PDA

View Full Version : Does Lightwave render show its age?



Original1
03-02-2004, 07:29 AM
Take a look at this article listed on Flay

http://www.zaon.com/company/articles/3d_rendering.php

This is a direct quote:


Lightwave: Lightwave's native built-in renderer is the only native renderer actually suitable for production film work, but its core architecture is definitely now showing its age. Lightwave rendering is slow compared to all of the other options available, and while it can produce wonderful results it has more shortcomings than the other highend rendering options—not necessarily surprising, given its age and native development. Lightwave's worst shortcomings are foremost its speed, closely followed by texture filtering, motion blur, and displacement quality. Still, the pricing (free with Lightwave and with unlimited network farming licenses) and tight integration with the core package are great strengths.



Please before we go too far the idea is not to start a flame war, but a sensible disscussion of HOW what we have can be improved.

I'm curious, is Lightwave renderer slower because of the 128 bit pipeline.

On Viper

Looking at Worley's latest offering its clear there is room for improvement.

What about integrated model painting like bodypaint,editing(painting) of Open exr and HDRI images or Lightwaves flx format, b-spline and edge modeling.

meshmaster
03-02-2004, 08:29 AM
some of that article may be true... I've seen the article before and think it's biased against lightwave and doesn't tell the whole story.

jin choung
03-02-2004, 08:30 AM
i thought that article was apt and telling as well. but i think the answer is basically so simple as to preclude asking the question:

IT'S ALWAYS ABOUT THE MONEY.

sure, we can go into details and point out such things like it would be nice if the lw renderer can be like REYES renderers and render higher order surfaces directly into pixels so as to get really nice displacement and such.

sure, we could talk about researching the latest and greatest technologies... heck, we can even get people who regularly read and understand siggraph papers.

sure, we could talk about doing a ground up rewrite and just radically re-architect the whole thing like what maya is to power animator and what xsi is to si.

BUT ALL THAT IS IRRELEVANT. WHAT YOU WANT TO TALK ABOUT IS IRRELEVANT.

BECAUSE IT'S ALWAYS ABOUT THE MONEY.

everything that you can think of that will improve beyond tweaks (like a really comprehensive rewrite of the renderer) takes a lot of money.

so at that point, you run into the issue of - ok, we can charge more but then we're going more head to head with the pricing of other apps like maya or max. if we change our prices drastically, will we INCREASE the money coming in or DECREASE it.

for me, one of the overriding reasons why i went with lw is INDEED THE PRICE and the everyman philosophy of newtek and tim jennison.

so pick one - increase the price or make do. i pick make do and if the pricing changes radically, i've just been raised out of the game and will have to fold.

that's my choice. others, including you may choose different. but just realize that NOTHING can come from NOTHING.

jin

mattclary
03-02-2004, 08:35 AM
Well said, Jin.

Ade
03-02-2004, 08:38 AM
How come Lightwave doesnt have Photon mapping when companies like maxon have implimented it into cinema4d?

Igu4n4
03-02-2004, 08:48 AM
While I've always enjoyed the LW rendering engine, and have been using it for 5 years now I believe, it does have several inherent faults, and that is due to age and perhaps a bit of technological stagnation. However, it also satisfies "most" needs, and more so it "is" a packaged solution.

I think the problem for many lightwavers is that it is the "only" solution (Fprime shows promise, but it's yet to see if this will be a viable alternative). I completely agree with Jin, creating a new renderer and packaging it with LW will move the scale of price/performance, and that is/has been one of LW's strongest points. However, I think you could satisfy both needs if the architecture was in place, and friendlier, to 3rd party rendering engines. The choice is then back to the consumer/user whether or not they choose to invest in a different rendering mechanism.

Obviously I'm over simplifying as creating that seamless communication pipeline for rendering alternatives might or might not be a cost prohibitive venture, and further it isn't as simple as just plugging a new rendering engine in as there is many layers of communication involved. But with more companies specializing in purely the rendering engines, it seems like a reasonable assumption to let them... and work with them to have that functionality available.

Ford doesn't put the best Audio packages in their vehicles, but they do the job. If you want more function/fidelity you go and buy a Pioneer or an Alpine, it shouldn't affect the initial price of the car, but it will cost more if you choose to change it. But then you have the option of listening to Britney Spears in enriched high-fidelity audio :P

(I'm sure this analogy will come back to bite me ;)

Steve.

Ade
03-02-2004, 08:52 AM
Good point...I do want the latest rendering features, lets hope Newtek will act soon..Didnt take c4d that long to get the newest stuff in their renderer. Maybe LW should get more updates rather than .5's every 1 year?

jin choung
03-02-2004, 09:01 AM
hey, i think the $300 HASH animation master has photon mapping too.

why? ya ask why? hey why does lw have things that cinema 4d doesn't have like skytracer or hypervoxels?

why?

that's a silly question. the answer is clearly:

BECAUSE!

------------------------------------------------------------------------------------

again, it's kinda pointless to ask. the answer will invariably be money. not everyone in programming can do everything. given that, the developmental path of an app can (especially in small companies) evolve along the lines of the developers proficiencies and not necessarily in the idealist 'BEST PATH'. again - money.

also, if you plunk enough R&D dollars into one path, it's tough to scrap it and start from scratch. we live in a world where you have to lie in the bed that you've made. in the last company i worked for, my personal mantra and the mantra that i always stressed to my coworkers is that no matter what management tells you, YOU HAVE TO GUESS RIGHT THE FIRST TIME! well, not if you're a huge company that allows a lot of room for PLAY.... but if not, you have to guess right the first time.

it's always about money. even when it comes to programmer robustness, competence, depth and breadth etc - even IGNORANCE and INCOMPETENCE are functions of money.

seriously people... it's not much more exotic than that. it's not mysterious and it's not INTERESTING.

why? money.

jin

Panikos
03-02-2004, 09:13 AM
Ade, Lightwave HAS photonmapping.
;)

jin choung
03-02-2004, 09:16 AM
ade,

guess what? more updates = more money!

and that does get filtered down directly to the user. again, that may be preferable to you. have an update every half year that you pay $500 for. alias seems to be on such a frequent for more track. but then that would raise many (including me) out of the 3d market.

and then, the question becomes, if they DO raise prices, will they make more or less money.

another relevant question is that if they raise prices, can they becomes BEST IN CLASS... ever? is that even a realistic possibility? if not, then you have an app that may cost about the same as maya but is not as refined... and that may not be a winning solution.

listen, lw's roots are DEFINITELY SHOWING and it has for a while now. it's heritage from the amiga toaster days have indeed followed it until now.

heck, i took to lw instead of 3ds max during the 5.6 days because lw reminded me of the DOS based 3d studio!

much about lw IS primitive. and it's toaster heritage is REALLY limiting the app starting with version 6. the interface is just haphazard and it is NOT WELL DESIGNED.

but given all that, i think that lw occupies a nice NICHE. sure, it's not as well thought out and designed as maya or xsi. don't kid yourselves folks - lw is NO ONE'S IDEA OF REFINED SOFTWARE DESIGN. it is very much currently a frankenstein's monster with odds and ends grafted onto whatever space there is left. but it IS cheap and it IS primitive - and what that means is that it IS simple - simple enough so that one person can know the whole shebang inside and out to almost its complete depth.

same would be difficult for any maya user to claim. and so, you have an affordable app that can get you where you need to go - just maybe with more blood sweat and tears. that is a good tradeoff imo.

so, my evaluation is that lw's PRIME SELLING POINT is its 1) SIMPLICITY AND ACCESSIBILITY and 2) PRICE.

and because it works with polys, it can INTEROPERATE in a multi app environment... and this is one thing that just totally killed hash animation master.... they went for ZERO POLYS as a selling point but man that messed up their potential in the mass market.

and i don't know c4d but the fact that it's not used much in big budget or famous projects may be indicative of something.
------------------------------------------------------------------------------------

if lw decides to maintain their current selling points, the BEST THING THAT THEY CAN DO is like one of you said - OPEN UP THE ARCHITECTURE so that it can plug and play with other renderers, animation apps, modelers, etc.

even that would cost money but that would probably be the best bang for buck solution.

jin

Jockomo
03-02-2004, 11:11 AM
Ya know, every time I see that guy's site, I think there is something fishy about it. Anyone actually know someone who works for Zaon, or know a real project they worked on?

Original1
03-02-2004, 11:14 AM
Originally posted by jin choung
much about lw IS primitive. and it's toaster heritage is REALLY limiting the app starting with version 6. the interface is just haphazard and it is NOT WELL DESIGNED.

jin

I think I might not totally agree there Jin.

D'ont forget 6.5 was supposed to be a complete rewrite of the code base.

but 6.0 introducced HDRI which has only just gone into version 6 of 3D max

Its an interesting dynamic, if you change too much you alienate your user base. If you don't change enough you fall behind.

