints that I have is that some of the UI can't be resized! For example, in Layout the Object Properties window still cannot be resized and you can only see 2 letters on each tab!!!


Yes exactly. I really don't understand why LW3DG was not able to fix this within 3 years. The GUI and User Experience is a major disappointment for me. Don't even want to install on my laptop with 15" 4k display, no point.

139032 139033 139034

Text tool in Year 2018, this is just unbelievable.

139035

Or the Volumetrics OpenGL preview

139036

And OpenGL has performance issues and stutters (dual Nvidia rig) using LW2018, it's much slower than other 3D applications and looks worse.

Airwaves
01-01-2018, 02:50 PM
It is explained here:
https://docs.lightwave3d.com/display/LW2018/Removing+Noise+workflow

Thank you for this. I appreciate it.
I turned off global illumination which helped a lot but I admit this is seeming too daunting a task at the moment to jump to 2018. I do such simple animations that I am starting to wonder if 2018 will be add more work. I am not defeated yet .... I still have 30 days to try it so I hope I can find a way to make it work for me.

grabiller
01-01-2018, 03:15 PM
Thank you for this. I appreciate it.
I turned off global illumination which helped a lot but I admit this is seeming too daunting a task at the moment to jump to 2018. I do such simple animations that I am starting to wonder if 2018 will be add more work. I am not defeated yet .... I still have 30 days to try it so I hope I can find a way to make it work for me.

I'm not sure the noise issue you have would be specific to LW2018 or even LW itself, it is a general issue to any montecarlo raytracer or pathtracer. Perhaps in your case you are stumbling on area lights (aside GI) which are usually the first "noise generators" you encounter (aside glossy surfaces and DOF or MB wich are normally off by default).

First be sure to use decent camera settings like adaptive sampling min/max 1/16 and a threshold around 0.025. Then check your lights, raise light sampling if needed.

There are scenes in the Content that are correctly set and do not generate visible noise (Furry.lws, ../..).

erikals
01-01-2018, 03:18 PM
regarding "cut names"

while not ideal, what other option would you suggest?
please make a mockup, show what you think it should look like.
text bubble solution < nope
wider window < wouldn't help much, unless you'd make it really wide
vertical text < no way.


Or the Volumetrics OpenGL preview
what's wrong with that? [kidding] yes, that needs an update.

OpenGL performance issues > i believe these are being looked into.

adk
01-01-2018, 03:23 PM
It's there for me.

Thanks for chiming in Snosrap. All the supposed new presets are also missing so it seems I have a bad install ?
You'd think that after some 15+ years I would know how to cleanly install LW.

PS: New install fixed it all. No idea why the previous one failed ?

dyls_E
01-01-2018, 03:28 PM
For leaves, just try PBSDF and add Thin in the Surface Editor... ;)

B hey thanks. I tried that, the surface looks good. but i was having a lot of trouble with transparency around the edge. shadows don't seem to respect the trans layer and were just casting shadows for the whole plane. Clip maps worked, but i usually only use them for distant objects, and ones that go over the limits of transparency limits. I've found a bit of a hack by mixing a standard material with 0 trans, and 1 refraction with the trans layer as alpha. but the edges don't look right any more.

*i must of been doing something dumb... it's working fine now

hypersuperduper
01-01-2018, 03:29 PM
I'm finding a number of small improvements in modeler that I really did not expect. little things like edit edges respecting symmetry. The new render workflow is still pretty daunting though.

MichaelT
01-01-2018, 03:30 PM
It is explained here:
https://docs.lightwave3d.com/display/LW2018/Removing+Noise+workflow

It is explained badly.

Here (you need to activate noise reduction before you can use it):

First:
139037
139038

Then render, and *then*:
139039

I would prefer if there was a quick option for all of this. Like a button.

Off to sleep.. good night all :)

dyls_E
01-01-2018, 03:31 PM
moving the PzForLW plugin folder to support/plugins will make it work.

I'm having difficulties with Vue xStream however. ah i was wondering if i would need to wait for poser to work :) i've never been able to get xStream to work with lightwave very well. (or even by itself lately)

dyls_E
01-01-2018, 03:40 PM
BTW, did they end the charter program? I've been seeing the products as "UPGRADE -CHARTER," but for 2018 I only see "UPGRADE." Anyone else on charter who also upgraded to 2018?
pretty sure the 2015 was the last upgrade in the charter program, tho we haven't had to pay for an upgrade for over 3 years, so it's not exactly an expensive program haha

grabiller
01-01-2018, 03:46 PM
It is explained badly.

Here (you need to activate noise reduction before you can use it):

First:
139037
139038

Then render, and *then*:
139039

I would prefer if there was a quick option for all of this. Like a button.

Off to sleep.. good night all :)

And also just a "Star Wars" button ? ^^

The documentation mentioned above is a general methodology to spot and fix noise issues in your scene, not a Noise Reduction Filter documentation, it is much more general. The Noise Reduction Filter is not a magic wand that will free your mind - and images - from troubles and issues automagically, for instance it does not address antialiasing issues (what they call Alpha noise). It is just the "Cherry on the Cake" that will add the final touch - getting rid of the remaining noise - without the need to go crazy on sampling. But your scene must be quite clean enough for this to work. Wrongly used you will end up with Impressionnist Paintings or just a blurry mess.

That said, once you get used to it, it is pretty straight forward, you choose the buffers that need denoising (not all will need it so it makes sense to have on/off switches there), then you enable the denoising filter feature at render time and that's it. You don't really need the third step "render then.." once you have found the settings that generally work for you. Only for specific cases if you really need to. Try with the default settings but with Tolerance at 0.03 instead of 0.1, and you should be safe for most of the situations.

prometheus
01-01-2018, 03:50 PM
Yes exactly. I really don't understand why LW3DG was not able to fix this within 3 years. The GUI and User Experience is a major disappointment for me. Don't even want to install on my laptop with 15" 4k display, no point.

139032 139033 139034

Text tool in Year 2018, this is just unbelievable.

139035

Or the Volumetrics OpenGL preview

139036
.

the opengl volumetric preview, could be way better, former we only had some wirerepresentation of volumetrics, now we got a flat sized area showing it..so that is not good enough...should be like dynamite has it..o fumefx..they do that right.


https://vimeo.com/9127056

Chris S. (Fez)
01-01-2018, 04:56 PM
Yes exactly. I really don't understand why LW3DG was not able to fix this within 3 years. The GUI and User Experience is a major disappointment for me. Don't even want to install on my laptop with 15" 4k display, no point.

139032 139033 139034

Text tool in Year 2018, this is just unbelievable.

139035

Or the Volumetrics OpenGL preview

139036

And OpenGL has performance issues and stutters (dual Nvidia rig) using LW2018, it's much slower than other 3D applications and looks worse.

The primary focus for the release was rendering and I feel they have delivered in a big way. The VPR is super fast and scales beautiful on multi-core machines. However, I am having trouble isolating and refining noise and am hoping there are individual and global AA multipliers for GI, lights and shadows that work like Arnold. Probably just missing some settings but so far I prefer the minimalist workflow in Fprime and Modo progressive rendering. There seem to be all the pieces here for a progressive production renderer and I feel they should move on that immediately.

The new surfacing is extremely flexible and powerful if you are comfortable with nodes. Feel lost without IFW textures though.

The cell shading and edge-rendering control is extremely flexible and powerful if you are comfortable with nodes.

QT with UI scaling did not make it into this release but is apparently a priority. I agree, very disappointing and should be top priority.

True Art has a great Nodal text tool. Again, quite powerful if you are comfortable with nodes. I agree, a respectable Text Tool needs to be native and expanded with all the features of the new Layout primitives. Having said that, the TextPlus tool in Max with modifier stack is as straightforward as it is powerful and makes short work of most text tasks. I'd like the native tool to shoot for similar functionality. I can only imagine C4D offers such functionality already.

I am getting great performance in my initial tests of 2018. Quick 3,000,000 poly test object deformed with bones compares very favorably to Max 2018.4. There has gotta be a conflict or driver issue on your machine. Or maybe save in the new LW format and reload?

Overall I am satisfied with progress in performance (though it is not the beast I was expecting and hope they continue to optimize the new engine) and very pleased with rendering and surfacing...once I figure out how to refine the noise and convince Richard to update IFW textures.

wingzeta
01-01-2018, 05:09 PM
Just messing around with the new shading system. I wanted to try the Principled shader with textures, so I took a tiling rusty metal texture I had, and plugged it in with cubic mapping. I have a map that is driving the metalness channel, to separate the metal from rust. Trying to drive the roughness with a map was harder to get a good result, so I dropped it, and just set a number. Need to figure that out. Also, tried Dielectric for the windshield, but wasn't able to get any reflections, so I ended up using a delta for the glass, with a texture driving Spec and Roughness. The headlights are just a dielectric, with no texture. The new system seems pretty good, I just have to figure it out.

Still setting a lot by eye, but I'm hoping to learn some "correct" PBR settings for one material vs. another. This render is just with default settings.
139040

RebelHill
01-01-2018, 05:13 PM
Trying to drive the roughness with a map was harder to get a good result, so I dropped it, and just set a number. Need to figure that out.

Have a peek at how I handle that sorta thing here...


https://www.youtube.com/watch?v=H-HLcq98LXM

wingzeta
01-01-2018, 05:19 PM
Have a peek at how I handle that sorta thing here...


https://www.youtube.com/watch?v=H-HLcq98LXM

Thanks for that! I will give it a try.

Nicolas Jordan
01-01-2018, 06:14 PM
Something very annoying that I notice is if you open a scene all the objects surfaces are expanded in the surface editor instead of being collapsed. Is there any option to have them all collapsed by default?

Is there a good reason for this to be this way by default?

Marander
01-01-2018, 06:31 PM
the opengl volumetric preview, could be way better, former we only had some wirerepresentation of volumetrics, now we got a flat sized area showing it..so that is not good enough...

I'd like to show how intuitive volumetric, fluid and particle editing with render preview and OpenGL can look like with a screenshot but I'm not going to mention other software again. The contrast could not be any bigger.

I really like the render output of LW2018 volumetrics but the way it's presented outside VPR and the limited abilities is just not for me.

Marander
01-01-2018, 06:49 PM
Feel lost without IFW textures though.
...
QT with UI scaling did not make it into this release but is apparently a priority. I agree, very disappointing and should be top priority.
...
True Art has a great Nodal text tool.
...
I can only imagine C4D offers such functionality already.
...
Or maybe save in the new LW format and reload?


Thanks for answering. I'm comfortable with nodes and I see this as a good step for LW to go more that direction.

Yes IFW2 (or similar procedurals) are essential for me too.
...
There are some new dialogs like Render Settings that look good, that would have been fine for me.
...
TA Text, yes I know but still basic compared to other native text tools that exist since years in other 3d packages.
...
That is a good idea, I didn't save in the 2018 format first (big LW 11.6 scifi scene).

