Page 1 of 4 123 ... LastLast
Results 1 to 15 of 50

Thread: A message for those still holding on to LW 2015....

  1. #1
    Super Member
    Join Date
    Sep 2003
    Location
    Washington DC
    Posts
    3,648

    A message for those still holding on to LW 2015....

    Please upgrade.
    We were holding on to upgrade because of a ton of legacy projects we were still supporting. But this month we bit the bullet and upgraded.

    And boy we have been quite impressed. And I really like how easy is for me to do nice reanders with the new render engine.

    I wonder if this experience is shared by others.

    _R
    This message does not reflect the opinions of the US Government, CG Networks or CGTALK.com. The opinions expressed on this posting are on my own volition.

  2. #2
    RETROGRADER prometheus's Avatar
    Join Date
    Aug 2003
    Location
    sweden stockholm
    Posts
    15,566
    Quote Originally Posted by robertoortiz View Post
    Please upgrade.
    We were holding on to upgrade because of a ton of legacy projects we were still supporting. But this month we bit the bullet and upgraded.

    And boy we have been quite impressed. And I really like how easy is for me to do nice reanders with the new render engine.

    I wonder if this experience is shared by others.

    _R
    Mixed feelings, for some product renders, great..when it comes to fireflies, the workflow isnīt as straightforward to deal with certain surfaces, like glossy surfaces.
    And fiberfx..I can not use due to itīs very very slow rendertime.

    It is in fact the slower rendertimes, and fireflies and how it deals with noise that is pushing me to use blender cycles more, so No I do not think the new render engine has been as succesful as it could have been.
    Workflow wise I think it took some steps back.

    New volumetrics, much more realism of course, but a step back in workflow and a loss of certain features, much more tedious to use a hypertexture...which now requires entering nodes, you can not add the new volumetrics to points either..which is very bad, and you do not have any velocity effects either, sprites are not implemented, you have to use the older hvīs..but you can not render it in vpr with newer volumetrics if you have clouds and if you have sprites..that is a no no and requires rendering in passes.
    Particle age setup is much tedious to set, as well as random particle size...there is no way to controll it other than entering the actual emitter and set a -+ scale on the particle emitter, instead of as with old hv a simple random hv size.
    You could go nodal and set up nodes for the randomnes, but you see..that is also an evidence that some workflow took some steps back.

    and surfacing is also lacking that direct connection, if using pbr materials..always requires a nodal setup for simple images, and bumps.

    For realism ..much better
    For workflow and speed....not that exciting.
    You could of course argue that the new materials will get you there faster without tweaking the surfaces so much, true..but was it necessary to avoid implement a better connection
    in the surfacing and in the volumetric setups?

  3. #3
    Electron wrangler jwiede's Avatar
    Join Date
    Aug 2007
    Location
    San Jose, CA
    Posts
    6,672
    Quote Originally Posted by prometheus View Post
    You could of course argue that the new materials will get you there faster without tweaking the surfaces so much, true
    Agreed, for a subset of materials, anyway. Materials with glossy/refl/refr surfaces often appear to require more tweaking now, as do materials requiring significant SSS contribution. So while many types of materials "get you there faster without tweaking the surfaces so much", many others appear to now require at least as much (if not more) tweaking as before.

    That said, moving from LW2015 to LW2019 adds quite a bit more than just the new render engine. While I do agree the new render engine introduced some workflow problems, new feature capabilities over LW2015 such as OpenVDB support add quite a bit of upgrade value as well, yielding (IMO) a net benefit upgrading from LW2015 to LW2019.
    Last edited by jwiede; 11-21-2019 at 03:42 PM.
    John W.
    LW2015.3UB/2019.1.4 on MacPro(12C/24T/10.13.6),64GB RAM, NV 980ti

  4. #4
    Registered User
    Join Date
    May 2012
    Location
    Virginia
    Posts
    474
    Quote Originally Posted by jwiede View Post
    Agreed, for a subset of materials, anyway. Materials with glossy/refl/refr surfaces often appear to require more tweaking now, as do materials requiring significant SSS contribution. So while many types of materials "get you there faster without tweaking the surfaces so much", many others appear to now require at least as much (if not more) tweaking as before.
    Definitely agree here. Specular and SSS seem to take a fair amount of tweaking, in particular, at least in my experience.