There have been a number of disscussions regarding the departure of some bods to Luxology, but who wrote the code for the renderer, and how can it be improved?

for example do we need photon mapping?

If Cinema 4D is so great, how come its market penatration, (going by the article at the start of the thread) is so low.

I'm not saying Newtek doesn't need to change the $64,000 question is how and why.

Intelligent answers on a postcard please:D

Original1
03-02-2004, 11:17 AM
Originally posted by Ade
Good point...I do want the latest rendering features, lets hope Newtek will act soon..Didnt take c4d that long to get the newest stuff in their renderer. Maybe LW should get more updates rather than .5's every 1 year?

There have been a, b and c updates along the way, all free
and the .5 update was free.

unlike others who charge you for .1 updates

The Issue is that there has been a hiatus in real development in all 3D apps for the last couple of years No one has really been breaking new ground.

Phil
03-02-2004, 12:44 PM
The way I have tended to view this is that the major vendors' provision of mental ray as one of their default render systems in new releases is a recognition that R&D for this part of the system is becoming less cost effective.

The R&D has already been done by mental images and it's far cheaper in terms of programmer time and release schedule slippage to simply provide a translation system between your animation product and the mental ray renderer. It keeps your customers happy, the output from different packages identical and it also provides distributed rendering.

I'm not sure how much of LW would be compatible with this kind of thing, though, given that both maya and max have issues with mental ray rendering for certain features they offer. HVs might be an issue, but assuming there's no translatable method, there's always the option to use dense sprites instead. *shrug*

GeorgeDittmar
03-02-2004, 02:23 PM
the guy that owns zaon hangs around scifi-meshes sometimes. they are based in washington and right now are working on a rpg book.

mlinde
03-02-2004, 02:56 PM
Originally posted by jin choung
everything that you can think of that will improve beyond tweaks (like a really comprehensive rewrite of the renderer) takes a lot of money. I'm not going to dispute this, but I have a question for you. Would you pay the standard $495 upgrade for Lightwave if the only upgrade was a new rendering engine? At some point LW has to undergo significant rewrites. All software does, or it ends up like Microsoft Office.

Second comment jin -- you hate the interface, but you aren't putting forth anything better. From what I've seen, the folks at NT actually listen to us, so put forth some ideas, if you hate it so much. I haven't had the opportunity to work on XSI or Maya, but the 3D software I have spent time in besides LW was not as feature-rich because the features didn't fit easily into the interface. I'd rather have a huge, dented stainless steel box of tools than a small, pretty box with two screwdrivers and a hammer.

DarkLight
03-02-2004, 03:20 PM
Originally posted by Phil
both maya and max have issues with mental ray rendering for certain features they offer.

This is mainly due to the fact that mental ray had been added and they had to write convertors for any texture/shader features for anything that is not compatible with mental ray. Softimage gets around this problem by not having thier own renderer and developing the app to support mental ray compatable features.

I'm sure if Newtek were to add support for 3rd party renderers such as MR, then they would come across the same problems as both Max and Maya.

Ade
03-02-2004, 04:08 PM
Originally posted by Panikos
Ade, Lightwave HAS photonmapping.
;)

Since when? and How?

DarkLight
03-02-2004, 04:15 PM
Originally posted by Ade
Since when? and How?

As far as i'm aware, photonmapping is and always has been part of Lightwave's radiosity solution.

mkiii
03-02-2004, 04:28 PM
Originally posted by Ade
How come Lightwave doesnt have Photon mapping when companies like maxon have implimented it into cinema4d?

Advanced render is an extra module that costs an additional £359 to the base cost of the app unless you buy the XL bundle, which costs more than Lightwave, at £1449 & very nearly as much as Maya Complete. Don't get me wrong - C4d is a v nice app, but I don't think the renderer is any better than LW just yet - altho' who knows what the future will bring. It has improved a great deal over the last couple of years.

And isn't Photon Mapping just another name for Radiosity rendering, calculating the way light bounces slightly differently?

riki
03-02-2004, 04:39 PM
I think Newtek has the vision to make it happen. But more likely in their own time.

peteb
03-02-2004, 05:19 PM
I must admit I think the renderer is really slow, but the results are still outstanding. People are going on about cost and how it will cost people more to have a new renderer. I don't think this should be true. Software has to progress to keep up to date. Why should we have to pay so much more for developers to just do their jobs properly? You don't expect to pay a huge amount of difference for a new game that's come out using new tech do you? It's a companies job to stay competitive, arguing that it's cheap won't cut it if it just isn't up to the job anymore. I'm not saying that Lightwave isn't I'm just putting a point across. As for other apps being a lot more advanced. I don't know about the latest versions of Max but I know that a few years back we were looking a using max to do some new tech in a game we were developing at the time. One of our programmers happened to mention what a mess the code of max was in. Basically through the years stuff gets re-written, added and left so you just end up with lots of bits bolted all over the place. So I don't think it's just Lightwave that suffers from this.

GeorgeDittmar
03-02-2004, 05:46 PM
i found lightwaves render to be slower than max's but then i learned max used a diff type of renderer than lightwave and that i prob had my settings to high. lol

Phil
03-02-2004, 06:10 PM
Originally posted by DarkLight
This is mainly due to the fact that mental ray had been added and they had to write convertors for any texture/shader features for anything that is not compatible with mental ray. Softimage gets around this problem by not having thier own renderer and developing the app to support mental ray compatable features.

I'm sure if Newtek were to add support for 3rd party renderers such as MR, then they would come across the same problems as both Max and Maya.

Yep. I'm very much aware of that. The question arises particularly for LW because plugins can already throw up problems due to limited communication. Whilst the 8.0 release is intended to ease some of these problems, it's a question as to whether external renderer support would be reliable, were it to be possible to add it to the existing system. *shrug*

I fear that much of the power available via mental ray, etc. would be wasted due to the limited surfacing facilities, etc. in any case. Unless NT could find a way to really deliver the power of external renderers (e.g. PRMan, not just mental ray), then they would probably be on a hiding to nothing. In any case, the argument is purely academic at this point. :)

Ade
03-02-2004, 07:02 PM
So compared to Vray, Photon mapping is just another word for Radiosity?

I was sure radiosity and photon mapping were two different things as explained by a friend of mine whos been getting into Vray in max?

Anyone clarify? Lets here it from some LW programmers..!

Titus
03-02-2004, 07:51 PM
Who better to explain this than mental images' Thomas Driemeyer:

http://www.highend3d.com/render/archive/sp.3d?mail_id=898

hrgiger
03-02-2004, 08:00 PM
Photon mapping is more then just standard global illumination. Photon mapping is real world lighting artifacts such as color bleeding, caustics, prisms, etc.... We have ways to fake it in Lightwave but they're not built into the radisosity solution. Lightwave's renderer does need a kick in the butt. Hopefully after 8 is released, Newtek will focus more attention on the render features and speed.

riki
03-03-2004, 01:33 AM
Isn't 8.5 is meant to be the Modeler Build?

NanoGator
03-03-2004, 02:35 AM
Man, I'm still exploring what 7.5 has for me... I mean faster renderer would be nice, but $500 buys me another render node.

Personally, my big time sink isn't rendering so much, it's the development of an animation. Antyhing they can do to speed that along will be more useful to me than a marginally faster renderer.

I can't judge LW's 'behindness' but I do think it's a very powerful tool to use. Speed isn't just how fast pixels can be thrown on the screen.

hrgiger
03-03-2004, 03:13 AM
I'm with Nanogator on this one. I'm sleeping or doing other things while Lightwave is rendering. Now with FPrime coming out, previewing is going to take no time at all. I'm just for improvements within the interface and working with Lightwave, espeically animating.

Nemoid
03-03-2004, 03:37 AM
Into an article about Vray, i read that photon mapping is a faster technology wich allows to calculate faster with radiosity, but in a not 100% accurate/brutal way, that is to say, avoiding to calculate smth but giving good results however in less time. maybe a comparison between a complete total radiosity calculation with all the bounces and so on, phisically correct and a photon mapping rendering would show some difference.
at the end, the article said that an accurate calculation always takes a certain amount of time, while photonmapping needs less time , with good results , but doesn't reach 100% the accurate calculation.

However, I'm sure that Nt will work on the rendering soon. 8 release has been focused into CA and workflow issues, with new modeler tools too. but i think we will see a focus on the rendering engine soon too. its the next area of the app to work up.

riki
03-03-2004, 03:45 AM
I wonder how PrimeTime sorry FTime sorry FPrime holds up on heavy renders.

I think we should be able to produce better work if we can get realtime feedback as we tweek.

Nemoid
03-03-2004, 05:18 AM
Now, this is surely right!
with a so good and fast preview of the final rendering, you can tweak better your work, without render tests and then let the Lw rendering do the final job with no prob.