But as I mentioned in a previous post, the OpenGL performance of a fbx scene is very poor compared to my main app (outdoor scene and character with rig and face morphs).

prometheus
01-01-2018, 07:28 PM
I'd like to show how intuitive volumetric, fluid and particle editing with render preview and OpenGL can look like with a screenshot but I'm not going to mention other software again. The contrast could not be any bigger.

I really like the render output of LW2018 volumetrics but the way it's presented outside VPR and the limited abilities is just not for me.

Please do not mention pointless references that goes to a void of nothing to refere to, it serves no good.. I assume you mean houdini open GL, which does that in the openGL, and fluids isn´t something to discuss, since we do not have that natively, turbulenceFD does a pretty decent job in the openGL for that, except for that it can not render both fire and smoke, as blender and houdini can.

prometheus
01-01-2018, 08:00 PM
I noticed that Distant light doesn´t seem to be able to react withing the volumetric scattering, I thought All lights should be able to do that..or am I wrong and have missed some setting here?

Snosrap
01-01-2018, 08:22 PM
I noticed that Distant light doesn´t seem to be able to react withing the volumetric scattering, I thought All lights should be able to do that..or am I wrong and have missed some setting here?

Yeah - they have the volumetric settings but they don't do anything. Actually I'm surprised they would be volumetric as Distant lights don't really have a position in 3D space, just rotation.

dyls_E
01-01-2018, 08:25 PM
kinda wish they left the old texturing system in there too, it was really handy for setting up a decent preview for openGL then use the nodes to overwrite the settings for the render. I don't think bsdf node is really giving me the kind of preview i want without using vpr.
139048

Nicolas Jordan
01-01-2018, 08:40 PM
Does anyone know how to do a blurry reflection in Lightwave 2018?

Do we get blurry reflections now by adjusting roughness on the BSDF material? What about blurred reflection on a standard material?

Shiny_Mike
01-01-2018, 08:41 PM
Just plug your textures into a standard material node, and plug that into "OpenGL" material input of the main "Surface" node. You can keep your main material for rendering and 2nd standard for previewing :thumbsup:


kinda wish they left the old texturing system in there too, it was really handy for setting up a decent preview for openGL then use the nodes to overwrite the settings for the render. I don't think bsdf node is really giving me the kind of preview i want without using vpr.
139048

- - - Updated - - -

The roughness value will control the apparent blur of the reflection.


Does anyone know how to do a blurry reflection in Lightwave 2018?

dyls_E
01-01-2018, 08:43 PM
Just plug your textures into a standard material node, and plug that into "OpenGL" material input of the main "Surface" node. You can keep your main material for rendering and 2nd standard for previewing :thumbsup:



- - - Updated - - -
HEY!!! awesome. i did not see that there!!! exactly what i wanted!

prometheus
01-01-2018, 08:52 PM
Yeah - they have the volumetric settings but they don't do anything. Actually I'm surprised they would be volumetric as Distant lights don't really have a position in 3D space, just rotation.

You know, I used distant light for the godray volumetric effect, but with hypervoxels items for catching the rays, so I do not see why it should be so darn hard to implement this.


This is a distant light in lw 2015 with hypervoxels large enough to cover the scene and act as volumetric density to catch the light...this is something I thought just activating volume scattering would take care of in the new lw 2018.


https://www.youtube.com/watch?v=DbkvT_HgKRs

Terragen and vue can pull this off with realistic distant sun, feels like I go back to square one in the quest of getting a decent sun, I thought I heard all lights could be volumetric, but obviously that was wrong, and it will still be tricky to set up sunlight to behave realisticly in distance with soft shadow and volumetric lighting, guess I have to try the old add a volumetric item again to make things volumetric

Snosrap
01-01-2018, 08:56 PM
[QUOTE=prometheus;1530602]You know, I used distant light for the godray volumetric effect, but with hypervoxels items for catching the rays, so I do not see why it should be so darn hard to implement this./QUOTE]

Well maybe it's a bug- the controls are there.

MichaelT
01-02-2018, 01:04 AM
And also just a "Star Wars" button ? ^^

The documentation mentioned above is a general methodology to spot and fix noise issues in your scene, not a Noise Reduction Filter documentation, it is much more general. The Noise Reduction Filter is not a magic wand that will free your mind - and images - from troubles and issues automagically, for instance it does not address antialiasing issues (what they call Alpha noise). It is just the "Cherry on the Cake" that will add the final touch - getting rid of the remaining noise - without the need to go crazy on sampling. But your scene must be quite clean enough for this to work. Wrongly used you will end up with Impressionnist Paintings or just a blurry mess.

That said, once you get used to it, it is pretty straight forward, you choose the buffers that need denoising (not all will need it so it makes sense to have on/off switches there), then you enable the denoising filter feature at render time and that's it. You don't really need the third step "render then.." once you have found the settings that generally work for you. Only for specific cases if you really need to. Try with the default settings but with Tolerance at 0.03 instead of 0.1, and you should be safe for most of the situations.

The denoiser is something many would like to use.. a lot of the time. Nobody is talking about some magic wand etc.. here. I don't "need" to use the denoiser at all. I can simply use the full precision image, and post the whole thing. But that is not what I'm talking about here.

Marander
01-02-2018, 01:24 AM
Please do not mention pointless references that goes to a void of nothing to refere to, it serves no good.. I assume you mean houdini open GL, which does that in the openGL, and fluids isn´t something to discuss, since we do not have that natively, turbulenceFD does a pretty decent job in the openGL for that, except for that it can not render both fire and smoke, as blender and houdini can.

No that's not what I was referring to (I only use the Apprentice version of it until now) and I don't find its OpenGL representation specifically good.

I was talking of Cloud editing with volume primitives (they can be parametric, polygonal, particles or nulls) and displayed nicely in OpenGL as well as IPR.

Sorry not going to post references / screenshot of other apps again (I'm sure you know what I'm m working with now), I'd like to respect Steve's new forum policy.

My point is that there are many ways of displaying or modifying Volumetrics in OpenGL other than a white aliased solid disc. The LW 2018 UI and UX is just a dealbreaker for me.

Marander
01-02-2018, 01:27 AM
The denoiser is something many would like to use.. a lot of the time. Nobody is talking about some magic wand etc.. here. I don't "need" to use the denoiser at all. I can simply use the full precision image, and post the whole thing. But that is not what I'm talking about here.

Denoising should be one click in the render settings (with additional options of course). Again, there are many examples out there that do this properly.

Marander
01-02-2018, 01:45 AM
Due to the fact that LWNext / 2018 did not deliver for me it still looks and feels like a 90s application I'm stopping my involvement and interest in it now.

If there's an update for a reasonable price that fixes these UI / UX / Workflow issues and offers something unique, I will have a look again.

Best thing that happened to LW is its community. Some very nice people here, I wish you all the best in 2018 (the software and the year) and beyond!

grabiller
01-02-2018, 03:32 AM
Denoising should be one click in the render settings (with additional options of course). Again, there are many examples out there that do this properly.

Well this is exactly the case. One click and it works, *granted* you have selected the buffer(s) on which to apply it, and again, it makes sense because you don't want to apply the denoiser on buffers that don't need it. Perhaps here, and I agree, the switch should be on by default on the default final render buffer, yes.

Aside from that, it is exactly like DOF, MB, GI, Volume, you enable the switch and it works.

Then you have the additional options, and the possibility to adjust those without the need to re-render the frame, which is pretty cool.

Now, I believe the real question is: Does the new Noise Reduction Filter delivers ? It's a bit early to give a definitive answer but it seems to do a pretty good job, granted you adjust the default settings.

But if you are more concerned about how many clicks it takes then I'm afraid you missed the point.

DogBoy
01-02-2018, 04:08 AM
am i missing something, was expecting these 2 to look almost the same?

the one on the left should look almost the same as the one on the right, 2018 new openGL feature. Can someone please test from the sample content pbr, rock or stone I think its called.

from the new docs
OpenGL Improvements

With the PBRGLSL Shading Method, display in OpenGL matches VPR as closely as possible. The screenshot above shows the Principled BSDF material in OpenGL and VPR side by side. Apart from the shadows evident in VPR, the shading matches closely.

I'm getting far closer OpenGL to VPR than you. So am assuming it is your card/settings
139053

Photogram
01-02-2018, 04:51 AM
Look pretty good already :)

next_n00b
01-02-2018, 05:32 AM
I'm getting far closer OpenGL to VPR than you. So am assuming it is your card/settings
139053

It looks like it is only working with Quadro drivers. Does not work on my iMac 10.13 with Nvidia GTX. It does work with Quadro on Win7, but not with Radeon Pro WX on Win7.

Sean Martin
01-02-2018, 05:45 AM
It looks like it is only working with Quadro drivers. Does not work on my iMac 10.13 with Nvidia GTX. It does work with Quadro on Win7, but not with Radeon Pro WX on Win7.

I have the same iMac and card as you, so it doesn't work!

DogBoy
01-02-2018, 05:49 AM
It looks like it is only working with Quadro drivers. Does not work on my iMac 10.13 with Nvidia GTX. It does work with Quadro on Win7, but not with Radeon Pro WX on Win7.

Um, I'm using a GeForce not a Quadro, on Win 10.
Sounds like the usual issue with Apple and OpenGL.

next_n00b
01-02-2018, 05:49 AM
It is weird because Dinosaur_WLK_GL does not work on my Win7 with Radeon Pro WX while Dinosaur_WLK_GL _Subdivided DOES work with Textured Shaded Solid and PBRGLSLShaders.

DogBoy
01-02-2018, 05:52 AM
I have the same iMac and card as you, so it doesn't work!

How do you know you are on the same iMac and card?

next_n00b
01-02-2018, 05:55 AM
I have a 2012 i7 gtx650m imac.

next_n00b
01-02-2018, 06:24 AM
I have checked one more time this PBR_Stone_Ball scene an conclusion is;

it does work OK with Quadro on Win7, PBRGLSLShaders method

and

it does NOT working OK on my iMac or/and Win7 with Radeon Pro WX

Otterman
01-02-2018, 07:06 AM
PBR_Stone_Ball scene

Sorry Am i missing something, is there 2018 content to download? Where is it?

rustythe1
01-02-2018, 07:11 AM
when you purchase upgrade it will be in your account

Otterman
01-02-2018, 07:16 AM
when you purchase upgrade it will be in your account

Oh ok, not with the trail version then?

Im on a mac and having the issue with installation so I'm stumped already.

next_n00b
01-02-2018, 07:17 AM
I have a Trial version and it is in my account, too.

Otterman
01-02-2018, 07:18 AM
Cool beans. Hopefully when I get it installed I can register it then grab the content. Eager to try.

Cheers Lightwavers!

next_n00b
01-02-2018, 07:20 AM
If you have problems unziping install try unziping with Stuffit Expander.