    Rendering is still agonizing; the renders themselves are beautiful, but getting there is just punishing, especially when you are trying to generate a few hundred frames of animation.

    I've said it before, but I'll say it again; a native GPU render solution would take the sting out of so many processes in one fell swoop.

  5. #5
    RETROGRADER prometheus's Avatar
    Join Date
    Aug 2003
    Location
    sweden stockholm
    Posts
    15,566
    Quote Originally Posted by jwiede View Post
    Agreed, for a subset of materials, anyway. Materials with glossy/refl/refr surfaces often appear to require more tweaking now, as do materials requiring significant SSS contribution. So while many types of materials "get you there faster without tweaking the surfaces so much", many others appear to now require at least as much (if not more) tweaking as before.

    That said, moving from LW2015 to LW2019 adds quite a bit more than just the new render engine. While I do agree the new render engine introduced some workflow problems, new feature capabilities over LW2015 such as OpenVDB support add quite a bit of upgrade value as well, yielding (IMO) a net benefit upgrading from LW2015 to LW2019.

    I think my main reason for upgrading, though I wasnīt happy about fiberfx rendering, or the new workflow in surfacing and in volumetrics...was in fact the new promising VDB tools and itīs many features which is indeed exciting..but it also feels like itīs in the cradle state when it comes to OpenGL display and simulation speed, improving that and some framework around the tools would very nice..but it really needs to take off at a level where it can compete with other fluid tools, such as TFD, or blender fluids at least...for my interest of maintaining a possible future upgrade investment.

    The VDB tools are there to import and export some of the volumetrics in lightwave..where blender I think doesnīt have it that so simply implemented to import, unless using octane...not sure about the latest experimental builds in blender..will have to check that up.

  6. #6
    Super Member
    Join Date
    Sep 2003
    Location
    Washington DC
    Posts
    3,648
    Thanks for the great replies. BTW to the Forum leader, I did not mean to post my message here, I wanted to post it here:
    https://forums.newtek.com/forumdispl...8-LW-Community
    This message does not reflect the opinions of the US Government, CG Networks or CGTALK.com. The opinions expressed on this posting are on my own volition.

  7. #7
    I think the only thing I miss from LW 2015 is NPR-style rendering, where usually you could get 2-4 seconds per frame renders if your scenes were on the light side (simple diffuse surfaces/cel-shading). Right now I reserve Lightwave renders for visual effects work for live film, since beefier systems like Octane tend to add too much overhead to be worth it for that purpose. I would like to be able to do full CG productions entirely in Lightwave in the future, the renderer just isn't quite there yet for animated content of that nature.
    Professional-level 3d training: Ryan's Lightwave Learning
    Plugin Developer: RR Tools for Lightwave

  8. #8
    RETROGRADER prometheus's Avatar
    Join Date
    Aug 2003
    Location
    sweden stockholm
    Posts
    15,566
    I miss volumetric lights in sprite mode, and the direct acess of hvīs ..and a sprite mode for that as well to work with the other render engine in one go, but itīs the same for blender..the blender old engine has sprites for rendering particles a certain way..which you can not do in cycles, I guess itīs something in that nature of the PBR engines that makes it difficult to implement things like sprites.

    The volumetrics lights in sprite mode I used to whip up some nebula, or just plain space gases stuff that looked very nice and rendered within a few seconds, while not fully volumetric and at the level of what you have to use today with true full volumetric items, it was enough for certain shots..but now you just can not get away with anything less than 35-minutes and probably more per frame..unless using octane that is.

    So basicly itīs some of the sprite functions in the old renderengine from hvīs and volumetric lightīs I miss, though we can employ hv sprites seperately ..but not at the same render pass as the other elements in the scene.