I'd not be too much surprised if F-Prime could do much more than the things showed till now, especially if you can save frames sequences. since you can move geometry, and bones, and lights... and all this with AA...

maybe this will substitute the current Lw renderingi some degree ? :)

DarkLight
03-03-2004, 06:44 AM
Originally posted by Nemoid
maybe this will substitute the current Lw renderingi some degree ? :)

Even if FPrime does not handle animations, i bet a lot of people who only produce still images will prefer to use it as a render.

Jaffro
03-03-2004, 11:51 AM
I think FPrime's main aim on the market is for tweeking, not as a final render even though it can do that. I think FPrime demonstrates that LW has enough power still to create amazing tools. LW renderer, although old, is still powerful enough for most needs.

Didn't Dan Mass do some exporting to renderman for rendering? I think it is possible, although sometimes hard, to render with what you like. But LW renderer can handle most things, and yea sometimes you'll have to do a 'work-around' to get the same result, but at the end of the day money is the factor.


In reply to the origional question; Does Lightwave render show its age?, No i dont think it does at all. It should look a LOT older than it does. There is room for improvement, and she's slowing down compared to the other top renderer's out there, but I doubt most people want much more than a speed/accuracy increase as LW's current renderer is beautiful.

Thats just my pov, at the end of the day i dont think it matters that much. We as users should be requesting the features we would like and NT as developers should be listning - this is happening so there's no problem!

Nemoid
03-03-2004, 12:34 PM
Even if the article quoted isn't so positive towards Lw renderer i agree with you, Jaffro. it should look way older than this. this also means that it stays on the market for sure , between all the others. and you have unlimited render nodes to build your renderfarm too.

I'd like to see rendering speed, AA and motion blur improved, but i also think they WILL be improved.

Another thing I'd like to see is some good support for 3rd party renderers, especially Mental Ray for more power when you want it.

But It's true. Lw renderer gets your job done in most cases, and with a good output for sure. A photographer i know compared rendering between works made with Maya's rendering from a good expert, and Lw renderings too. he said that Lw rendering is very good concerning the output and light quality. this is not speed, but indeed is a fundamental thing.

as for FPrime, a good preview tool for sure way better than what other apps have at their disposal right now. this will make 3D work really better and fun with good previews in NO time for sure.

all the rest, if there is smth more, is a great add on to a really magic tool .

scott_krehbiel
03-03-2004, 12:49 PM
Responding to some comments posted a bit further up the thread, I was in a group of people learning from a former ILM artist. He was at first interested in Lightwave's radiosity rendering, but then found C4D and loved it. He's now pushing Cinema 4D pretty hard, but =only= for its radiosity rendering. He loves the fact that you just turn it on, and it renders nice and pretty. I must admit that my experiements with radiosity in Lightwave weren't as pretty or easy as how he set up his work in C4D.

When I asked him about modeling, he said that he only models in FormZ.

It's kinda funny how many shops will model in one package, animate in another, render in a third....


Now here's a question - what would you all think if Newtek came out with a dedicated high-end modeler? An extra option that you could purchase, for maybe $799 more. Then you could have your full-on NURBS modeling, editable parametric object history, etc. With the editable history, you'd be able to do a bi-rail sweep, then decide to change the rails, and your mesh would update in real-time.

The cool thing about this solution is that it would give Newtek some incentive to develop higher end tools, and then these tools could slowly filter down into the consumer Lightwave version as newer stuff is added. I think that right now, Newtek may need a little financial incentive to get the real high-end development going. I hope that they're still above that scary red line in the accounting books.

One more point is dealing with features that lots of other companies are very quick to add - new mathematical models for doing all the cool stuff. If you go into Max and turn on radiosity, you'll find -what 5?- different algorythms to choose from. Go to texture an object, and you've got a bunch of shader models.

I'm not talking about algorythmic textures, like wood or burlap, I'm talking about Phong vs. Blinn vs. Oren-Nayer-Blinn vs. Lambert vs. Metallic. These are the math models that determine how an object is lit, not just what color it is. There's a big difference. I wish that you could choose a different shading model in the texture editor, and then have all your controls right there - not needing to go to the plug-in controls, like you do for the plug-in shaders.

I don't mean to accuse Newtek of resting on their laurels, but other companies are saying "Radiosity! Great idea! Here's 6 ways of doing it!"

But then again, I'm greatly looking forward to seeing the new editors for textures and hopefully even shaders!

Don't bother with me - I just ramble here
Scott

Nemoid
03-03-2004, 01:30 PM
This would be a good product for sure. there are alot of good solo modeler apps, but one more can be good as well, especially if introduces NURBS, ngons, edge weighting and releted edge tools , better splines and a lot of great enhancements. since modeler is an app and Layout another, NT could do this for sure, i think.

as for Cinema, is a good product wich is enhancing more and more. at ILM they have a XSI pipeline,but as we all know, great studios use different apps, most of wich propietary apps to work.
they used Lw as well with good results.

DarkLight
03-03-2004, 03:06 PM
There is a SDK for Mental Ray so it should be possible to interface LW to Mental Ray to some degree.

I'd give it a try myself if i had a copy of if, but it's a little out of my price range to get just for that.

jin choung
03-03-2004, 04:50 PM
scott,

the overwhelming answer would be, "WHAT THE HELL FOR?"....

as you said, there ARE apps like formZ... along with a boatload of other apps like AMAPI, REALSOFT, RHINO, and a crapload of things from alias.

since this is the arena staked out by others, in which newtek has NO MARKET PRESENCE as well as NO TECHNICAL RESOURCES (!!!) (aside from a non robust implementation of SDS, we are still almost strictly a POLY ONLY APP!!! these other apps RELISH the very idea of researching new and interesting primitive types!). that kinda expertise doesn't grow on trees.

and also, even if they ACQUIRE the talent needed to slap something like this together, waddya think, would it be easier for formZ to continue putting out even more refinement or would it be easier for the newtek modeler to hit that same level of refinement from a standing start?

BTW, combined with the fact that there is no standardized implementation of SDS or (perhaps more importantly) TEXTURED SDS, and nothing standard about edge, vert, face weighting, this is one of the reasons why i think modo is doomed.

not to mention the whole money thing again.... difficult to get best in class anything if you charge little. and if you don't charge little, what exactly would newtek be capitalizing on?

not rocket science. not advanced economic theory. it really is this prosaic... it really is this boring.

------------------------------------------------------------------------------------

speaking of advanced economic theory though - i was watching charlie rose a few weeks back and there was this guy from harvard business school talking about a new book of his.

but he said that one way to BREAK IN as a newby into a new field is by making something that is "NOT AS GOOD BUT CHEAP".

as an example, he used japanese automakers in the early eighties. they got a foothold by putting out affordable crap.

by doing so, they are not competing for customers who are IN THE MARKET for the product. not at all.

instead, they are grabbing the attention of people who never even considered that the product was a possibility at all for them.

sure, people may like the quality of a swiss watch technology but everybody's interested in CHEAP.

and in exploiting that market, you can get a foothold and ramp up like the japanese automakers.

this BTW is one of two reasons why my app of choice is LW. :)

problem with 3d apps though is that you can't assume a HUUUUUUGE market. this is definitely not on par with consumer electronics and 3d content creation is not an everyday thing for most folks... nor does it look like the immediate future holds conditions for this to become any different.
------------------------------------------------------------------------------------

oh, and to the guy who said that i should make reccommendations about the interface... er... where have ya been? i do that constantly?

one of the biggest 'features' of the new 6.x generation rewrite was that new tools can simply and seamlessly be added as .p plugins.

guess what, that ended up being one of our greatest double edged 'features' cuz the developers ended up giving us a vast array of really useful tools that ended up in CONSTRUCTION ADDITIONAL!

i ***** about that every other day. and this is one of my few gripes that don't require siggraph level knowledge to tackle. the principle is very easy, very clear:

IF YOUR INTERFACE CANNOT COMFORTABLY AND ELEGANTLY ACCOMMODATE ALL YOUR FEATURES, SOMETHING IS WRONG!!!

further (pet peeve)

you should not have organizational tabs that don't have a clear DESCRIPTIVE, USEFUL MEANING!!!

what the hell does CONSTRUCT or DETAIL MEAN (?!?!?!?!) within the context of tabs like MULTIPLY or MODIFY (two really excellently apt tabs).

it just requires someone who likes things to make sense.

i wouldn't even be averse to MODIFY1, MODIFY2, MULTIPLY1, MULTIPLY2, etc....

and before some smart guy again decides it's wise to inform me that i can setup the menus to my heart's content... i am a proponent that the default menu should be an example of multipurpose elegance (as opposed to specialization which is where customization SHOULD come into play... as opposed to having to mess about just to make the thing make sense).

as for the 'they listen'.... yah they do. but for a while now, they've either been unable (money, time, other resources) or unwilling (insanity) to comply with my nipple hardening wisdom.

jin

Phil
03-03-2004, 05:56 PM
Originally posted by DarkLight
There is a SDK for Mental Ray so it should be possible to interface LW to Mental Ray to some degree.

I'd give it a try myself if i had a copy of if, but it's a little out of my price range to get just for that.

You'd still be stuffed in trying to get HVs, etc. to render AFAIK. It seems that these are inaccessible to plugins - shaders cannot be applied to them, for example (not the ones I have tried before now in any case). Similarly with the ground fog plugins, etc. I just don't see that kind of openness for devs and so we'd end up with a very limited pipeline.

I am not going to hold my breath though and I'm really on the fence about upgrading LW at the moment. If I absolutely knew what was on the cards for renderer and surfacing system updates during the 8.x lifetime (and very rough estimates for dates), I would perhaps be happier.

The absence of any fix information (or update for 7.x) for reported and confirmed issues with LW's current renderer has me very concerned and has dampened my enthusiasm for 8.0 considerably. Similarly with other bug fixes - there simply is no information and it cannot be argued that releasing this information would be problematic. If release is, as mentioned elsewhere, March 22, then I fail to understand the logic as NT must know what got fixed and what's still broken at this point in time.

Panikos
03-04-2004, 04:14 AM
Ade, LW uses Photonmapping. Arnie Cachelin wrote that many many times but you can conclude it yourself as soon as you understand what is photonmapping.

Photon mapping is just an algorithm of GI calculation, let say the most dominating out of all.
You can read more in here
http://graphics.ucsd.edu/~henrik/

Henrik invented photonmapping.

LW's Radiosity is the most accurate than other applications, maybe the slowest. Additionally the caustics calculation is splitted from the radiosity, so you can have caustics and/or/without radiosity. (This is not good imho)

A photonmap is an image that resembles your rendered image, where photon values are stored. Currently there is no way to view the photonmap in LW, but imagine that it looks like the difference of an image without GI and an image with GI.

Nemoid
03-04-2004, 05:20 AM
Jin,

I've never used Form Z, there are alot of solo modelers apps, but I don't agree another modeler wouldn't find a place in the market for sure. Even as it is right now.
Also, it would be very interesting to see a modeler with the ease of use of Lw modeler, but with alot of good features wich could interest many users, with different kinda workflow. for example, introducing NURBS could sound not so a great idea at a first glance, cause in the industry, subpatches are what they want the most, for the many advantages they bring, but it would be efficient and interesting for many users as well. this is valid for many other things.

The implementation od SDS in Lw current modeler doesn't support ngons, edges and the other things you said, but as a modelling workflow offers the possibility to do well and good in less time however. this because of the philosophy of the modeler in general. this is also the reason why many studios adopt and adopted Lw modeler for movie production. In lw modeler is currently the strongest part of the app.

For these reasons a lw solo modeler would be.. interesting.

Now, we dunno if Nt could afford such a task right now, but its a good idea IMO.

About the 3d apps market, its clear that everyone likes cheap.
Even Bill Gates! he likes cheap ! for sure! :D
The Lw philosophy has ever been that of being both cheap and good at a certain level.
There are also very expensive apps , though, just because in the movie industry, for example, production speed is what they need and want. always. this means alot of money saving. so, you can see, its a strange market, not so huge and difficult to face for sure.
As an user, for example i think cheap crap is not always better than a little expensive good app, just because the money i could spent now, can return afterward in time save, flexibility and more.

The ideal position is sometimes in the middle, being as cheap as you can and as good as you can.


As for the interface ,plugs and other things. here i agree. The current situation is also caused by the fact that Lw has never been developed into a linear, stratified and clever way. the main reason of this is at the start, just because it was born like 2 applications, wich were joined to work toghether more and more afterwards. Now, this is a strange born and evolution indeed.

At 6.0 time, Lw was "rewritten", but the structure of the app, was not rethought entirely to allow the user to work into a modern fashion.Into one environment, for example. Some lack of boldness maybe. Some lack of money too? I dunno this.

But i know that this is the cause if the current situation. This is more , more important than a simple UI rearrangement and redesign, because at the end it reflects on your work into a massive way.
There's only one good and valid solution to face this task, and its a complete, clever rewrite that Nt could start to do into a parallel pipeline, especially near the end of Lw8.x cycle, as i always said, bringing some parts of the code into a new well thought environment, recoding other parts, making all of this working smoothly.

Ade
03-22-2004, 08:33 AM
What makes Lightwaves renderer more accurate than others on the market?
I rememebr once someone stated Lightwaves renderer is 192 bits...

What was it exactly and how does that help?
I read cinema 4d is 128 bit?

Whose is more accurate?

LW, 3ds (lightscape renderer), c4d (based on cebas), Maya or xsi?

scott_krehbiel
03-22-2004, 09:15 AM
Ade, I've had some experience on this one - not at the rendering level, but at the simple object movemement level.

When I was using Max for games, (versions 3 and then 4) I had strange problems with what I believe were bit-depth errors: Take an object that's aligned at 0,0,0. Rotate it +5 degrees on Y. Then rotate it -5 degrees on Y. Now where would you expect it to be pointing??? 0! Of course. But where would it actually read? -0.000097 Every stinkin' time. I'm pretty sure that it was Y rotation that would do this, but I may be remembering incorrectly. I try to suppress the memories of my Max days.

I also noticed that when two objects were close, one occluding the other, there would be render errors. Even though you knew that the painting on the wall was closer to the camera than the wall was, the painting would pop in front of and behind the wall every other frame. It wasn't a matter of object centroids, but I think the package just didn't have enough bit-brains to decide which polygon was closer and stick with that decision.

I don't know if Max has fixed this in more recent versions (and don't everybody start flaming me for using Max - it was my employer's decision and not mine). The most recent version that I touched was v4.1

Anyway, this is the advantage of having more bit depth in your app. The same type of principle would hold true for the bit-depth of the renderer. In a very small nutshell, more values of brightness/darkness means prettier exposures. For instance, if you have a scene that you lit a certain way, but want to put a serious gamma curve on it to get it looking right, having lotsa bit depth in your render means you won't get posterization after you do your gamma corrections. Also, the fact that your light rays have all that bit depth determining the directional vector means that your reflections and refractions will be much more accurate. Imagine having "guesstimaded" reflections.

Now imagine being able to render a city scene and not have the frigging brick walls do some stupid transparency moire zebra stripe "oh I'm invisible! No I'm not. Yes I am!" trick. (Sorry, some of the Max memories have been uncorked. I knew I shouldn't have gone there. See, now my eye's started twitching!)

Now I'm sure that others will chime in and say "but the algorythm is as important as bit-depth" and I'll say ahead of time that they're probably right, and their reasons are probably way over my head.

I do know that C4D has some really pretty radiosity, and I'm seeing more and more shops advertising positions using "Maya and Cinema4D." Bloody amazing - animating in Maya and rendering in C4D. Who would have thought?

Scott

Ade
03-22-2004, 09:19 AM
Whats the lightwave advantage then in its renderer?

scott_krehbiel
03-22-2004, 10:30 AM
Originally posted by Ade
Whats the lightwave advantage then in its renderer?

Well, I think that based upon all the technical details communicated in my last message, the answer becomes very clear: I have no clue. :p

No, all joking aside, there are tons of trade-offs. In my opinion, I'm happy that objects don't pop back and forth, fighting as to which is really in front of the other. I also like the fact that even very dark objects which aren't lit very well, still have lots of lighting data. You could basically "over-expose" your image like a photographic print, and bring details out of the shadows.

But as I said, it's a trade-off. Maybe the lighting algorythms are just better in other packages. Maybe other packages support more illumination models than phong - stuff like Blinn, Oren-Nayar Blinn, Lambert, etc.

I hope other people can give good examples of how Lightwave's pipeline helped them at render time.

Oh, here's one example though: I was working on a friend's independant film, and I had to put 650 sundials all over these hills for a crane shot. I'd made these sundials about as optimized as I thought best, and went to render - in Max. Later, I decided to take the project into Lightwave and render there. The render times in Lightwave were 1/7th the render times in Max. So at least in the speed category, Lightwave is (or was) about seven times faster than Max. I don't know about the current version though.

SplineGod
03-22-2004, 08:16 PM
Originally posted by Ade
Since when? and How?
Heres a quote from Arne Cachelin that was on the old Newtek Forums. Notice the date. Lightwave has apparently had photon mapping for awhile. :)

By Arnie (Arnie) (65.184.20.249) on Wednesday, October 24, 2001 - 11:11 am:


Lightwave also uses photon maps, which are basically structures to store and use global illumination calculations efficiently. As the most mature commercial photon mapping software, LW has refined the features to allow interpolation to replace more exhaustive sampling, and has also added a mode for ignoring diffuse interreflection and only using the environment for illumination. These features make global illumination rendering more usable by making it faster. Adding more explicit bounces vastly increases the rendering time, while adding very little to final renders. The entire effect of all higher order bounces is controlled by the ambient intensity (which is different when radiosity is enabled). Using a global non-directional lighting amount to account for bounces not explicitly calculated is a solution recommended by Jensen, since higher order diffuse bounces create illumination which tends to be non directional, and weak. When radiosity is enabled, this is the ONLY use of the ambient lighting, so you should use it to set the amount and color of the dominant diffuse reflection.

KouunnoHito
03-23-2004, 12:28 AM
Originally posted by Original1
There have been a, b and c updates along the way, all free
and the .5 update was free.

unlike others who charge you for .1 updates

The Issue is that there has been a hiatus in real development in all 3D apps for the last couple of years No one has really been breaking new ground. pixologic has been breaking ground

scott_krehbiel
03-23-2004, 12:33 PM
About the photon mapping and Lightwave's lighting algorythms, I'll say this: I've seen global illumination from C4D, and it's really pretty and it's really fast! I think it's much faster than Lightwave.

Is anyone here familiar enough with C4D to comment on the types of global illumination available? I know that other packages will often give you a choice of 4 or 5 algorythms that you can choose from.

P.S. I'm not slamming Lightwave... I know that they pioneered in this area.

pixelinfected
03-23-2004, 04:35 PM
lw is a good engine, but it's old, and show perfectly it's age.
it lack in :

-managing high resolution (slow and memory waster)
-high memory request (over all resource)
-slow high obj number manager
-slow high poly mover
-too sloppy in passage from layout to screamernet (originale engine to net render engine
-slow GI implementation (check fprime)
-no BSP tree implementation on big poly scene
-no antialias by low and high filter on render engine (like renderman and mental ray)
-no ibryd motion blur (only time slice mb, slow and not enought good, and unique time they try to do a 2d MB, with postpro MB they did a very big water hole...)
-no texture optimization in render
-no good multipass exporting

some one tell me that newtek not have enought money and developer to develop a new engine?

ok, i must tell you that in italy there was a boy which develop at 17 years old, a new render engine, renderman compliant, that is so bad, like developer, that it adapt for maya, max and more, and now work for pixar.

no excuse, the only reason is lack of organization, 6 was only a rebuild, not a total real rewritten of software (if they did that like they told, they add undo multilevel, which every professional application have, i teach to add a multilevel undo on developing exercise of my student during third week of course)

but without a total rewritten software, add a multilevel undo is a desperade mission.

now newtek restart witha new team and i hope they organize better all.

good luck to new dev team, which, i hope, show the muscle with lightwave [8]

iacopo

pixeltek
03-24-2004, 07:27 AM
I realize that this is premature, but LW9 is based on new core architecture. Changes will probably affect every aspect of LW across the board, including the renderer.

mattclary
03-24-2004, 09:06 AM
Originally posted by pixeltek
...but LW9 is based on new core architecture...

And what oracle did you go to to get this revelation? :rolleyes:

I know everyone is HOPING for a rewrite, but you state it as fact, which I think is VERY premature.

Nemoid
03-24-2004, 10:55 AM
Maybe he's a prophet! (or someone from the Nt dev team) :D

mattclary
03-24-2004, 12:57 PM
Then he should be bound by an NDA.

mkiii
03-24-2004, 04:54 PM
Or just bound.

pixeltek
03-24-2004, 06:35 PM
Hey guys. This is a nice and interesting thread, so save the sarcasm. I go to Siggraph year after year, and this way I get to meet and chat with a lot of folks.

First of all, I thought that this was by now common knowledge. Secondly, this information is probably easy enough to verify from Newtek employees, some of whom stated matter of factly that this current core architecture has pretty much run its course and that they had bumped into the limits of what they can do with it. At that point it was mentioned that a core rewrite is already underway and that much energy and resources are expended in that area. Just as most of you already suspected, the core architecture is getting long in tooth, and to no one's surprise, LW8 (and potential fractional upgrades) should be the end of the line.

Personally I thought that was really neat info and gave me something to be very excited about for the future of this SW. I remember that it came up when I returned to the Newtek booth from a Maya demo of their new fluid dynamics sim. At that time it was said that the current LW architecture would not lend itself easily to incorporate this, but that the new one had no problems handling this and more.

That's about it. Nothing psychic or prophetical about that info. Just enjoy it and anticipate: great things are coming our way!!

pixeltek
http://www.cosmic-pearl.com/

pixeltek
03-24-2004, 06:52 PM
Let me add one more thing on what I just said. No one is probably more secretive about development than automotice manufacturers, yet virtually everyone knew for several years that a new Corvette and Mustang was under development. It only heightened everyone's anticipation, but until the first "spy" photos (usually they are tacitly permitted to happen) came out, rumors were abundant, yet no one had a clue of what the final output was going to be, until the companies finally let the cat out of the bag, so to speak.

By letting people know that a core rewrite was underway, is a similar situation. On one hand lots of people can excitedly walk around and anticipate the exciting future release, whenever it will take place, and on the other hand, the company is viewed that it is far from stagnating, but expending much time and resources to release a totally fresh, leading-edge product some time in the future. We could not even guess when that release will be or what will be included, but remember when, I think it was Softimage, did their core rewrite? They did not have a new release for several years until they finally fielded XSI. I think it was the longest release hiatus for any 3D SW package ever. Just to give you some caveats regarding this issue. Conversely, C4D even though being relatively new, already did a core rewrite for their latest release. Different strokes....etc

pixelinfected
03-25-2004, 03:43 AM
it's the hope of all lw user, but until i not see it i cant' believe.
last time they tell that lw was rewritten (6), it's only a restyle, but the core was same, that is the reason for actual problems.
i hope the new team start from zero, a new architecture.

jin choung
03-25-2004, 04:29 AM
no, the sarcasm is RICHLY DESERVED....

WHO told you? HMMMMmmmmmmmm?

sorry to burst your bubble but there's a lot of folks hanging around on convention floors who are full of crap and merely wannabe hangers on....

you were not clear in your post. are you saying that a NEWTEK EMPLOYEE directly told you that they are doing a total tear down?

if so, i want this to be the very next words in your next post:

NEWTEK TOLD ME THAT THEY ARE DOING A TOTAL TEAR DOWN, GROUND UP, FROM SCRATCH REWRITE.

and if it's sooooooo easy to get a confirmation from newtek staff... GO AHEAD! give that a try!!!

so if you speak truth, please, cite your source.

jin

Matt
03-25-2004, 04:47 AM
Answering the thread question . . .

In terms of quality, difficult to say, in the right hands I've seen some amazing work from LightWave that matches work from other renderers with all the bells and whistles.

In terms of speed and features I think it is (definately speed).

Just my 2 cents!

Exper
03-25-2004, 04:52 AM
Originally posted by Matt
In terms of speed and features I think it is (definately speed).Maybe we'll have a little aperitif with the announced LW[8] "Faster Rendering"! ;)

Nemoid
03-25-2004, 04:59 AM
Pixeltek.

i really hope you're saying the truth, but...you know,words are only words....Verba volant, scripta manent, romans used to say.
you have no proof of what you're saying.

I think no one of NT team assured you with documents that Lw wil be rewritten for release 9.0.
Nt have a new team so this is possible, but nobody from Nt posted infos, or statements to validate what you're saying.

That's why nobody here thinks or is authorized to think Lw will have a core rewrite for release 9.0 for sure.


to return on topic: i think Lw rendering will be addressed after 8.0 release. its the next area to work up.

Matt
03-25-2004, 05:09 AM
Originally posted by Exper
Maybe we'll have a little aperitif with the announced LW[8] "Faster Rendering"! ;)

I'm really hoping so, I really am!

I also hope the OGL speed is improved too, right now I'm working a scene that is dreadfully slow, odd thing is, even if I set the objects to bounding box it's still slow when moving stuff around, not as bad, but it should whip along.

I just find it annoying when you have to move something very slightly, you get no feedback then suddenly whoosh! it's a few metres away!!!

Exper
03-25-2004, 05:30 AM
Originally posted by Matt
I also hope the OGL speed is improved too...We'll see...
take a look here:
http://www.newtek.com/products/lightwave/product/8/features.php#OpenGLEnhancements

Matt
03-25-2004, 05:33 AM
Thanks Exper, was aware they said they had addressed it, I just hope it's blazing fast now!

Exper
03-25-2004, 05:35 AM
Every "not moaning" user is hoping too! :D

pixeltek
03-25-2004, 06:55 AM
Hi jin.

I have been following your interesting posts for a long time (in and out of exile), but just in case I am stepping into someone's ricebowl here, I will not cite these sources (yes sources - I heard this from more than one). No, these were not wannabees either, but people we all know and respect. You can do with this whatever you want. I will bow out of this discussion with a friendly smile and wish you and the rest of the respondents all the best. I'm happy with this information and looking forward to exciting things to come. Let's leave it at that.

Peace,

pixeltek
http://www.cosmic-pearl.com/

Nemoid
03-25-2004, 08:39 AM
No prob. if you are sure of what you say,you probably are right.