Otterman
01-02-2018, 07:40 AM
Yeah Tried that and Keta. Still no joy. Also tried Chiltons Terminal command but it's bricked. Will have to try on my machine at home. Cheers

Sekhar
01-02-2018, 07:46 AM
I have checked one more time this PBR_Stone_Ball scene an conclusion is;

it does work OK with Quadro on Win7, PBRGLSLShaders method

and

it does NOT working OK on my iMac or/and Win7 with Radeon Pro WX

Yeah, it is working on my Windows 10 with GTX 980. Looks like NormalMap is off by default, but once you flip it...nice!

next_n00b
01-02-2018, 07:47 AM
Stuffit Expander 15.0.3a build 3720 worked well for me.

next_n00b
01-02-2018, 08:10 AM
Do you mean "NormalMap Channel" = Checked? If so, does not work for me. I have deleted all config for 2018, played with normal map, mipmap on/off, linear/srgb ... all the stuff I usually do to setup normal map.

raymondtrace
01-02-2018, 08:33 AM
Quite shocked at the Mac unzip debacle.

That's more of an issue with Apple. Apple's native unarchiver has always been lacking. There's a lot of ways to produce a ZIP archive and Mac is blind to many of them.

https://theunarchiver.com/

Sekhar
01-02-2018, 08:51 AM
Do you mean "NormalMap Channel" = Checked? If so, does not work for me. I have deleted all config for 2018, played with normal map, mipmap on/off, linear/srgb ... all the stuff I usually do to setup normal map.

Yeah, below is what I'm seeing without NormalMap Channel set and with it set. Again, this is on Windows 10 and GTX 980.

139057139058

Vladimir545
01-02-2018, 09:19 AM
Hello all and sry for my english.

Begin testing render in trial 2018.

139059

139060

I liked the render very much. For me, extremely fast-as in normal mode, and gi interpolated, bruteforce is also quite fast.
**excellent flexible shading.

Everything is very quickly, predictably and visually adjusted.


Thanks to the developers for the great work.


About Complain here. "The application from the 90s-ies Ui))", why overload the interface ??. Rebuild the Ui? -buds will be bugging and this will affect stability on 100%.

Everything is fine. Do not break what works and do not change the concept-Layout, Modeler. This package Lightwave always favorably different from its competitors, its approach to work.

Last time I had such pleasure from research softimage xsi.

After xsi package was "killed" one "modern" company with very "modern" ui in their products.

Thanks again for the release Lw 2018, I'm delighted. Only The price could really be little cheaper)

bazsa73
01-02-2018, 09:22 AM
I have tortured FFX thoroughly and no crashes so far even with VPR on.
--EDIT--
So far I'm pleased with this new FFX. I see potential in it. Thanks NT!

Nicolas Jordan
01-02-2018, 10:25 AM
Is anyone else having trouble with 2018 keeping your preferences from session to session?

I'm finding it's not remembering things I set in the preferences like Tablet, Mip Mapping toggle to off etc.

Sekhar
01-02-2018, 10:56 AM
Is anyone else having trouble with 2018 keeping your preferences from session to session?

I'm finding it's not remembering things I set in the preferences like Tablet, Mip Mapping toggle to off etc.

No problems here: saving all the stuff like layout, frame rate, tablet, etc. fine. This is on Windows 10. I wonder if it's to do with where it's trying to write the config files in your case, if there's an issue with perms or something.

next_n00b
01-02-2018, 10:56 AM
Win7, same workstation

Radeon Pro and Quadro

sadkkf
01-02-2018, 11:00 AM
Win7, same workstation

Radeon Pro and Quadro

Well, crap. I put a Radeon into my new system from an IT friend's recommendation. Wishing now I went with NVidia.

next_n00b
01-02-2018, 11:08 AM
I don't think that Radeon driver is to blame, because some scenes do work well with PBRGLSLShaders. I mentioned before Dinosaur_WLK_GL _Subdivided scene.

Nicolas Jordan
01-02-2018, 11:14 AM
No problems here: saving all the stuff like layout, frame rate, tablet, etc. fine. This is on Windows 10. I wonder if it's to do with where it's trying to write the config files in your case, if there's an issue with perms or something.

Ok thanks, I will have to look into it and see if I can figure out what going on. I'm on Windows 10 as well.

rustythe1
01-02-2018, 11:38 AM
well i seem to be getting a lot of issues with transparent and reflective surfaces, there is all sorts of visible banding coming through, transparent turns pure black when nothing is behind it (even though the backdrop is white)
you can see here the floor seems to be reflecting some sort of pixelated image, but its only set to ray traced, its not the backdrop image as when i turn that on you can see it over the top of the big pixels.
not looking good at the moment ( cranking rays up for each type changes nothing)
139069

MichaelT
01-02-2018, 11:44 AM
That "banding" looks like a reflection of a low resolution image. Is there an image there that we don't see?

rustythe1
01-02-2018, 11:51 AM
yes, it appears that when the surface is set to raytrace it can still see the backdrop if its added to the environment light, so setting the MIS samples of that light way up starts to remove it, unfortunately seems to impact render times greatly so using a gradient backdrop rather than an image seems to be a better option when using plain surfaces. still cant fix the transparency though

MichaelT
01-02-2018, 02:03 PM
The blackness in the transparency is probably due to you having too few bounces. I think I understand that you want the backdrop to be invisible on selected objects.. correct? If so, I don't think that is possible. Maybe if you write a script to tell the object to ignore any rays coming from the backdrop (or selected object(s)) could solve it? But that would probably violate a lot of things. And cause problems, as you would now have to collect rays that were blocked by the same objects etc.. I tried setting up something myself.. but I don't see the problems you have. So I'm a bit uncertain as to how you actually set it up:

http://forums.newtek.com/attachment.php?attachmentid=139072&d=1530798
139072

rustythe1
01-02-2018, 02:32 PM
well actually in this instance the backdrop is white, its just not showing in the transparency? so the only way to get the backdrop to show white is to put a physical plane all the way around the background (a bit like you above) as it seems to not be able to show the backdrop colour either, i.e. materials only show physical elements in the scene (the texture environment must be seen as a physical element as that does render through transparency, its just the colour and gradient backdrops do not?), but also like i said in the other thread, the standard material node behaves as expected so anything composited or in the back drop shows as it should. i would have thought being as effects and arch viz comping would have been such a big thing it would have been something that had been spotted before (especially as that content scene clearly shows the error in the spyder car!)

MichaelT
01-02-2018, 02:57 PM
You do see backdrop color in the glass in my image (if you look closely) if that is what you meant? Do you mean that if I put a transparent item using a standard shader in front of the cubes, I will get issues?

rustythe1
01-02-2018, 04:15 PM
sorry, no i mean it looks like you have some walls around the floor and glass cubes, those would show through the transparency as they are being traced, what i have is a flat floor, then the cube on top, nothing around apart from in "backdrop options" the backdrop is white, then the glass becomes black,

my second issue was
if you add a texture environment to "backdrop options" then in "composition options" set to use "backdrop colour" and change that to white, you still see the textured environment,

MichaelT
01-02-2018, 04:29 PM
Ok. the blackness is again probably due to you having too few bounces in your rendering settings, so I suggest you check them out. As for the backdrop still being visible. That is probably due to the white box not being set to double sided. Which means the backdrop doesn't "see" the cube, and simply passes by it. But if checking that box (double sided) and it still doesn't block the backdrop despite being blocked.. then yes... that would be a bug and should be reported. Because that means those rays travel through all geometry unhindered.

Also, the environment light will sample the backdrop if inside geometry. Luckily though you can exclude objects from its effects. Hope that helps.

update: Missed this comment of yours: "composition options" set to use "backdrop colour" and change that to white, you still see the textured environment," ... hmm yeah, that does sound like a bug to me.

But just to show what I mean with my comment before the update. I still have an active backdrop.. but it is being blocked by the geometry being double sided. Also.. enough bounces removes blackness. There is a near invisible standard shader box in front:

http://forums.newtek.com/attachment.php?attachmentid=139090&d=1530823
139090

rustythe1
01-02-2018, 05:01 PM
like i said, no walls around the objects, its just a simple floor and cube, so 6 rays should be enough for 1 transparent object (i have cranked it to 20 and no difference) see image its a very simple set up,
139091

Wade
01-02-2018, 07:15 PM
Shot in the dark but - have you tried double sided?

139094

MichaelT
01-02-2018, 07:37 PM
like i said, no walls around the objects, its just a simple floor and cube, so 6 rays should be enough for 1 transparent object (i have cranked it to 20 and no difference) see image its a very simple set up,
139091

I got it.. I think we are talking around each other... Like I said.. I think that your "composition options" set to use "backdrop colour" and change that to white, you still see the textured environment" really is a bug, and you should report it. Sounds like it is ignoring color if you have a texture.
In any case.. have you checked that you have enough ray bounces?

http://forums.newtek.com/attachment.php?attachmentid=139095&d=1530860
139095

Too few bounces means the ray will just cut off.

Kryslin
01-03-2018, 12:25 AM
I'll just ask straight up : I've been trying to reduce noise in my FFX test render without using the noise filter, so I hit the docs... And they keep mentioning Shading Samples...

Where is this setting located? It's not on the render options panel, not on the camera options panel...It's mentioned, but not shown where you can change it.

rustythe1
01-03-2018, 02:42 AM
yep ive set them all up and down, lots of rays, 2 sided makes it worse as all the internal surfaces become black if they are in contact with others (that was a known issue in 2015, i think highlighted to matt)

Kryslin, the settings you are looking for are in the window above your post, in 2018 different settings affect different surface types, so you can have clean shiny surfaces, but noisy matt surfaces, the first tab has the "ray" setting which is a sort of global setting, default is 1 but pump it up and it will clean renders up a lot. maybe watch RHs videos as he describes the affects of each one quite well.

rustythe1
01-03-2018, 02:49 AM
what happens to yours if you take these walls away? and yes i think its a bug too, and it causes problems with the alpha channel, use a texture environment and alpha is as it should be, use just backdrop colour and it becomes solid so you cant map out the black, i can think of many work arounds, but i should not have to, and as i often have to work on the fly live will cause delays when trying to get instant feedback (just have to use standard node for glass but not sure how that impacts the overall renders)
139101

rustythe1
01-03-2018, 03:04 AM
ok, its defiantly a bug, but not actually with the material, it appears that changing something else (i don't seem to be able to repeat it at the moment) was preventing any sample settings from updating to the selected settings, i just set the scene up with the exact same settings and it now renders as it should, it looks like the refraction setting was not changing (even though i had it set to 64 at one point) must have been stuck at 5 by what i can work out, what ever the problem is it also seems to come with scenes from 2015 as that's where i first saw the problem, unfortunately i cant share them as I'm stuck with NDA

RomanS
01-03-2018, 04:18 AM
Anyone installed the MacOS trial yet? I downloaded the trial but installer says it is damaged and will not run.

I tried downloading multiple times and it still will not run, so it looks like MacOS trial is corrupt.


Yes, there is something wrong with the Mac-Installer when downloading from a Mac ;-)
But downloaded from a Windows machine, it works. Wired but true.

Roman

OFF
01-03-2018, 04:46 AM
I got my license and played with demo scenes and my own scenes.

Here are my impressions. I liked the fact that new volumetrics were added, new exellent shaders for surfaces, a little better OGL performance.

But overall I'm very upset with the speed of rendering. I expected really some breakthrough - not only in the field of beautiful shaders, but also in a qualitative jump in speed and quality of rendering.

Another disorder is related to the performance of OGL when working with character animation. There are also very small positive changes.

The same goes for the FFX. Even a simple character - a squirrel covered in short hair - makes the LW suffer for 10 minutes. The rendering itself takes hardly 2-3 minutes in this process.

In general, I'm certainly glad of the release of the LW 2018 (Next) version. In any case, as a platform for future improvements - this can only be welcomed. But at the same time, since there are no hints of a road map for the development of LW - many small studios and freelancers (like me) will have to puzzle over how to speed up the working process. Either by purchasing additional render servers, or by purchasing third-party renderers - like Octane render for example (which is not very cheap). There is still hope for an early release of Kray 3, adapted for LW 2018. Wait and see.

Asticles
01-03-2018, 06:07 AM
I got my license and played with demo scenes and my own scenes.

Here are my impressions. I liked the fact that new volumetrics were added, new exellent shaders for surfaces, a little better OGL performance.

But overall I'm very upset with the speed of rendering. I expected really some breakthrough - not only in the field of beautiful shaders, but also in a qualitative jump in speed and quality of rendering.

Another disorder is related to the performance of OGL when working with character animation. There are also very small positive changes.

The same goes for the FFX. Even a simple character - a squirrel covered in short hair - makes the LW suffer for 10 minutes. The rendering itself takes hardly 2-3 minutes in this process.

In general, I'm certainly glad of the release of the LW 2018 (Next) version. In any case, as a platform for future improvements - this can only be welcomed. But at the same time, since there are no hints of a road map for the development of LW - many small studios and freelancers (like me) will have to puzzle over how to speed up the working process. Either by purchasing additional render servers, or by purchasing third-party renderers - like Octane render for example (which is not very cheap). There is still hope for an early release of Kray 3, adapted for LW 2018. Wait and see.

Well, what I've seen is that the way of working with noise and samples has changed. Since there is no antialias render on each camera sample pass, you cannot see how well is it going from the camera settings. It's much better to keep them low and adjust samples on other areas, like it is shown on the denoising guide.

Also, all the tweaking should be done from VPR. A pity there is no (or I don't know how) vpr zoom of the render region. Is it possible?

It is a paradigm shift. IMHO

MichaelT
01-03-2018, 06:46 AM
I got my license and played with demo scenes and my own scenes.

Here are my impressions. I liked the fact that new volumetrics were added, new exellent shaders for surfaces, a little better OGL performance.

But overall I'm very upset with the speed of rendering. I expected really some breakthrough - not only in the field of beautiful shaders, but also in a qualitative jump in speed and quality of rendering.

Another disorder is related to the performance of OGL when working with character animation. There are also very small positive changes.

The same goes for the FFX. Even a simple character - a squirrel covered in short hair - makes the LW suffer for 10 minutes. The rendering itself takes hardly 2-3 minutes in this process.

In general, I'm certainly glad of the release of the LW 2018 (Next) version. In any case, as a platform for future improvements - this can only be welcomed. But at the same time, since there are no hints of a road map for the development of LW - many small studios and freelancers (like me) will have to puzzle over how to speed up the working process. Either by purchasing additional render servers, or by purchasing third-party renderers - like Octane render for example (which is not very cheap). There is still hope for an early release of Kray 3, adapted for LW 2018. Wait and see.

You can get it to render really fast. But you need to get into settings and manually tweak them to get there. For example, you can begin by setting everything to as low as possible, and work yourself up from there.. depending on what you need in your scene. But if you have volumetrics, expect a lot of time added. Because dealing with 3d volumes is a whole other monster :) My point is though, that you can get render speeds in the seconds. But since the rendering engine is entirely replaced, and is now PBR.. it does take a bit of work to get there. You should however, not have to do that very often. Unless you change scene requirements often :)

MichaelT
01-03-2018, 07:22 AM
what happens to yours if you take these walls away? and yes i think its a bug too, and it causes problems with the alpha channel, use a texture environment and alpha is as it should be, use just backdrop colour and it becomes solid so you cant map out the black, i can think of many work arounds, but i should not have to, and as i often have to work on the fly live will cause delays when trying to get instant feedback (just have to use standard node for glass but not sure how that impacts the overall renders)
139101

139114

zapper1998
01-03-2018, 09:25 AM
2018 is kinda clunky, I think its me...
Performance Camera Crashes LW...
Memory hog as usual...Wish the new Render engine would use GPU... Future request. someday we hope..
Noticed a render slow down, some scenes fast some slow ...
Only been tinkering with 2018 for a few days.. getting use to the new GUI panels and such...
Previous scenes from 2015 need to be redone to work in 2018 because of Octane..

But I do think once i get the flow worked out looking good

jeric_synergy
01-03-2018, 09:56 AM
Report every bug. Have you reported the Performance Camera?

+++++++++++++++++++++++++++++++++++++++

I see a lot of concerns about render times, usually with imported 2015 scenes.

Which of the included Content scenes show off the speed of the new render engine, as a contrast to un-tweeked legacy scenes? :confused::confused:

Kryslin
01-03-2018, 10:08 AM
Kryslin, the settings you are looking for are in the window above your post, in 2018 different settings affect different surface types, so you can have clean shiny surfaces, but noisy matt surfaces, the first tab has the "ray" setting which is a sort of global setting, default is 1 but pump it up and it will clean renders up a lot. maybe watch RHs videos as he describes the affects of each one quite well.

Thanks, got it. Playing around with the various settings, I'm getting some decent results at speed.

Asticles
01-03-2018, 10:39 AM
Edited

BeeVee
01-03-2018, 01:10 PM
Report every bug. Have you reported the Performance Camera?

+++++++++++++++++++++++++++++++++++++++

I see a lot of concerns about render times, usually with imported 2015 scenes.

Which of the included Content scenes show off the speed of the new render engine, as a contrast to un-tweeked legacy scenes? :confused::confused:

What Performance Camera? There's a Perspective Camera and an Advanced Camera (which is slow, and always has been, because of its single-threaded nature).

B

gar26lw
01-03-2018, 06:54 PM
surface baking cam is multi threaded, right?

Thomas Helzle
01-03-2018, 07:05 PM
I'm pretty speechless from the total lack of major improvements in the modelling area. Same old clunky modeller. Outch.
I was expecting at least an initial implementation of poly modelling in Layout.
Instead the one area in LW I had the least problems with and actually liked the most - rendering - was brought up to snuff, now doing what everybody else is doing too (PBR)...

What I totally don't get: What is so earth-shattering in LW 2018 that it couldn't be talked about?
This update won't even raise a brow with any of the people in the industry I met over the last couple of years.

I use Redshift. They release a new version almost every week. They talk freely with their users and keep them up to date. There is a basic roadmap and if something is getting delayed, users are told what's up. And yes, sometimes that hurts.
I use Houdini. They release a daily build with latest bugfixes. They talk freely with their users, engage even on Discord, sometimes rant most funnily. Nobody dies. We're all human. It's actually fun. Sometimes there is a roadmap, sometimes there isn't, but the last major update (16) was mindblowing. No problems with trusting them to come up with more great stuff in the future.
I use Rhino. They have an open beta of Rhino 6 running for some years and many of the devs communicate and ask for feedback on their forums. Bugs are fixed, problems are discussed. Not the most innovative software in the world, but quite versatile and solid for what it is.

Now: What on earth is so unbelievable special about an outdated 3D software like Lightwave playing catch-up that it can't be talked about?
I must be missing something absolutely basic here.


I'll check the demo some more, but right now I have a hard time justifying even 295$ with the little actual use I get out of LW.
Not knowing where the whole application is headed doesn't help AT ALL with this.
If I knew the next milestone would be comprehensive modelling in Layout I might see this as an investment in a glorious future, but after the last couple of years I lost pretty much all the illusions about LW I ever had...

Cheers,

Tom

Sekhar
01-03-2018, 07:26 PM
I'm pretty speechless from the total lack of major improvements in the modelling area. Same old clunky modeller. Outch.
I was expecting at least an initial implementation of poly modelling in Layout.
Instead the one area in LW I had the least problems with and actually liked the most - rendering - was brought up to snuff, now doing what everybody else is doing too (PBR)...

What I totally don't get: What is so earth-shattering in LW 2018 that it couldn't be talked about?
This update won't even raise a brow with any of the people in the industry I met over the last couple of years.

I use Redshift. They release a new version almost every week. They talk freely with their users and keep them up to date. There is a basic roadmap and if something is getting delayed, users are told what's up. And yes, sometimes that hurts.
I use Houdini. They release a daily build with latest bugfixes. They talk freely with their users, engage even on Discord, sometimes rant most funnily. Nobody dies. We're all human. It's actually fun. Sometimes there is a roadmap, sometimes there isn't, but the last major update (16) was mindblowing. No problems with trusting them to come up with more great stuff in the future.
I use Rhino. They have an open beta of Rhino 6 running for some years and many of the devs communicate and ask for feedback on their forums. Bugs are fixed, problems are discussed. Not the most innovative software in the world, but quite versatile and solid for what it is.

Now: What on earth is so unbelievable special about an outdated 3D software like Lightwave playing catch-up that it can't be talked about?
I must be missing something absolutely basic here.


I'll check the demo some more, but right now I have a hard time justifying even 295$ with the little actual use I get out of LW.
Not knowing where the whole application is headed doesn't help AT ALL with this.
If I knew the next milestone would be comprehensive modelling in Layout I might see this as an investment in a glorious future, but after the last couple of years I lost pretty much all the illusions about LW I ever had...

Cheers,

Tom

It's a bit harsh and unfair to judge 2018 based on modeler as virtually all material (i.e., important) changes are in layout - hope you get to check it out, there is a lot of discussion and many pointers here on the new (and exciting to many of us) stuff.

shrox
01-03-2018, 07:40 PM
I am having real concerns about the surface material changes and other points I have read about. I am wondering if it might be better for me just to keep my 2015v running and just get FBTurbulance.

samurai_x
01-03-2018, 09:29 PM
I am having real concerns about the surface material changes and other points I have read about. I am wondering if it might be better for me just to keep my 2015v running and just get FBTurbulance.

The change was pretty obvious if you saw the blog. The focus was again the renderer and shading in lightwave 2018. No way previous models, scene lighting would be compatible with 2018. And going with a bidirectional pathtracer using cpu and not gpu, don't' know if its the right direction.