    Also miss the hypertexture effect speed, velocity mostly and viper to preview it fast.
    Viper is gone
    the volumetric light in sprite mode is gone.
    hypertexture speed effect is gone....you could do other setups to simulate the effect but takes more time to set up, either with null referencing or plug speed time node in to position slots for example.

    I also did some little effects with growing networks of polylines and hypervoxels sprites, and clip mapping to map symbols, and make point joints and then also have some points expand with hypertextures and the hypertexture effect..that was kind the charm of old hypervoxels that it was probably the Easiest and fastest systems to work with compared to other software when setting up sprites or volumetrics, you could compared to modo, blender, houdini...none of those did the workflow easier than Lightwaves hypervoxel menus in the legacy HV menu.

    So I am a bit sad the workflow of the newer system went the way it did, by not implementing sprites,an not implementing volumetrics that can be applied on to points or vertexes, only nulls and particles.
    and more difficult to keep track on the elements you edit, you are required to select the element in order to edit it..using the scene editor is a must for that now, but you can not move other objects while keeping focus on the volumetric editing, the old hv menu allowed you to move whatever you want in the scene while keeping the editing settings live in front of you in itīs own window.

    As I have said before, I am not complaining about the realism in the volumetrics..which of course is much better, Initially I complained about renderspeed at larger sizes, still not the best..but I have come to terms with how it mostly work with the step length.

    I am curious to how they look at rendering of particles and items in sprite modes, if they plan and if they actually can implement it in this new Modern render engine.

    As for surfacing, I donīt like having to double click and feed image nodes in to every pbr material, and I donīt like the lack of no default projection setup when trying to use anisotrophy..you have to set it up each time, though you should probably adapt to using an anisotrophy preset already having that connected, not sure if there was a preset material shipping with that?..otherwise they should build a couple of those to help newbies.


    And glossy surfaces..they need to do something about that.

    Undo system
    UI fixing
    and modeling in to layout next please, direct editing of curves and making them 3d shapes would be nice, animatable..none destructive and supporting svg formats.

  9. #9
    RETROGRADER prometheus's Avatar
    Join Date
    Aug 2003
    Location
    sweden stockholm
    Posts
    15,566
    Quote Originally Posted by Ryan Roye View Post
    I think the only thing I miss from LW 2015 is NPR-style rendering, where usually you could get 2-4 seconds per frame renders if your scenes were on the light side (simple diffuse surfaces/cel-shading). Right now I reserve Lightwave renders for visual effects work for live film, since beefier systems like Octane tend to add too much overhead to be worth it for that purpose. I would like to be able to do full CG productions entirely in Lightwave in the future, the renderer just isn't quite there yet for animated content of that nature.
    I wonder what nature content it is you want? can you ellaborate on that?..if you got the time and can keep it short

  10. #10
    Quote Originally Posted by robertoortiz View Post
    Please upgrade.
    We were holding on to upgrade because of a ton of legacy projects we were still supporting. But this month we bit the bullet and upgraded.

    And boy we have been quite impressed. And I really like how easy is for me to do nice reanders with the new render engine.

    I wonder if this experience is shared by others.

    _R
    Yeah, the render engine is simply gorgeous - but slooooow. Actually the first pass it takes is really pretty fast, its the AA afterwards that is slow. I do interior stills and product shots (no animation) so TBH the slow speed isn't terribly annoying, it's the crashing while rendering in 2019 that is killing me. The shading system is excellent! So much easier to get great looking materials than in previous versions and the patina and edge shaders are very powerful, plus throw in the free Anything Goos and you can really do some amazing things. The interface improvements are welcome too. I like the Flat Menu too - find anything really fast.
    Tim Parsons
    Technical Designer
    Sauder Woodworking Co.

    http://www.sauder.com

  11. #11

    I think the only thing I miss from LW 2015 is NPR-style rendering
    it's a bit yes & no. LightWave 2019 can pull off NPR that 2015 didn't stand a chance to pull off.
    but sure understand what you mean, in regards to rendertimes.

    its the AA afterwards that is slow.
    Yep!  
    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

  12. #12
    RETROGRADER prometheus's Avatar
    Join Date
    Aug 2003
    Location
    sweden stockholm
    Posts
    15,566
    Quote Originally Posted by erikals View Post

    it's a bit yes & no. LightWave 2019 can pull off NPR that 2015 didn't stand a chance to pull off.
    but sure understand what you mean, in regards to rendertimes.