BTW : nice SW models there!! :)

pixelinfected
03-25-2004, 10:11 AM
Originally posted by jin choung
no, the sarcasm is RICHLY DESERVED....

WHO told you? HMMMMmmmmmmmm?

sorry to burst your bubble but there's a lot of folks hanging around on convention floors who are full of crap and merely wannabe hangers on....

you were not clear in your post. are you saying that a NEWTEK EMPLOYEE directly told you that they are doing a total tear down?

if so, i want this to be the very next words in your next post:

NEWTEK TOLD ME THAT THEY ARE DOING A TOTAL TEAR DOWN, GROUND UP, FROM SCRATCH REWRITE.

and if it's sooooooo easy to get a confirmation from newtek staff... GO AHEAD! give that a try!!!

so if you speak truth, please, cite your source.

jin

lw is actually in blind way. the old core is too close and hard to change.
lw is in the same situation of softimage some years ago, before to start sumatra, now XSI project.

with old and crap core, it survive thanks to a great society, phoenix (italian society) which develop a bunch of plugin which help and cover every needings of softimage user.
Every softimage user remember the phoenix tools cds and help fournished. and actually part of XSI code is Phoenix dev.

For Lw there is Worley, and now NewTek and Worley start to collaborate to develop a next lw sdk (that mean they work to a new core and next lw generation)

be cause from my old experience like developer (c, c++, visual basic, java, and more), i know that is more simple to restart a new code, and use old part like function rewritten as dll and plugin, then modify an old code that is not object oriented, and many other thing which all modern software must do to continue to live).

all i told is only from my experience with lw (i use it from over 11 years) and many other experiences.

i not have direct contact from new dev team which give me segret info... i would but i haven't ;-)

pixeltek
03-25-2004, 10:12 AM
Thanks, Nemoid. Its' one step at a time. For now we'll enjoy LW8 for all it's worth, and then, once the excitement abates, which should be a good long while - and the price is certainly right - then we'll turn our collective minds and eyes toward the future and the rumbles and rumors of the next full version LW. At that time you will remember what you heard here and what that entails. IMO, this will be huge. Anticipation (wasn't there a song?) is half the joy of getting there.

pixeltek
http://www.cosmic-pearl.com/

jin choung
03-25-2004, 02:45 PM
look pixelinfected,

i agree that we NEED a ground up rewrite. i can see the ADVANTAGES of that as well. i can see the RATIONALITY of it.

BUT THAT DOES NOT MEAN IT'S GOING TO HAPPEN!!!

money, money, money, money, money, money, money, money....

so for heaven's sake, don't go around saying that it WILL happen! unless someone with some authority told you that.

say that you'd LIKE for that to happen - that you'd expect it - that 'it only makes sense'....

but not that it will without a doubt.

jin

KouunnoHito
03-25-2004, 10:54 PM
Jin,
I'm going to have to disagree with your money angle R & D is the heart of a company with the goal to maintain some level of competitiveness in it's market. So the "it takes money angle" isn't working for me in ANY arguement you've used it in. If newtek doesn't maintain some sort of edge they can fall out of the spotlight. And they're playing catch up as it is now, so they need to get it in gear.

jin choung
03-25-2004, 11:20 PM
hey,

disagree all ya want. but facts are facts....

alias, softimage, sidefx and the other large 3d app companies aren't where they are with THEIR r&d simply out of sheer luck....

their formidable r&d COSTS MONEY. really! i [email protected]# you not... r&d costs lots and lots of money!

they charge MUCH MORE (and much more STRICTLY AND FREQUENTLY) and they have a MUCH LARGER user base by way of major fx houses.

they HAVE THE RESOURCES to be where they are.

lw charges LESS to a relatively fewer number....

you're saying they have to remain competitive. THAT'S A GREAT SENTIMENT!

so where are they gonna get the money to pay for the sentiment?

JUST BECAUSE YOU HAVE TO DO SOMETHING DOES NOT MEAN THAT YOU AUTOMAGICALLY GET THE MONEY TO DO IT!

unless you live in a markedly different universe than i do.

soooo... tell me....

THEY MUST DO AGGRESSIVE R&D TO REMAIN COMPETITIVE YOU SAY....

so... either they HAVE the money... or they have a way of GETTING the money (don't kid yourself... it's a vast sum to do a ground zero rewrite)....

enlighten me.

archiea
03-25-2004, 11:29 PM
As i've said before, NT should charge for X.5 updates. If you can't afford it, don;t get it. but your budget shouldn't impede its development.... professionals will pay for that pace of progress...

jin choung
03-25-2004, 11:42 PM
well,

just speaking for myself, if they start charging more, i'll stop paying.

lw meets a budget niche for me. i have about four major reasons why lw is my app of choice and one of the most significant is COST.

if they start charging for .5s i'm out... and will probably just go to maya which i use at work and i like quite a lot.

THAT in itself is not a deciding factor for newtek.

if they feel that they have to raise prices, that is their prerogative.

but what IS a deciding factor is whether they think that they will ACTUALLY EARN MORE MONEY BY CHARGING MORE.

they have to do the math on how much more they would make from paying customers who stick vs. the money they would LOSE by customers who won't... like me.

it's a tough call and just charging more frequently like you suggest may make the situation worse rather than better.

jin

KouunnoHito
03-26-2004, 12:12 AM
R&D cost money true enough but your not getting the point, without R&D their'd be no product or a severly outdated one.

I don't see the swelling cost from a gound up rewrite that you do, so please elaborate on what I haven't considered.

jin choung
03-26-2004, 12:33 AM
????????

!!!!!!!!!!!

????????

WHAT?!?!?!?!?!

a ground up rewrite would be an ENORMOUS, GINORMOUS EXPENSE!!!

are you kidding?!

how LONG do you think that that would take? basically, it's the equivalent of writing a COMPLETELY NEW 3D APP FROM SCRATCH!!!

especially since most of the current programmers have VERY LITTLE EXPERIENCE AT ALL with the current code base! most of them were brought on specifically at the beginning of the eight development cycle.

so... conservatively, that would be 4 or 5 years. 4 or 5 years without the ability to update the current product and therefore, not have anything new to sell during that time.

oh, you say that they CAN upgrade the current product simultaneously? well, that requires more programmers. or we can just work the current ones 24/7 until one of them dies....

this is also assuming that the current engineers are even CAPABLE of such a task. if not, hiring people of such caliber doesn't come AT ALL CHEAPLY!

i can go on... ya want me to?

jin

mlinde
03-26-2004, 10:23 AM
Originally posted by jin choung
so... conservatively, that would be 4 or 5 years. 4 or 5 years without the ability to update the current product and therefore, not have anything new to sell during that time. Hmmm, sounds like something called XSI to me :eek:

KouunnoHito
03-26-2004, 11:42 AM
Originally posted by jin choung
????????

!!!!!!!!!!!

????????

WHAT?!?!?!?!?!

a ground up rewrite would be an ENORMOUS, GINORMOUS EXPENSE!!!

are you kidding?!

No and this isn't what I WANTED to hear from you:rolleyes: I have a good idea of expenses that would be ocurred but the queston is't how much total but how much more would be spent over regular R & D costs. Speaking of which what do you believe the designated goal is for R&D ?


Originally posted by jin choung
how LONG do you think that that would take? basically, it's the equivalent of writing a COMPLETELY NEW 3D APP FROM SCRATCH!!!

especially since most of the current programmers have VERY LITTLE EXPERIENCE AT ALL with the current code base! most of them were brought on specifically at the beginning of the eight development cycle.

so... conservatively, that would be 4 or 5 years. 4 or 5 years without the ability to update the current product and therefore, not have anything new to sell during that time.
Have you ever programmed before but the benefits far outweigh the negatives first ; They lost some of the original programmers who were most familiar with some of the code, 2nd the program is showing it's age and it's much more expensive to maintain Structured programming(as it's been said it is) over a modular OO code 3rd the new programmers their would have a great understanding of the application and how it works. 4th Future upgrades would be a lot easier and even upgrades to core functionality would be easier. Which should mean more tightly integrated code. As far as 4-5 years they would have to have hired some true idoits for it to take that long unless what they had planned was truly revolutionary. They aren't started completely over and this is something you forget or don't understand.


Originally posted by jin choung
oh, you say that they CAN upgrade the current product simultaneously? well, that requires more programmers. or we can just work the current ones 24/7 until one of them dies....

this is also assuming that the current engineers are even CAPABLE of such a task. if not, hiring people of such caliber doesn't come AT ALL CHEAPLY!

i can go on... ya want me to?

jin No actually I don't because your just screaming and not presenting any worthwhile "facts" that would be oblivious to me.

pixelinfected
03-26-2004, 04:49 PM
all that words confirm my ideas

newtek must release more frequently lw.

every years one pay update, and many free debugging version during the year.

the cost of year upgrade could be 300$ (usually we pay 495$ every 18 month, usually), then mean 300$ at year is not a so expansive cost for the update).

in this way newtek could have an year constant money to support and expand its dev team. if they continue to work in the actual way, they can use correctly their budget. that mean no good work, fast release to get money, and less happy user

pixeltek
03-26-2004, 04:57 PM
"look pixelinfected,

i agree that we NEED a ground up rewrite. i can see the ADVANTAGES of that as well. i can see the RATIONALITY of it.

BUT THAT DOES NOT MEAN IT'S GOING TO HAPPEN!!!

money, money, money, money, money, money, money, money....

so for heaven's sake, don't go around saying that it WILL happen! unless someone with some authority told you that.

say that you'd LIKE for that to happen - that you'd expect it - that 'it only makes sense'....

but not that it will without a doubt. "
-------------------------------------------------------
jin aka. obnoxious know-it-all,

I realize now that you really don't care what I think, nor what anyone else thinks, as long as you get to screech your opinion out in front of everyone. An of course yours is the only one that counts and is worth considering. As of right now, you are on "ignore". The price for reading your crap is getting too high. I guess I thought there was something more behind that name, but maybe you are not the jin I remember, but a cheap imposter. Identity theft, it happens all the time.

Be that as it may, just for YOU and YOU ALONE, I recant everything, I will call the well placed (and well-known, BTW) NT people with whom I had these talks liars, and promise that it was all a figment of my imagination, if it just shuts you up. Ok?

Happy now?

Good.

Now go and live with your miserable, bitter, bloviating and egotistical self. You've had your wish.

End of discussion. There will be no further reply, nor even interest in anything else you might have to say.

pixeltek

pixelinfected
03-26-2004, 05:13 PM
just a few second Pixeltek,
probably for my bad english i seem too much impositive, and not express my correctly idea.

i'm afraid if i not express correctly, that
(from your answer i understand that)

i work in the 3d industry from over 14 yeras, and i see many 3d software died, i love lw, and like a lot what i can do with it, i express my opinion (probably with wrong word) for a possible way to have a future.

i talk about a commercial strategy be cause like many people talk in the forum, first a society see that, than they could prepare the dev planning, not be cause i'm a commercial man (i'm the opposite of commercial).

the problem of cost is for me important, and in europe is more higher, actually the update of lw, which cost 495$ in usa, with change and tax grow to 495$ +240$
but i work with it, and if i must change software, i need time to learn, time when i can't work, then it cost more than a little adjust of price (which no one oblige me or you to buy it).

anyway,
pax guy,
have a nice day.

jin choung
03-26-2004, 05:19 PM
i'm devastated....

i'm sooooooo saaaaaaad.

boo - fing - hoo

as you started with the name calling, i could in fairness respond with a long string of really heartfelt obscenities directed at your ignorant little head, i will refrain.

as for 'obnoxious know-it-all'... hey, i'm not the one that stated a complete unknown with absolute certainty. if you claim you've seen a penguin fly, I WANT PROOF.

SHOW ME!

and if not, shut the f up already.

jin

jin choung
03-26-2004, 05:26 PM
oh,

and kh, you keep telling me the benefits of a ground up rewrite...

I COMPLETELY AGREE!!!

but just cuz i think it's a good and sound and reasonable idea doesn't mean it's going to happen!

also, i am arguing that the DIFFERENCE between current development and a total tear down, start from scratch rewrite would be enormous!

time is money and if it takes 4 or 5 years without the ability to sell something else, that is a HUGE sum of money.

money is a fact.

what 'facts' are you presenting? we're all bsing... but some of you guys are bsing as if you have a crystal ball or something.

jin

tokyo drifter
03-26-2004, 05:50 PM
Originally posted by pixeltek
jin aka. obnoxious know-it-all,

I realize now that you really don't care what I think, nor what anyone else thinks, as long as you get to screech your opinion out in front of everyone. An of course yours is the only one that counts and is worth considering. As of right now, you are on "ignore". The price for reading your crap is getting too high... :D :D :D

KouunnoHito
03-26-2004, 06:28 PM
Originally posted by jin choung
oh,

and kh, you keep telling me the benefits of a ground up rewrite...

I COMPLETELY AGREE!!!

but just cuz i think it's a good and sound and reasonable idea doesn't mean it's going to happen!
Let me clarify something, I do not believe lw9 will be a complete rewrite I have no facts to support such a idea and I never claimed it. What I did do is dismiss your notion of "money" as the reason why lw9 wouldn't be a complete update. And as a reaon for the non exising updates new popular features that popping up all over.


Originally posted by jin choung
also, i am arguing that the DIFFERENCE between current development and a total tear down, start from scratch rewrite would be enormous! WHAT 'm saying is simply R & D is a big part of any company that wants to remain competitive in their industry. Take Electric Image for instance they use to be a bigger player in VFX industry now look at their current position. If Newtek doesn't upgrade the tech and rely on their past reputation then they'll fade away, out of the forefront . Look at the recent head to heads comparisions, lw is showing it's age. Look at moves made by XSI, alias and especially discreet.


Originally posted by jin choung time is money and if it takes 4 or 5 years without the ability to sell something else, that is a HUGE sum of money.

money is a fact.
Again I doubt it would take a competent team 4-5 years to update LW, they have to maintain some sort of compatibility. So a lot of the code would be ripped out, some completely replaced and by a couple *.5 you'd have the completed and much needed fully modular updated code. And if the current code base isn't OO then they really don't have a choice. It can cost more to maintain and update structured code than it would to move to OO code. And again if it's simply just too expensive, why would companies spend money on R&D in the first place ? Simply R&D is the life blood of any competitive company, so money withheld can easily cause a company to fall behind. The age old saying is "It takes money to make money" without a new Complete rewrite I can't see LW lasting long as a viable competitor in this industry. Competitors are already treading in on Lightwaves space;
Maxon is neck and neck with LW.
Alias, XSI and Discreet have closed the Gap in performance and ability in rendering tech.

So while money is a factor it shouldn't be the only one considered.


Originally posted by jin choung what 'facts' are you presenting? we're all bsing... but some of you guys are bsing as if you have a crystal ball or something.
jin Ok replace facts with rational thought, but when you start a post of with dozens of !'s and ?'s it's hard to take you seriously. If you do want facts go look up the financial reports of a "few" companies.

jin choung
03-26-2004, 09:30 PM
you guys are killing me....

it might be a language thing but clearly, we're not communicating clearly.

!!!!!!!!

I AGREE THAT LW NEEDS TO BE UPDATED! I AGREE THAT WE *SHOULD* HAVE A COMPLETE REWRITE LIKE XSI TO SI, MAYA TO POWER ANIMATOR! I AGREE THAT WE RUN THE RISK OF FALLING BEHIND! I AGREE WITH THAT. COMPLETELY!

!!!!!!!!!

you think it's affordable. fine. that's your opinion. i DON'T think it's affordable. that's just MY OPINION. you asked me to elaborate so i did. and i don't really care if you take me seriously. you are free to think whatever you wish....

but i am challenging the statement that somebody made that IT WILL HAPPEN!

unless newtek told them that, HOW WOULD THEY KNOW?!?!?!

did god tell them?

the magic fairies?

jin

p.s. tokyo drifter, if you've got something to say, let's hear it....

private
03-26-2004, 10:15 PM
Originally posted by KouunnoHito
As far as 4-5 years they would have to have hired some true idoits for it to take that long unless what they had planned was truly revolutionary. They aren't started completely over and this is something you forget or don't understand.

Actually, it could be argued that the former development team had more than 7 full working versions of R&D plus over two years developing a module that hasn't been released yet. As an animation module is in the works, that can be counted in the time, however, a renderer hasn't been mentioned. Four to five years is a fairly accurate estimate.

I seriously doubt that the new programers are idiots (or idoits as you call them). This is a chance for Newtek to take a new direction. With other companies focusing on more of a modular development, I would hope that one day, Newtek can finally make the change, showcasing a unified application instead of the Modeler/Layout system there is now.

As far as the current renderer showing its age, I would say in the area of speed, yes it is. The quality in can render out is still excellent, but comes at a slower rate than renderer specific companies.

R&D takes money, and in the 3D application world, it takes a lot of money and time. Thinking that is doesn't is pretty naive.

KouunnoHito
03-26-2004, 10:43 PM
Originally posted by jin choung
you guys are killing me....

it might be a language thing but clearly, we're not communicating clearly.
Totally agree first let's get some thing's clarified;

What's your idea of a complete rewrite ? I have a feeling my take on such a thing is a bit different than yours. I think you feel they should bring in coders to strenghten the development team and junk everything and start over. My thinking is that they will dissect the code retain what they need update/replace those portions and integrate them into future releases(to assure compatibility and fix bugs) as they rewrite core portions of the program. But most importantly keeping everything modular as they do it.


Originally posted by jin choung
but i am challenging the statement that somebody made that IT WILL HAPPEN!

unless newtek told them that, HOW WOULD THEY KNOW?!?!?!
that's fine I simply think money isn't a factor in it not happening.