I'm pretty speechless from the total lack of major improvements in the modelling area. Same old clunky modeller. Outch.
I was expecting at least an initial implementation of poly modelling in Layout.
Instead the one area in LW I had the least problems with and actually liked the most - rendering - was brought up to snuff, now doing what everybody else is doing too (PBR)...

What I totally don't get: What is so earth-shattering in LW 2018 that it couldn't be talked about?
This update won't even raise a brow with any of the people in the industry I met over the last couple of years.

I use Redshift. They release a new version almost every week. They talk freely with their users and keep them up to date. There is a basic roadmap and if something is getting delayed, users are told what's up. And yes, sometimes that hurts.
I use Houdini. They release a daily build with latest bugfixes. They talk freely with their users, engage even on Discord, sometimes rant most funnily. Nobody dies. We're all human. It's actually fun. Sometimes there is a roadmap, sometimes there isn't, but the last major update (16) was mindblowing. No problems with trusting them to come up with more great stuff in the future.
I use Rhino. They have an open beta of Rhino 6 running for some years and many of the devs communicate and ask for feedback on their forums. Bugs are fixed, problems are discussed. Not the most innovative software in the world, but quite versatile and solid for what it is.

Now: What on earth is so unbelievable special about an outdated 3D software like Lightwave playing catch-up that it can't be talked about?
I must be missing something absolutely basic here.


I'll check the demo some more, but right now I have a hard time justifying even 295$ with the little actual use I get out of LW.
Not knowing where the whole application is headed doesn't help AT ALL with this.
If I knew the next milestone would be comprehensive modelling in Layout I might see this as an investment in a glorious future, but after the last couple of years I lost pretty much all the illusions about LW I ever had...

Cheers,

Tom


Been saying the same thing for over a year. The renderer was never the problem. I was expecting some mesh tools in layout. But I've accepted it now, that will never happen. Unification is a pipe dream. 3 teams have already tried since Luxology team. Can't be done unless you have infiinite time and money.

gar26lw
01-03-2018, 10:57 PM
I am having real concerns about the surface material changes and other points I have read about. I am wondering if it might be better for me just to keep my 2015v running and just get FBTurbulance.

what kind of concerns?

Sekhar
01-03-2018, 11:14 PM
The change was pretty obvious if you saw the blog. The focus was again the renderer and shading in lightwave 2018. No way previous models, scene lighting would be compatible with 2018.

I was doing some test renders to compare 2018 with 2015.3. Here are the renders with 2015.3 and 2018 of the same 2015 Superhero scene...as you can see they look nothing like each other. This is using the same standard surfaces in 2018 as in the original scene. The rendering was much faster in 2018 though, for what it's worth (1m 57s vs. 22m 34s in 2015.3), though I'm sure they'll be closer once you apply all the adjustments for 2018. What's the NewTek recommended process for using existing scenes/objects in 2018 as it looks like they will need work?

139153139154

gar26lw
01-03-2018, 11:17 PM
that’s one massive speed increase. what’s it like with the new pbr shaders applied?

Sekhar
01-03-2018, 11:49 PM
that’s one massive speed increase. what’s it like with the new pbr shaders applied?

No, this is using the original surfaces, which is my point in that it looks like we can't just load an existing scene or object without significant re-work...unless I'm missing something here. To be more accurate: Lightwave created a copy of the 2015 scene in 2018 format, so it may have changed something in the surface settings, but they seem to be the same.

Anyway, I did a quick change of a couple of the surfaces to the new BSDF (PBR) material to match the 2015 render a bit more, see below. The render time did increase to 8m 13s as I expected, but it is still way less than the 2015's 22m 34s, so I'm very confident that there is a big speedup in 2018. There is a lot to check out, for sure.

139156

samurai_x
01-04-2018, 12:25 AM
I was doing some test renders to compare 2018 with 2015.3. Here are the renders with 2015.3 and 2018 of the same 2015 Superhero scene...as you can see they look nothing like each other. This is using the same standard surfaces in 2018 as in the original scene. The rendering was much faster in 2018 though, for what it's worth (1m 57s vs. 22m 34s in 2015.3), though I'm sure they'll be closer once you apply all the adjustments for 2018. What's the NewTek recommended process for using existing scenes/objects in 2018 as it looks like they will need work?

139153139154

Just looking at the two renders. Lw 2018 would render faster because its a lot less final looking with tons of noise. The lw 2015 render on the right took 22m?? I think that's way too long for something so simple. Probably 5 mins max with I7 cpu fullhd size.

EVeryone has to start from nearly scratch with shading and lighting with lw 2018. Two totally different render engines.

erikals
01-04-2018, 12:30 AM
the 2018 doesn't have the same Reflection Blur, add more RefBlur, and render-time in 2018 should go up.

AntiAliasing also need to be increased in 2018 to match (or add more Gi Rays, or both)

render quality (versus speed) should be better in 2018 though.

MarcusM
01-04-2018, 01:01 AM
the 2018 doesn't have the same Reflection Blur, add more RefBlur, and render-time in 2018 should go up.

AntiAliasing also need to be increased in 2018 to match (or add more Gi Rays, or both)

render quality (versus speed) should be better in 2018 though.

Hi erikals. I know you like test that stuff so maybe you can do video ;] In tests LW 2018 I saw that in Modeler, Expand selection is much faster than in 2015.3.

erikals
01-04-2018, 05:08 AM
nice, i'll give it a test, see what happens :)
thanks for suggestion. :)

OFF
01-04-2018, 05:57 AM
Hmm, it turns out that if I work in the Principled Material mode - then I can not see the texture on my models? Since textures assigned through nodes are not displayed in OGL.

DogBoy
01-04-2018, 06:44 AM
Hmm, it turns out that if I work in the Principled Material mode - then I can not see the texture on my models? Since textures assigned through nodes are not displayed in OGL.
What, in 2018? of course they should be (with the exception of Skin). If you can't see a texture, try using a second material going into the OGL Material input to test.
Here is Principled in OGL139164

Thomas Helzle
01-04-2018, 06:49 AM
It's a bit harsh and unfair to judge 2018 based on modeler as virtually all material (i.e., important) changes are in layout - hope you get to check it out, there is a lot of discussion and many pointers here on the new (and exciting to many of us) stuff.

Well, that's my whole point - all the new stuff is basically about rendering which wasn't the main area of concern with LW to begin with.
Since the update to LW 6.0 from 5.6, rendering was one of the best parts of LW and back then actually DID raise some brows in the industry (HDRI! WOW!). So now the gap between the (capable) renderer and the absurdly outdated and unconnected (in a lot of ways) modeller opens up even more.
If I would be able to see this as a clear strategy towards a _completely_ overhauled LW in any reasonable timeframe, I'd be happy to support it, even if I wouldn't personally use it in it's current state (I actually did that for a long time since I switched to XSI in 2004 for my main 3D app). But since there seems to be no clear vision whatsoever (or at least none that would be shared with the users) I fail to see why I should support this dabbling about anymore.

I mean seriously: Good Renderers are out there like stars in the Milky Way. You can buy them and plug them into whatever 3D app you like if you use one of the more used ones. What the 3D market needs the least is _another_ renderer that does the same old PBR stuff like everybody else, especially on the CPU but has no up to date infrastructure behind it to make it stand up in the current market as a complete solution to creating beautiful and innovative work from scratch in an up to date workflow.

All the examples I've seen so far are somewhere between cheap looking crap and stuff I was able to do years ago somewhere else already. I haven't seen a single image that would make me go "wow" or make me want to try LW if I wouldn't be a user already.
The cell shaded stuff looks like something from the 90ies, even game engines do that better these days.
In realtime.

There was a time when Maya users exported to LW to render. But today I can't see more than maybe a handful of diehards doing anything like that and go to the trouble of dealing with the completely outdated way of handling objects and scenes in LW. So I fail to see who is supposed to be the target audience for this software.
If you want to save money, use Blender, it has come a long way and also has a very capable renderer, is in constant open development and does modelling much better than LW.
If you want to work in the industry, use anything _but_ LW.

And instead of running an open beta with good communications with the userbase it looks as if betatesting only just began and involves advance payment...

This whole company seems to be as stuck in the last century as the software.

This may become the first time since LW 5.6 where I won't update.
I just can't see this going anywhere - and I even kept faith after the Core debacle...

And if 2018 breaks backwards compatibility with existing scenes, that would make it even more useless to me (I will have to do more testing to see how it holds up there).

And believe me, I'm very very restrained in my posts here, I would have much harsher words for this mess if the people doing it would sit in front of me (not the individual programmers doing their fine work, but the people in "management" or whatever we are supposed to call it).

This endless and slow way of dying makes me very very sad for LW, which once was at the forefront of innovation and now feels like a long-dead zombie dragging along it's brittle bones from sheer lack of being able to die in peace and dignity.

Cheers,

Tom

MarcusM
01-04-2018, 07:46 AM
One of my tests. Rendered in full HD. Standard materials. Changes I made only in lighting, described lower.
Image in attachment is especially low res.

LW 2015.3 - 1m 37s
Ambient Intensity -40%
GI Intensity 85%

LW 2018 - 3m 9 s
No Ambient Light (darker areas)
GI Intensity 120% (almost the same result brightness like in 2015.3)

For example when I must render 10 throws to print on A3, in big resolutions, it will be difference in hours.

mummyman
01-04-2018, 07:55 AM
One of my tests. Rendered in full HD. Standard materials. Changes I made only in lighting, described lower.
Image in attachment is especially low res.

LW 2015.3 - 1m 37s
Ambient Intensity -40%
GI Intensity 85%

LW 2018 - 3m 9 s
No Ambient Light (darker areas)
GI Intensity 120% (almost the same result brightness like in 2015.3)

For example when I must render 10 throws to print on A3, in big resolutions, it will be difference in hours.

Well, they did say that the standard material in 2018 is going to be slower.. so you'll have to do a convert. Even then, from what I think I understand...it'll be slower. Did you try converting it to the new Principled shader?

MarcusM
01-04-2018, 08:00 AM
Well, they did say that the standard material in 2018 is going to be slower.. so you'll have to do a convert. Even then, from what I think I understand...it'll be slower. Did you try converting it to the new Principled shader?


My goal was the same result and test 1 to 1. I will not convert this propably in near future. There is to many different materials and models im using also for different things etc.

Just showing my results. BTW simplest way to fake Ambient light?

samurai_x
01-04-2018, 08:15 AM
One of my tests. Rendered in full HD. Standard materials. Changes I made only in lighting, described lower.
Image in attachment is especially low res.

LW 2015.3 - 1m 37s
Ambient Intensity -40%
GI Intensity 85%

LW 2018 - 3m 9 s
No Ambient Light (darker areas)
GI Intensity 120% (almost the same result brightness like in 2015.3)

For example when I must render 10 throws to print on A3, in big resolutions, it will be difference in hours.