    Yep!  
    Thanks for sharing your vids again Erikals, Despite this thread urging to upgrade... I think I would advise you to stick to 2015 for a while though, and instead save your money foremost for a new computer, so you do not have to stand there in shame when presenting something on
    youtube it will also make it more attractive to watch your vids where you can focus only on other tasks ahead, though in this case your old computer may be related to the tips of cutting rendertimes.

    So ..look at new computer..a mix of good nvidia gpu card and cpu, for any future Lightwave, meanwhile you can still have fun with blender with much new techology anyways, PBR Materials are there as well and itīs not much difference in techology, though I do not now how your computers GPU is handling cycles?
    And while you are buffering and waiting for mister dollar, I think we may see another important feature release for Lightwave, and you can get a better picture of where you think it is going with improving the renderer, and if there soon or ever will pop up some Layout modeling tools.

  13. #13
    Electron wrangler jwiede's Avatar
    Join Date
    Aug 2007
    Location
    San Jose, CA
    Posts
    6,672
    Quote Originally Posted by erikals View Post

    it's a bit yes & no. LightWave 2019 can pull off NPR that 2015 didn't stand a chance to pull off.
    What NPR functionality do you believe LW2019 offers, that weren't doable in LW2015 with its NPR support (incl. UnReal Xtreme2, and nodal control over object edge rendering)?

    I'll just remind folks there were threads like this one, when LW2018 released. While LW2019 subsequently improved certain NPR surfacing aspects, and offered much, much superior NPR preview versus prior versions, there are still areas of NPR (esp. w.r.t. object-level edge rendering controls, etc.) where LW2015 offered more direct/efficient solutions than either LW2018 or LW2019.
    Last edited by jwiede; 11-24-2019 at 08:01 PM.
    John W.
    LW2015.3UB/2019.1.4 on MacPro(12C/24T/10.13.6),64GB RAM, NV 980ti

  14. #14

    What NPR functionality do you believe LW2019 offers, that weren't doable in LW2015 with its NPR support
    I know 2019 offers a better render overall.
    - Better Metal shading
    - Way better FiberFX renders
    - Way better Volumetrics in terms of fog/sunrays

    tho sure, lots can be done in 2015 also, and perhaps even faster / easier.

    doing something close to this was impossible in 2015.



    ok, not saying 2019 can match that 100%, but definitely waay closer in 2019 than 2015.

    and the LW2019 renders are superfast....!   [insert ironic line here]  
    Last edited by erikals; 11-24-2019 at 08:07 PM.
    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

  15. #15

    Thanks for sharing your vids again Erikals, Despite this thread urging to upgrade... I think I would advise you to stick to 2015 for a while though, and instead save your money foremost for a new computer, so you do not have to stand there in shame when presenting something on youtube
    u Welcome!  
    and I've got a i7 3930k standing in the corner [turns head in shame...] haven't had time to move the hardware, install Win7 etc etc...

    So ..look at new computer..
    Overclocked Ryzen 1700 is next up, should be 600% times faster.


    though I do not now how your computers GPU is handling cycles?
    just fine.  

    And while you are buffering and waiting for mister dollar, I think we may see another important feature release for Lightwave, and you can get a better picture of where you think it is going with improving the renderer, and if there soon or ever will pop up some Layout modeling tools.
    yes, with my starving artist mentality, there won't be any upgrade before 2020.  
    LW vidz   DPont donate   LightWiki   RHiggit   IKBooster   My vidz

Page 1 of 4 123 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •