[tom]

Imagine you're a product designer having to render such a thing (just like above dragon example):
http://lh4.ggpht.com/_i3R-qMVpIRw/RhRX- ... f00051.jpg
http://farm2.static.flickr.com/1290/644 ... b94bb8.jpg
http://www.lotioncrafter.com/files/cach ... f8d048.jpg
http://image.ec21.com/image/globalglass ... Powder.jpg
http://www.tradeplumbing.co.uk/assets/i ... e-fros.jpg
http://www.stonesink.net/images/gd-41frosted.gif

Would you say your engine is not capable of doing this and so you?
Simulation is not all about textured walls and plastic surfaces...
Today there are realtime (literally) gaming engines doing what you say at that level.
A physically-based render engine should do something more so the designer
would be able to make the proof of design. I'm still waiting for the render of
above dragon but of course you won't care if you're after rendering only walls.

What's different here?
http://igad.nhtv.nl/~bikker/
http://www.youtube.com/watch?v=IH11D7qYh10
http://www.unrealtechnology.com/feature ... =rendering

[Tok_Tok]

So for now it seems GPU renderers are best for previewing purposes. How long does it take to have a finished render, also with refractive materials or reflected caustics? At what resolution? But even for previewing purposes how does the export/import work? If you want to just translate/scale/rotate any geometry, how long does the export take each time for a moderately complex scene before you can actually start to get the 'realtime' preview? Or say you want to reposition an emitter in the scene. How many of the GPU renderers currently have real emitters and not only IBL?

You are talking like Octane exists for years and it's very strange that they don't have these features yet. Ofcourse they don't have it because their still in a Beta stage, you simply can't expect
them to be as good as Maxwell. But what they do have is the speed and from there they can build up. Over time they will implement the features Maxwell also has, but when Maxwell chooses not to
use GPU rendering now, they will lose in the end because they still lack the speed.

As i understand, for now, Octane doesn't have caustics, transparant shadows and caustics(correct me if i'm wrong). Also rough glass isn't one the features yet, that's why no one has shown the dragon Tom.

For now GPU renderers are great for preview renders, but as soon as all improves they will be just as good as Maxwell. Maybe not 100% physically correct but i wouldn't mind.

[big K]

tom, i don´t know what this is about.
i am pretty sure that there are scenes where maxwell is still struggeling (pool caustics and sun through dialectrics...) and other engines don´t, so what do you want to proove with this one example?
if it can´t render this it is bad ? (this was said about maxwell as well in the past; see above)

glass with roughness and the like is about to be implemented on the next update of Arion (i think, don´t hit me if i am wrong, but it is for sure on the to do list)

of course, there scenes that get into trouble with the GPU renderers. (Arion has for example no displacment so far)
i have shown some examples, there are also quite a lot of production design examples. (which normally is even better for Arion as less materials and ram is needed)

[tom]

As i understand, for now, Octane doesn't have caustics, transparant shadows and caustics(correct me if i'm wrong). Also rough glass isn't one the features yet, that's why no one has shown the dragon Tom.

Forget it. Which other unbiased engine do you think would achieve it?

tom, i don´t know what this is about.

This is all about you're comparing apples with oranges. When they are at the same quality level, wake me up and then I will let you know where Maxwell is.
Until then >>



We're always after more speed as everybody else but, we will not degrade our quality and/or reduce or core features for jumping on GPU, XPU, etc immaturely.

[Half Life]

I love it -- but it won't work... they've got their mind so focused on speed that any argument that uses quality or feature comparison is like throwing water on a duck.

I've said it before I want REAL performance -- to me that means the best bang for my buck and my time, Maxwell provides that in spades...

Best,
Jason.

[JTB]

This thread started as a simple question.... What are we waiting from NL... not if Octane or Arion or... is better than Maxwell... This is not a public discussion forum, we all are NL clients... so we've already chosen Maxwell... No matter what problems GPU rendering may have today, we all want it.... as preview renderings? as final renderings? it is not very important...
There is one thing that comes to mind though, and it seems to be accepted by some companies... Biased and unbiased methods in one application... Octane, Thea Render are the two examples I know of... And with GPU it is probably a new fashion since biased rendering can have great and instant results while unbiased methods can be used for the final product.
NL has already stated that there will be no biased Maxwell, ever.... Maybe they have to reconsider because Maxwell's quality and easy of use, the fast scene setup, the great materials, multilight.... combined with GPU and CPU power and especially with a liiittle biased preview engine, would be a great product...

[tom]

NL has already stated that there will be no biased Maxwell, ever.... Maybe they have to reconsider because Maxwell's quality and easy of use, the fast scene setup, the great materials, multilight.... combined with GPU and CPU power and especially with a liiittle biased preview engine, would be a great product...

No, there will be no biased Maxwell, ever. However, Maxwell may offer faster preview and final render solutions with or without GPU or with something else. The only fact you should always know is Maxwell will not fake things, reduce its abilities to be faster. This is and was always possible if we think otherwise.

[Tok_Tok]

I love it -- but it won't work... they've got their mind so focused on speed that any argument that uses quality or feature comparison is like throwing water on a duck.