Lw 2018 looks dimmer, duller. Probably fake it with environment light with no shadows.
People really must start from nearly scratch with lw 2018 shading and surfacing. No way around it.

mummyman
01-04-2018, 08:23 AM
Did you see Andrew Comb's video post about using the environment light? I know it's not the same.. but maybe you can do a workaround somehow. I'm planning on keeping certain projects in 2015 until I get up to speed and comfortable using 2018 for a full-feature project. Gonna be a big learning curve!

bazsa73
01-04-2018, 09:05 AM
Spherical reflection image seam cannot be animated anymore. Envelope button is missing. Why?

patrickbod
01-04-2018, 09:19 AM
Hi,

I had the same pb.
I have found a workaround solution : download the file in a windows system and execute it in your Mac. It works fine

MarcusM
01-04-2018, 09:25 AM
I have desire to report as a bug missing Use Texture and Remove Texture buttons in Texture Editor.

zapper1998
01-04-2018, 09:28 AM
Report every bug. Have you reported the Performance Camera?

+++++++++++++++++++++++++++++++++++++++

I see a lot of concerns about render times, usually with imported 2015 scenes.

Which of the included Content scenes show off the speed of the new render engine, as a contrast to un-tweeked legacy scenes? :confused::confused:


Yes i did report the bug
and it was the performance camera crashing scene, all other camera's work fine

yup

erikals
01-04-2018, 10:04 AM
Hi erikals. I know you like test that stuff so maybe you can do video ;] In tests LW 2018 I saw that in Modeler, Expand selection is much faster than in 2015.3.

neat stuff   :)


https://youtu.be/R2KaIpv3AvQ

stiff paper
01-04-2018, 10:16 AM
This endless and slow way of dying makes me very very sad for LW, which once was at the forefront of innovation and now feels like a long-dead zombie dragging along it's brittle bones from sheer lack of being able to die in peace and dignity.
I just wanted to say that I agree with pretty much everything you've said. Also, I too would be much ruder about it, but everybody seems to be happy, and I don't think there would be any point.


...not the individual programmers doing their fine work...
This is where we differ. I think that ever since the Luxology team left, the problem has specifically been the programmers. Management's problem has always been that it's loyal to those programmers.

For me, personally, this feels like the absolute end. Not a new beginning.

Ciao everybody. It's been a gas, but I doubt I'll be around any more.

Ztreem
01-04-2018, 10:27 AM
Well, that's my whole point - all the new stuff is basically about rendering which wasn't the main area of concern with LW to begin with.
Since the update to LW 6.0 from 5.6, rendering was one of the best parts of LW and back then actually DID raise some brows in the industry (HDRI! WOW!). So now the gap between the (capable) renderer and the absurdly outdated and unconnected (in a lot of ways) modeller opens up even more.
If I would be able to see this as a clear strategy towards a _completely_ overhauled LW in any reasonable timeframe, I'd be happy to support it, even if I wouldn't personally use it in it's current state (I actually did that for a long time since I switched to XSI in 2004 for my main 3D app). But since there seems to be no clear vision whatsoever (or at least none that would be shared with the users) I fail to see why I should support this dabbling about anymore.

I mean seriously: Good Renderers are out there like stars in the Milky Way. You can buy them and plug them into whatever 3D app you like if you use one of the more used ones. What the 3D market needs the least is _another_ renderer that does the same old PBR stuff like everybody else, especially on the CPU but has no up to date infrastructure behind it to make it stand up in the current market as a complete solution to creating beautiful and innovative work from scratch in an up to date workflow.

All the examples I've seen so far are somewhere between cheap looking crap and stuff I was able to do years ago somewhere else already. I haven't seen a single image that would make me go "wow" or make me want to try LW if I wouldn't be a user already.
The cell shaded stuff looks like something from the 90ies, even game engines do that better these days.
In realtime.

There was a time when Maya users exported to LW to render. But today I can't see more than maybe a handful of diehards doing anything like that and go to the trouble of dealing with the completely outdated way of handling objects and scenes in LW. So I fail to see who is supposed to be the target audience for this software.
If you want to save money, use Blender, it has come a long way and also has a very capable renderer, is in constant open development and does modelling much better than LW.
If you want to work in the industry, use anything _but_ LW.

And instead of running an open beta with good communications with the userbase it looks as if betatesting only just began and involves advance payment...

This whole company seems to be as stuck in the last century as the software.

This may become the first time since LW 5.6 where I won't update.
I just can't see this going anywhere - and I even kept faith after the Core debacle...

And if 2018 breaks backwards compatibility with existing scenes, that would make it even more useless to me (I will have to do more testing to see how it holds up there).

And believe me, I'm very very restrained in my posts here, I would have much harsher words for this mess if the people doing it would sit in front of me (not the individual programmers doing their fine work, but the people in "management" or whatever we are supposed to call it).

This endless and slow way of dying makes me very very sad for LW, which once was at the forefront of innovation and now feels like a long-dead zombie dragging along it's brittle bones from sheer lack of being able to die in peace and dignity.

Cheers,

Tom

+1
Exactly my thoughts. I will still download and do some comparison between LW and Blender, but right now it feels like I will switch to Blender as my main app instead of LW.

MarcusM
01-04-2018, 10:35 AM
neat stuff   :)


https://youtu.be/R2KaIpv3AvQ

Yeah, that is it. Only one thing faster in Modeler.
For example moving 90k poly sphere is smoother in 2015.3 (On my gtx 1060)
How many more clicks user must take to put texture on actual default material (Principled).

gar26lw
01-04-2018, 10:54 AM
modo is now faster for initial setup. the ux needs tweaked to regain speed.

erikals
01-04-2018, 10:57 AM
For example moving 90k poly sphere is smoother in 2015.3
Modeler seems 100% the same speed here in 11.6 vs 2018


How many more clicks user must take to put texture on actual default material (Principled).
not sure, haven't gotten around to this just yet.

erikals
01-04-2018, 11:04 AM
modo is now faster for initial setup. the ux needs tweaked to regain speed.

modo can be neat, for exterior scenes and renders however i'd choose LW2018.
LW2018 +instancing / OD Tools instancing / Advanced Placement (needs new code) is hard to beat.
(vs cost and speed that is)

the only thing better would be LW2018 + Octane.


though sure, many nice things in Maya/Max/Blender/C4D/Modo/Houdini also

prometheus
01-04-2018, 11:28 AM
As someone here mentioned... "How many more clicks user must take to put texture on actual default material (Principled). "

I am a bit worried actually of what I see in terms of the workflow approach ..and what has been one of lightwave´s strengths when setting up things, from surfacing to working with hypervoxels.
I kind of get a feeling that Lightwave workflow strays away to a completly different approach, which means you have to look for your tools, your settings in a more hard digging way, like adding deformations (finding it in a logical way for easy acess) surfacing as mentioned, and also how we work with the new volumetrics, It could very well be that the team just needs more time..and may actually change the face of the volumetric properties, or access to those controls.

It´s way to early for me to say if I will upgrade, it´s only been a few days of testing, and I will do much more testing than previously just this upcoming weekend, if it weren´t for the great looks of the volumetrics..then I would have been very doubtful, as it stands now...that may be the foremost reason to upgrade, but still to early to say..and the volumetrics needs work in the form of UI, and some additional stuff I think.
have hardly tried the newer materials, just a few of them, hardly tried modeler, tested the lights, where I was a bit dissapointed that the distance Light, which is to be the best one for sunlight...Doesn´t seem to work for volumetrics, which I felt was odd since they adapted an overall volumetric scattering intergrator, and that kind of light is Important for me.

Tried some OpenVDB imports from blender, works..but seems buggy and crashes as often as it works.

Dpont´s node textures is broken, as are dp kit..though there are native replacements for some of that, how much I do not know..and it will probably be worked on natively as the team goes on and pick up request to fill what is broken with Dpont nodes.

Open gl rotations stutters..both in Layout and in modeler.

The render speed, not that much impressed with it..in some cases it seems slower, but so hard to compare due to what is on by default and newer settings, and newer more realistic materials.

Airwaves
01-04-2018, 12:05 PM
I was doing some test renders to compare 2018 with 2015.3. Here are the renders with 2015.3 and 2018 of the same 2015 Superhero scene...as you can see they look nothing like each other. This is using the same standard surfaces in 2018 as in the original scene. The rendering was much faster in 2018 though, for what it's worth (1m 57s vs. 22m 34s in 2015.3), though I'm sure they'll be closer once you apply all the adjustments for 2018. What's the NewTek recommended process for using existing scenes/objects in 2018 as it looks like they will need work?

139153139154

This is where my struggle is with going to 2018. When you compare the two 2015 looks so much better. I know you did a better one below but my point is that I am worried I will spend more time trying to deal with all the noise. I appreciate you posting this as it is helping me really decide if it is worth it to move to 2018.

Did you have to work on removing noise a lot?

Airwaves
01-04-2018, 12:08 PM
You know what would help me decide on 2018 is if I knew they would be doing updates to it in the near future to fix things but my concern would be the next update will be next year for 2019.

hypersuperduper
01-04-2018, 12:16 PM
I completed a little test project with a lw2018 Trial. A little blobby guy from modeling to surfacing, animating and rendering and except for a few glaring bugs that I expect the devs to address I am mostly happy with the new version, the gl performance and actually using the program seems somehow much zippier, the object properties panel is vastly improved. and modeler, surprisingly , while largely the same did recieve a number of smaller updates, some of which, I discovered purely by accident. the new materials and renderer are very nice and I have no doubt they will be easier to set up for realistic renders but they are not fast at all and I don’t yet really know what renders fast and clean easily and what doesn’t. In the old lw it was easy to get pretty quick vpr performance even if many surfacing options could slow things down to a crawl. In 2018 all of that is completely different. I need to find a new set of go-to solutions.

The new documentation is also much better designed, but it seems incomplete. It really should give at least a cursory explanation of every single tool or option that is visible as a button Or an input field in the standard lw2018 install. If it’s in the interface it needs to be in the documentation. Currently a lot of it, particularly older functions are just not even mentioned. Hopefully this will improve with user feedback.

As for upgrading, I almost certainly will, but I would like to see a maintenance update within the next couple weeks to fix a couple of the bigger braindeader bugs (installer that is incompatible with standard OSX unpacker, thinks like that) this would go a long way towards demonstrating that Newtek plans to support this version and their users like they need to and have suggested they will.

gerry_g
01-04-2018, 01:13 PM
2018 seems slow to update sometimes when tweaking values in vpr, one of the values in a gradient I was setting up refused to move or delete, I presume this is due to use of qt or whatever it’s called, it seems to be notorious on the Mac for buggyness

prometheus
01-04-2018, 01:33 PM
2018 seems slow to update sometimes when tweaking values in vpr, one of the values in a gradient I was setting up refused to move or delete, I presume this is due to use of qt or whatever it’s called, it seems to be notorious on the Mac for buggyness

Yes, I have found that to be an issue for me as well, and is unfortunately bad, when the new render vpr actually seem to become slower than 2015.