Originally posted by jin choung
did god tell them?

the magic fairies?
Maybe he use the ibm machine that let's you see into the future



Originally posted by private
Actually, it could be argued that the former development team had more than 7 full working versions of R&D plus over two years developing a module that hasn't been released yet. As an animation module is in the works, that can be counted in the time, however, a renderer hasn't been mentioned. Four to five years is a fairly accurate estimate.
The basic Program development cycle flows somewhat like this. Identify the needs/requirements, design the solution, get it approved, do the programming, debug, and document and update. I see the R&D portion possibly done by a different team altogether or some select few individuals off the team heading the charge into new functionality. However you can't say that develpment and maintenance of previous versions are R&D for a newer version that's backwards logic.



Originally posted by private I seriously doubt that the new programers are idiots (or idoits as you call them). This is a chance for Newtek to take a new direction. With other companies focusing on more of a modular development, I would hope that one day, Newtek can finally make the change, showcasing a unified application instead of the Modeler/Layout system there is now.
Ok first your putting words into my mouth, please try reading and comprehending a post before you reply one it. I never called them idoits, I said 1. "I doubt it would take competent programmers 4-5 years to update LW."

This is a compliment and confidence in them that they can get it done much quickier than that.

" As far as 4-5 years they would have to have hired some true idoits for it to take that long unless what they had planned was truly revolutionary."

Again another compliment as I believe A.) It would not take five years thus they couldn't possibly be idoits B.) What you and others believe would take 4-5 years just for a rewrite, I believe would allow them to release a revolutionary product.

I think I've made my point and I did not like the fact that you threw words in my mouth and used them to insult the developers. I feel you owe them a apology as one of them could have read what you said and been upset by it.



Originally posted by private As far as the current renderer showing its age, I would say in the area of speed, yes it is. The quality in can render out is still excellent, but comes at a slower rate than renderer specific companies. Not just speed it lacks certain features, I am by no means downing it. I am simply saying they aren't maintaining their edge as lighwave was/is reowned for it's renderer as the others offerrings are catching up and have some things LW doesn't.


Originally posted by private R&D takes money, and in the 3D application world, it takes a lot of money and time. Thinking that is doesn't is pretty naive. Are you saying that I'm naive ? If so I think you need to reread my post and try comprehending them this time. This will help greatly as it will cut down on your apparent misunderstandings.

Original1
03-27-2004, 05:24 AM
Originally posted by private
Actually, it could be argued that the former development team had more than 7 full working versions of R&D plus over two years developing a module that hasn't been released yet. As an animation module is in the works, that can be counted in the time, however, a renderer hasn't been mentioned. Four to five years is a fairly accurate estimate.

I seriously doubt that the new programers are idiots (or idoits as you call them). This is a chance for Newtek to take a new direction. With other companies focusing on more of a modular development, I would hope that one day, Newtek can finally make the change, showcasing a unified application instead of the Modeler/Layout system there is now.

As far as the current renderer showing its age, I would say in the area of speed, yes it is. The quality in can render out is still excellent, but comes at a slower rate than renderer specific companies.

R&D takes money, and in the 3D application world, it takes a lot of money and time. Thinking that is doesn't is pretty naive.

Having originally started this thread, the quality of the renderer is excellent, and is a tribute to the work done on it in Lightwave 6.0, but this thread was started in the hope of boosting a technical disscussion of waht works and what doesn't.

Also I was hoping to start a debete about future direction.

I realize that the programming team are busy right now, but I hope some of them might come on to the forums and give us some idea of the direction they might want to take the program in the next few years.

We all know we can fake radiousity with spinning lights, but are there things in the default settings which if tweaked speed the render up?

Is Lightwave renderer still powerful and current, but hidden to the newbies through lack of knowledge?

jin choung
03-27-2004, 07:41 PM
and finally, a word of sanity:


http://www.cgtalk.com/showthread.php?s=&postid=1233931#post1233931

A total rewrite from scratch at any point is not something under consideration, but a gradual change of the architecture with intermediate versions is much closer to what we are aiming for. Dr. Cross now oversees development of all NewTek software products and he believes very strongly that architecture must be addressed in every iteration of the development cycle in order to keep an application not only current to the latest in software technology but innovative, and from now on LightWave will have such an evolving architecture. Other folks may have other views about how development should be approached, but here at NewTek we are confident that this approach will yield the results that our users need in timely fashion.


__________________
Chuck Baker
Senior Director of Corporate Communications
NewTek, Inc.

tokyo drifter
03-29-2004, 02:02 AM
It's funny that after that moment of sanity, there wasn't any discussion about it in that thread. Seems like an important statement to me.

pixelinfected
03-29-2004, 03:49 AM
new dev team are a professional developer, but they start with a difficult situation, for few time of releasing, for pressure of all user, for pressure to continue a work of over 12 years of development, i think new dev team marvel us.

Exper
03-29-2004, 04:10 AM
Originally posted by pixelinfected
i think new dev team marvel us.Agree!
Just a little more patience! ;)

pixelinfected
03-29-2004, 05:12 AM
they need only a time to restart the work and organize a correct develop line.
they will do a lot of hard work, and need time to organize the line in the best way.

welcome new dev team, we attend to be amazed since two or three years from now, which new generation.
good work guys.

mattclary
03-29-2004, 06:26 AM
Originally posted by tokyo drifter
It's funny that after that moment of sanity, there wasn't any discussion about it in that thread. Seems like an important statement to me.

That's usually the way it works. Once you have some authoritative information, it shuts down speculation and rumor.

Exper
03-29-2004, 07:51 AM
Originally posted by mattclary
That's usually the way it works. Once you have some authoritative information, it shuts down speculation and rumor. Better than another moaning stream! ;)

Nemoid
03-29-2004, 08:46 AM
BTW i just replied to Chuck in the CGtalk thread above, that IMO, rewriting Lw into a PARALLEL pipeline could be better. however since this is a need, we can "talk" for months, but it have to be done. this work could require 3 years, and what to do in the meantime? enhancing lw with the current structure (knowing for sure that the rewrite will come, though.)

when Pixelinfected was talking about money, it was because a better management strategy of Lw would increase Nt capability to develop a rockin' solid app with ease. and assure in the meantime the users that

EVERY year they will have:
a major update- of the app with the current structure,
and EVERY month they will have :
a bug fix for FREE, with at least ONE bug fixed.

this will allow Nt to earn more money, to restore users faith in their good products and to develop the app better.
parallel or not parallel rewrite Lw pipeline.

i.e. - a good marketing and developing strategy IMHO.

Exper
03-30-2004, 08:10 AM
A total rewrite from scratch at any point is not something under consideration, but a gradual change of the architecture with intermediate versions is much closer to what we are aiming for. ....

Chuck BakerI'm agree with Nemoid...
but also this Chuck's statement is good step forward! :cool:

Nemoid
03-30-2004, 10:22 AM
Sure it is. :)

Jace-BeOS
04-20-2004, 06:48 AM
There are two things that lead me to become a Lightwave user at work and home:

1. The UI. Despite being proprietary and poorly organized/unified, the UI stays out of the way and lets you get work done mostly quickly (my real gripe is that many related settings are spread all over the place instead of being together). 3DS Max is an example of a UI that uses API widgets and standards but still sucks because it wastes so much of the user's time on manipulation of the UI instead of the content (the sliding panels being horrendous). I became comfortable with the UI in Lightwave much faster than any other 3D product I've used (I've tried a bunch and find them all mostly either too cute, like Carrara and Bryce, or too complicated, such as 3DSMax and Maya. LW just happened to work out well for me, even with the admission of having to hunt around a bit for things and customize the UI due to the defaults being less than ideal.

2. The renderings I've seen created with it. I love the look of the rendering engine's output (when people use it right, I should add, since I'm still new at this). Some of my absolute favorite renderings are done with Lightwave, including the fx in Firefly, Sci-FI Channel's Battlestar Galactica remake and Babylon 5 (though B5 sure has a dated look these days ;-) )

Oops... One more thing:

3. Cost.

The quality and cost ratio of Lightwave is second to none. Carrara comes close, but as I said the UI is way too far on the "reinvent the wheel" and cutesy side for me, as well as its turbulent history and much slower and less capable core.

Even if Lightwave were more expensive (which I am glad it is not), I would still be more interested in it than any of the others, including Maya, simply because I like the work I've seen done with it and I have an easy time using it.

If you like the quality of output, does it really matter all that much how all the internals work? I don't hang around while it renders, so after I do a few low quality previews, I set it in motion and go do something else. Speed isn't really an issue for me. I'll take "faster" whenever it comes. I'd like more physical accuracy (such as light behaviors... I don't think there's any way to model a laser light beam reflecting off of several mirrors), but again, the rendering engine is flexible enough that there's usually something to do to mimic what's needed.