The point is, that right now you are comparing apple with oranges, i totally agree. But how long do you think it will take for them to catch up. They don't have to reinvent the wheel.
I'm not blinded by speed and i'm not saying Octane is better, i'm simply trying to prove that GPU rendering isn't such a bad thing and Next limit should look at it from a objective point of view.
Not looking at how bad it is and what it can't do, but how it can make Maxwell better!

Looking at the reactions from Tom and Mihai Nextlimit isn't going to implement it, so i think the question is awnsered.

[tom]

Looking at the reactions from Tom and Mihai Nextlimit isn't going to implement it, so i think the question is awnsered.

We underline something clearly that we're not going to implement X technology (and this is not limited to today's GPUs) if such a technology is not mature enough to port Maxwell into. Today's GPU's and softwares supporting them may look exciting to you but it's nothing new to us and we have plenty of information/experience on this area in NL labs. It does not mean we are against GPU support or speed. We won't make a version of Maxwell just for rendering with half of its features but, you should always expect a fully armed leap from Maxwell using all features instead. This could be GPU or not, we're not excluding any technology at all.

[Frances]

"If I had learned to play [the piano], I should have been a great proficient." - Lady Catherine de Bourg Pride and Prejudice

[tom]

"It's different. When you perform with an orchestra it's different than performing with a piano soloist. They really have to be in sync with the conductor." ~ Marc Tourre

[Tok_Tok]

"There is a tide in the affairs of men, which, taken at the flood, leads on to fortune." - William Shakespeare

[Mihai]

You are talking like Octane exists for years and it's very strange that they don't have these features yet. Ofcourse they don't have it because their still in a Beta stage, you simply can't expect them to be as good as Maxwell.

And Arion has been in development how long? (just as another example). You agree most if not all GPU renderers that you're excited about are still too immature, but the big question here is will they ever be mature following this path, and in the mean time what will happen to CPUs?

But what they do have is the speed and from there they can build up.

In my opinion it's exactly the other way around. You start with the highest quality and try to optimize to gain speed, but if you start by cutting corners, you are already in a dead end and you have to make a lot of changes while still trying to figure out how to retain the speed you advertised to initially attract users. Quite impossible and the wrong way to start. Why do you think there are suddenly so many GPU renderers, all with about the same limitations/feature set so far? It makes me wonder. It also makes me realize it's much easier to start showing something when you have 40% of the usual features/workflow you're accustomed to in a CPU renderer but adding more than that doesn't seem an easy if possible task currently with GPUs.

As i understand, for now, Octane doesn't have caustics, transparant shadows and caustics(correct me if i'm wrong). Also rough glass isn't one the features yet, that's why no one has shown the dragon Tom.

For now GPU renderers are great for preview renders, but as soon as all improves they will be just as good as Maxwell. Maybe not 100% physically correct but i wouldn't mind.

"As soon as" implies what? Unknown. I agree they look good as a preview render, but even for this the workflow currently seems very lacking (disregarding all the missing features). Every time you want to make a small change you have to re-export your scene? How long does that take each time?

[Lars Magnusson]

We underline something clearly that we're not going to implement X technology (and this is not limited to today's GPUs) if such a technology is not mature enough to port Maxwell into. Today's GPU's and softwares supporting them may look exciting to you but it's nothing new to us and we have plenty of information/experience on this area in NL labs. It does not mean we are against GPU support or speed. We won't make a version of Maxwell just for rendering with half of its features but, you should always expect a fully armed leap from Maxwell using all features instead. This could be GPU or not, we're not excluding any technology at all.

Looking through the pages of this thread, and this was what I wanted to hear!

I'm all for using the GPU as a calculating complement to the CPU ( I'm just about to place an order for a new 6core machine with a GTX480)
But I bought Maxwell because the unbiased reality it can deliver.
So, I hope that if GPU support will come in the future, that Maxwell just handles it as if it was additional CPU cores, delivering part of the calculations to them.
(Is that even possible? The Cuda cores in an Nvidia card...can they really handle the same type of calculations as a CPU...?)

I was extremely disappointed when my previous render-software trueSpace (with a rather limited VRay) started to move more and more into realtime graphics.
Same with the CAD software I work as a reseller for, it uses the Mentalray render but instead of giving me all the functionality that Mentalray can offer, the focus now seem to be on realtime render. (and since my worklaptop isn't even close to the graphic performance I have at home I can barely use the realtime stuff.)

That said, my situation is different from most of you.
I bought Maxwell for fun/hobby use, and most of you use it commersially with deadlines and client expectations to consider, so I understand the urge you have for speed.
(sure, it doesn't actually matter to me if my "just for fun and private use" image takes all night to render or even more, but it would be cool if the final noise free result could be reached in a fraction of the time, since I tend to sit and stare at the renderwindow, fascinated with how the dots slowly builds up the image....it's like meditation )

[Frances]

"It's different. When you perform with an orchestra it's different than performing with a piano soloist. They really have to be in sync with the conductor." ~ Marc Tourre

I should not have expected you to understand the reference.