Thomas Helzle
01-04-2018, 02:45 PM
This is where we differ. I think that ever since the Luxology team left, the problem has specifically been the programmers. Management's problem has always been that it's loyal to those programmers.

Not sure if we differ that much. I have absolutely no clue about the internal structure and personage of Newtek/LW3DG so I assume that the management is in error if the software doesn't have a clear direction, since directing and focusing is usually a management job. If they don't do that, it's again their fault. Most programmers are too deep in their code to make good decisions in the general realm.

I also have no clue what led to the Luxologists leaving, but I found so much strange and oddly familiar stuff (not in a good way) in Modo in the time I used it (bought 601, tried to like it, didn't in the end) that I wondered if it wasn't a good thing to lose them. Totally over engineered piece of buggy and hard to use software. And Brad Peebler made me cringe - should probably sell used cars.

Probably it's a problem with people in general. Like William Vaughan was maybe good for selling LW to total noobs but had a big part in making LW the laughing stock of the industry with his crappy primitive Seus models. Rob Powers gave me hope there, since at least stuff suddenly looked professional. This release again is more back to the old standards of "developer art".

So in the end I probably have simply outgrown Lightwave and what it can be or wants to be years ago.

Now it feels like that hot cheerleader from back in school that you find is still a waitress in the same old diner in the same old town 30 years later...

Cheers,

Tom

Thomas Helzle
01-04-2018, 03:11 PM
I did another test: Catmull-Clark edge weighting in Modeller is just as broken as it always was from the first day of it's implementation - how many years ago?
As soon as you edit the geometry further after assigning CC weights, you get a totally broken mesh.

It think this software really has outlived it's "best before" date for about a decade.

Cheers,

Tom

gar26lw
01-04-2018, 03:25 PM
does layout understand points, edges and polys now?

- - - Updated - - -


You know what would help me decide on 2018 is if I knew they would be doing updates to it in the near future to fix things but my concern would be the next update will be next year for 2019.

yeah same. i want to know when bugs are to be fixed and have some idea of next set of features, with a timeframe. without that, it’s hard to take things seriously and commit.

Chris S. (Fez)
01-04-2018, 05:25 PM
I don't see the problem. Rendering and viewport performance was the focus and they delivered. The new PBR workflow means Lightwave can match industry standard surfacing. It scales extremely well on multicore workstations and network rendering is easy and free.

I do think there are too many settings to chase down. They should shoot for Frime progressive. Essentially no settings at all.

If LW is nothing but an animation and surfacing platform for Octane then it provides a fast, affordable solution. Have you compared the viewport performance to Modo? It's easily twice as fast for deforming meshes. Turn on subds on heavy meshes and it beats Max performance with similarly subdivided meshes. And this is 1.0 for the new engine.

The nodal surfacing allows exceptional control over lines and cell shading looks. If it looks like crap it is a limitation of time and artistry, not the software. Productions must always balance cost and quality:
https://www.google.com/amp/s/io9.gizmodo.com/the-first-trailer-for-the-batman-ninja-anime-is-a-glori-1820907529/amp

If people have constructive criticism, fine. But Steve should delete the posts with nasty hyperbole and enforce the forum rules. Folks have had fair warning. Be courteous or be banned...or, better yet, just move on.

All programs have strengths and weaknesses. Lightwave is a balance of efficient solutions and workarounds, just like the others. Embrace the learning curve for the new tools and kindly let NewTek know what could be better.

Little things like adding tone mapping to VPR would make a huge difference to day to day workflow.

gar26lw
01-04-2018, 05:42 PM
I do think there are too many settings to chase down. They should shoot for Frime progressive.

yes, progressive rendering and the ability to zoom in. i am not sure why they did not put those two features into 2018. they are there in modo for a reason. miss fprime; i wonder why worley stopped dev?

i would like to know the next focus of development, like the foundry is doing.
then we could focus input on those areas and maybe make more of a difference and be more constructive.

would be nice to see the image processing be part of vpr too. if you could pause the renderer, image processing is applied. that would need progressive rendering to work nicely.

Cageman
01-04-2018, 07:19 PM
Well, that's my whole point - all the new stuff is basically about rendering which wasn't the main area of concern with LW to begin with.

Hmm... LW renderengine was quite dated... it was basicly 20 years old. PBR workflow with tools like Substance Painter was, in the old renderengine, a huge workaround with thousands upon thousends of mouseclicks to get to something decent.

If you don't think LW renderengine didn't need to catch up, you have not used those other engines out there enough.



I mean seriously: Good Renderers are out there like stars in the Milky Way. You can buy them and plug them into whatever 3D app you like if you use one of the more used ones. What the 3D market needs the least is _another_ renderer that does the same old PBR stuff like everybody else, especially on the CPU but has no up to date infrastructure behind it to make it stand up in the current market as a complete solution to creating beautiful and innovative work from scratch in an up to date workflow.

Well... how much do you have to pay to have a farm of, lets say 200 machines, using Arnold? Modeling asside, since most of it is done, at least on our end, in Zbrush + Modo / Maya. Yes... LW is no longer the goto tool for modeling. But... the new render engine allowed us to complete a project during beta.




All the examples I've seen so far are somewhere between cheap looking crap and stuff I was able to do years ago somewhere else already. I haven't seen a single image that would make me go "wow" or make me want to try LW if I wouldn't be a user already.

And if I say that the ammount of time I put into shading things is this:

1) Load the objects
2) Load the textures exported from Substance Painter
3) Hook the textures into the BSDF Materal
4) Render
5) Done

What do you have in comparsion to that, in terms of the time you had to put into tweaking shaders etc... even for different lighting conditions, you would have to tweak the shaders... That ******** is all gone now, if you approach PBR with a proper approach. Don't give me the craptalk about not having to spend a lot of time with previous verisons of LW to tweak shaders to work with the lighting. Just don't, because if you do, you are lying.



And instead of running an open beta with good communications with the userbase it looks as if betatesting only just began and involves advance payment...

As I said, we've already finished a project using beta version of LW 2018. It is always risky to use BETA software in production, but... it turned out to be extremely stable and predictible for the renderingtasks at hand.

My take away from your hatered towards the focus on Rendering in this update, is that you are not fully aware of how an important change PBR is for making it a lot easier for the artists to achive the results.

That said, you still have the Standard Material to do things "oldschool" and go very artistic (as in not photoreal). That stuff is way harder in render engines like Arnold or Octane.

gar26lw
01-04-2018, 08:09 PM
yes, the renderer had to be updated. as cageman said, if you were using it and used other render engines, it was blindingly obvious. lightwave had lost its edge in that area and the rest of the app certainly could not make up for that. it was making the decision to jump a no brainer, even thought there would be a learning curve and you would be away from the beloved lightwave ui paradigm. now, they need to sort out the rest of the app a bit, especially the ux and some of the workflows as they are dragging things down.

but saying that, i agree with a lot of what you say.

OFF
01-04-2018, 08:23 PM
What, in 2018? of course they should be (with the exception of Skin). If you can't see a texture, try using a second material going into the OGL Material input to test.
Here is Principled in OGL139164

A good illustration of the consequences of smoking! )

samurai_x
01-04-2018, 08:32 PM
Rendering and viewport performance was the focus and they delivered.

Its basically a layout focus for the 6th time since lightwave 7.
That's not the major problem for lightwave. The major problems are never addressed and that is what's making users quit on lw.

erikals
01-04-2018, 08:47 PM
i'll just refer to Cageman's post, instead of writing things over again.

http://forums.newtek.com/showthread.php?155474-LW-2018-Comments-Opinions&p=1531333&viewfull=1#post1531333

samurai_x
01-04-2018, 09:05 PM
In anycase, lightwave will be fine.
There's still 187 users who will never quit on lw. Till death do as part thing. :D
https://www.facebook.com/groups/lightwiki/permalink/1284327375006592/

VermilionCat
01-04-2018, 09:31 PM
Glossy Reflections in Photoreal shading model, what exactly does this do in technical term? I don't seem to find the explanation.

Hail
01-04-2018, 09:32 PM
Not sure if we differ that much. I have absolutely no clue about the internal structure and personage of Newtek/LW3DG so I assume that the management is in error if the software doesn't have a clear direction, since directing and focusing is usually a management job. If they don't do that, it's again their fault. Most programmers are too deep in their code to make good decisions in the general realm.

I also have no clue what led to the Luxologists leaving, but I found so much strange and oddly familiar stuff (not in a good way) in Modo in the time I used it (bought 601, tried to like it, didn't in the end) that I wondered if it wasn't a good thing to lose them. Totally over engineered piece of buggy and hard to use software. And Brad Peebler made me cringe - should probably sell used cars.

Probably it's a problem with people in general. Like William Vaughan was maybe good for selling LW to total noobs but had a big part in making LW the laughing stock of the industry with his crappy primitive Seus models. Rob Powers gave me hope there, since at least stuff suddenly looked professional. This release again is more back to the old standards of "developer art".

So in the end I probably have simply outgrown Lightwave and what it can be or wants to be years ago.

Now it feels like that hot cheerleader from back in school that you find is still a waitress in the same old diner in the same old town 30 years later...

Cheers,

Tom

For f**k sakes they replaced the geo engine, renderer and added a deform stack.
If that doesn't mean anything to you then I don't know what else will.
As you said, you've probably outgrown lightwave, so help yourself. Go get a subscription of Max or Maya which btw would probably cost you a lot more than the upgrade price of lw 2018.
Some of us still enjoy using lightwave to make money and like the direction it is headed.
So please stop spitting on the cake!
Just move on already.
You will not be missed.

jeric_synergy
01-04-2018, 09:45 PM
The new documentation is also much better designed, but it seems incomplete. It really should give at least a cursory explanation of every single tool or option that is visible as a button Or an input field in the standard lw2018 install. If it’s in the interface it needs to be in the documentation. Currently a lot of it, particularly older functions are just not even mentioned. Hopefully this will improve with user feedback.

I know it seems easy to say "If it's in the UI, it should be in the dox", but give BeeVee a hand and as you come across Omissions (as in "Errata & Omissions") and use the bug report form on the bottom of every doc page to let him know about it.

Hail
01-04-2018, 09:49 PM
The major problems are never addressed and that is what's making users quit on lw.

Isn't that why they replaced the geo engine and started porting the UI kit to QT?
Good f**king lord... you guys are so thick it is infuriating!
A dose of common sense is all it takes to see the direction in which this is all headed.
Are you all that BLIND or just playing dumb?

samurai_x
01-04-2018, 10:09 PM
Isn't that why they replaced the geo engine and started porting the UI kit to QT?
Good f**king lord... you guys are so thick it is infuriating!
A dose of common sense is all it takes to see the direction in which this is all headed.
Are you all that BLIND or just playing dumb?

Always ask for a ROADMAP as the former President of Lightwave said so. That's commons sense. :D

Chris S. (Fez)
01-04-2018, 10:13 PM
Its basically a layout focus for the 6th time since lightwave 7.
That's not the major problem for lightwave. The major problems are never addressed and that is what's making users quit on lw.

Making the renderer compatible with modern "substance" standards addressed a major problem. The nice thing is we did not lose NPR for technical animation and illustration. Plus, more critically, the channel export for render passes is beautiful and sets the foundations for improved pass management

Many pipelines outsource to Zbrush. It was a hard choice but I think Newtek have invested in the right area. Even if rendering relies on Octane, Layout provides a fast interface for lighting, nodal texturing and rendering.

It will be interesting to see if 3rd Powers can leverage the expanded SDK but I hope 2.0 of the geo engine allows native element selection and manipulation in Layout. Long past due.

I still believe unification is the goal but suspect it will come in a different form than we imagined.

samurai_x
01-04-2018, 10:35 PM
I still believe unification is the goal but suspect it will come in a different form than we imagined.

A different form of unification? :D Outside the lightwave bubble, people really don't want a different form of unification.
187 people on facebook probably don't mind at all as long as they can retire in 5-10 years and then who gives a frak about lightwave by then, they're retired. :D
I think that's their ROADMAP and business plan.

And its not gonna happen anyway. Financially not feasible to do it at this point.

gar26lw
01-04-2018, 10:39 PM
For f**k sakes they replaced the geo engine, renderer and added a deform stack.
If that doesn't mean anything to you then I don't know what else will.
As you said, you've probably outgrown lightwave, so help yourself. Go get a subscription of Max or Maya which btw would probably cost you a lot more than the upgrade price of lw 2018.
Some of us still enjoy using lightwave to make money and like the direction it is headed.
So please stop spitting on the cake!
Just move on already.
You will not be missed.

what is the direction?

Cageman
01-04-2018, 10:53 PM
what is the direction?

What Direction is C4D, Maya, 3DS Max heading in? Houdini... ? Can you give me a solid example, except Blender (which is a free and community driven software), where the roadmap is layed out?

Can you tell me what features Maya 2019 will have, for example?

Chris S. (Fez)
01-04-2018, 11:01 PM
I think unification will take place incrementally over the next few releases. I'd guess we'll soon see Weight Painting and element manipulation for sculpting, morph compensations and creating clusters with falloff etc.

For the next release I imagine we'll get rendering refinements, an optimized geo engine 2.0, a QT interface port along with a modernized Graph Editor, Scene Editor and perhaps Render Pass Management.

But unified modeling tools...a ways off. In the mean time Modeler might be optimized and the Hub enhanced.

samurai_x
01-05-2018, 12:17 AM
What Direction is C4D, Maya, 3DS Max heading in? Houdini... ?

I think they're trying to split their appz like lightwave because majority of 3d users like split appz. :D

So you don't have to work with fluids natively. :D


https://s13.postimg.org/czd1sigqv/25498124_1737689416276291_4289830477361139974_n.jp g

Wickedpup
01-05-2018, 12:19 AM
does layout understand points, edges and polys now?



yeah, that is one thing I would like to know....

samurai_x
01-05-2018, 01:12 AM
I think unification will take place incrementally over the next few releases.

I think you will sing a different tune in a few years in this forum and the foundry forum.
Lets see what lightwave 2021 will have. :D
I have accepted it can't be done.

hypersuperduper
01-05-2018, 01:27 AM
Wait. I didn’t know this was the unification debate thread. Oh yeah I forgot, EVERY thread is the unification debate thread.

OFF
01-05-2018, 01:34 AM
If we try to believe in modern science, then, based on the fact that the universe occurred as a result of an explosion (who planted the bomb?) and all its parts are constantly flying away - the nature of reality in the division of everything and everything into fragments. Consequently, Newtek's desire to divide LW - strictly corresponds to the laws of the universe (if we believe all the stupid things that scientists are talking about). :p

VermilionCat
01-05-2018, 02:06 AM
The UI behaves really amateurish. Do something better next time.

Chris S. (Fez)
01-05-2018, 02:14 AM
I think you will sing a different tune in a few years in this forum and the foundry forum.
Lets see what lightwave 2021 will have. :D
I have accepted it can't be done.

I don't think NewTek is going to abandon modeling but it is obviously no longer a priority. They had a head start by pioneering subds and such but have since lost the advantage.

3rd Powers + TrueArt + LW cad serve many of my needs in Modeler and Layout. I still prefer Modeler for most projects but there are clearly superior options available for UVs, sculpting and procedural nondestructive modeling.

Maintaining Modeler is hopefully not impeding progress. For instance, if they can implement open subds in Layout then they should not hold back just because implementing subs in Modeler is problematic.

As far as I am concerned, Lightwave is a pipeline tool. I'd prefer a complete package but if development is stretched too thin then they need to at least temporarily prioritize on compatibility with go-to tools like Zbrush and practically universal formats like open subds.

If the new geo engine is truly point, edge, poly aware then they can methodically add modeling over the next decade. If not then...they should delete that blog post asap.

djwaterman
01-05-2018, 02:46 AM
Did you see Andrew Comb's video post about using the environment light? I know it's not the same.. but maybe you can do a workaround somehow. I'm planning on keeping certain projects in 2015 until I get up to speed and comfortable using 2018 for a full-feature project. Gonna be a big learning curve!

Can someone direct me to a link for that video?

Mastoy
01-05-2018, 02:53 AM
Can someone direct me to a link for that video?

https://youtu.be/72AAwSFx4nA

kopperdrake
01-05-2018, 02:56 AM
A different form of unification? :D Outside the lightwave bubble, people really don't want a different form of unification.
187 people on facebook probably don't mind at all as long as they can retire in 5-10 years and then who gives a frak about lightwave by then, they're retired. :D
I think that's their ROADMAP and business plan.

And its not gonna happen anyway. Financially not feasible to do it at this point.

What is it with you and the age thing samurai? So many posts you stick on here refer to the old timers, old guys etc. Broken record.

gar26lw
01-05-2018, 03:10 AM
i don’t know bout broken records, broken cds perhap?

MarcusM
01-05-2018, 04:04 AM
I must confess, not only expand selection is faster in Modeler 2018.

Imported in Modeler 125 Mb fbx, dozens materials, dozens layers, close to 2mln polygons, GTX 770, shading method - MultiTexture.
Highest average FPS during rotating in perspective.

LW 2018.0 - 111 FPS
LW 2015.3 - 4 FPS
LW 2015 - 7 FPS
LW 11.6.0 (ver. 2723) - 261 FPS
LW 11.6.0 (ver. 2714) - 263 FPS
LW 11.6.0 (ver. 2706) - 263 FPS
LW 11.5 - 116 FPS
LW 11.0 - 15 FPS

MichaelT
01-05-2018, 04:25 AM
(in regards to unification post earlier) Are we having this talk *again*. (rolls over in disbelief with a planetary facepalm)

VermilionCat
01-05-2018, 04:34 AM
Spot Info node is now gone? Why??

Ztreem
01-05-2018, 05:10 AM
I think unification will take place incrementally over the next few releases. I'd guess we'll soon see Weight Painting and element manipulation for sculpting, morph compensations and creating clusters with falloff etc.

For the next release I imagine we'll get rendering refinements, an optimized geo engine 2.0, a QT interface port along with a modernized Graph Editor, Scene Editor and perhaps Render Pass Management.

But unified modeling tools...a ways off. In the mean time Modeler might be optimized and the Hub enhanced.

Soon as in 3-6 years? People has been waiting for weight painting in layout since lw 7 and still it’s not there...a proper undo???
In lw 8 we got dynamics with hardfx (almost unusable for any quality work), LW 9 gave us a rendering update. Lw 11 & 2015 gave use dynamics again(Bullet). 2018 rendering update. I’m starting to see a pattern. Workflow wise LW gets very little attention and it starting to get slower and slower to get things done.

hypersuperduper
01-05-2018, 05:40 AM
Soon as in 3-6 years? People has been waiting for weight painting in layout since lw 7 and still it’s not there...a proper undo???
In lw 8 we got dynamics with hardfx (almost unusable for any quality work), LW 9 gave us a rendering update. Lw 11 & 2015 gave use dynamics again(Bullet). 2018 rendering update. I’m starting to see a pattern. Workflow wise LW gets very little attention and it starting to get slower and slower to get things done.

2018 did add a number of workflow improvements, no question. Multi select copy paste of item properties for example that in itself is a pretty big workflow change. Just try copying a displacement node network from one object to like 20 others in 2015.3 it’s step by step tedium. In 2018 you can copy/paste and reorder modifiers on multiple selection items. It works like a charm. You can’t edit them for multiple yet, but copy/paste is a godsend. That is nothing if not a workflow change. I would argue that after rendering, the workflow fixes were the next biggest change.

Works with nodal motion modifiers as well.

ianr
01-05-2018, 06:00 AM
LW 2018 WRINKLES (TIPS). (from lighwiki resource findings)

1)Add Viewport.... Now you can have a little VPR viewport off to the side while working in Modeller
and see whats happening with a quick hit of the sync button, with tools still active, too

2) Render Times increase.....

The CFG always lives on the C-drive Remember ? It needs a little update....(hopefully fixed in next point-patch)

The LW2018.0-64.CFG must be changed manually! (Lightwave and Hub must be closed for that! REMEMBER)

open in Word Pad > search for Segment Memory. You will see only set @ 1GB as default.

Alter/increase GB to 50% of your system Ram.

Save the file!!!

Close Word Pad ..>>Restart Lw2018

Thanks to : Ingo Rotfuchs & Mark Warner:jam::jam:

Chris S. (Fez)
01-05-2018, 06:01 AM
3rd Powers weight painting works great but I bet we see native weight painting and rudimentary element selection and manipulation in about a year.

I'd like to see bug fix and optimization releases every few months along with a yearly release. For this first year I think it is reasonable to expect a .5 release in 7 months or so with a focus on QT interface updates, rendering refinements, and geo-engine performance.

Ztreem
01-05-2018, 06:04 AM
2018 did add a number of workflow improvements, no question. Multi select copy paste of item properties for example that in itself is a pretty big workflow change. Just try copying a displacement node network from one object to like 20 others in 2015.3 it’s step by step tedium. In 2018 you can copy/paste and reorder modifiers on multiple selection items. It works like a charm. You can’t edit them for multiple yet, but copy/paste is a godsend. That is nothing if not a workflow change. I would argue that after rendering, the workflow fixes were the next biggest change.

Works with nodal motion modifiers as well.

That sounds good, even though a global node editor would be preferd. When you copy a motion modifier node from one object to another does it add that modifier or does it replace it if its allready there?

Hail
01-05-2018, 06:06 AM
what is the direction?

Modernizing the core, replacing old tools with new tools with better efficient workflows

hypersuperduper
01-05-2018, 06:09 AM
That sounds good, even though a global node editor would be preferd. When you copy a motion modifier node from one object to another does it add that modifier or does it replace it if its allready there?
You can pick. There is paste and paste into.