Any features you'd like to see implemented into Maxwell?
By JTB
#324596
MetinSeven_com wrote:When Next Limit will upgrade Maxwell to support hybrid CPU / GPU processing, I really hope they will keep full integration in the supported 3D editors, such as 3ds Max in my case. My biggest complaint against a hybrid processing renderer like Arion is that you have to export the scene to the external renderer and then all the changes you make (changing camera, settings, etc.) result in a new scene, so you end up with two versions of the same scene. I dislike that. I would very much prefer integration into Max, for example like V-Ray RT: as an ActiveShade renderer that realtime picks up changes you make to your 3ds Max scene.
+1000!
This is the most important....But I am not very optimistic...
User avatar
By Richard
#324598
Shaderlight even has this style of workflow coming out for SU which has to be the hardest of all for direct interaction!
User avatar
By choo-chee
#324599
I agree about the way arion works but I don't care since maxwell is so eay for me to set-up, all i want is SPEED and that's it. I adjust cameras, lights etc counting on my 15 years experience in CG. all I care about is SPEED. that's all that matters when you deal with unbiased engines. you know it will look good. the only question is when .
By stephen6
#325353
I vote for GPU rendering. Radion already has cards with 4GB of ram. This is the wave of the furture, and it's not going to stop. Video cards will always have many more processors than CPUs.

I'm seriously thinking about switching to Arion by Random Control.
Don't make me leave Next Limit!
By JTB
#325365
stephen6 wrote:I vote for GPU rendering. Radion already has cards with 4GB of ram. This is the wave of the furture, and it's not going to stop. Video cards will always have many more processors than CPUs.

I'm seriously thinking about switching to Arion by Random Control.
Don't make me leave Next Limit!
Well, there is another topic here http://www.maxwellrender.com/forum/view ... 97&t=34534 and I think there are many interesting opinions and examples (check the 2-3 last pages.)
User avatar
By Rickyx
#325738
As far as I understand Gpu is only one of the optimizations possible:
http://vimeo.com/12890944

This has been possible not only using Gpu+Cpu but also changing the way rays are calculated.
That means for me that mind goes far than brute force :wink:
By RK_art
#325902
I just tried octane render that is 100 % GPU-based and the render speed simply blows you away. :shock:
This is reeeeeally warp speed on my double-Geforce machine.
The software , material system and the plugin for C4D is still beta or under devellopment, but the current price of 100 $ for a licence and for the speed you get with that is incredible.
Maxwell is more evolved, but I will definetely add octane render to my apps and I think, when octane is in the 1.x versions, Maxwell will have a serious problem when you compare then the price-speed/feature relationship with octane.
Maxwell's licence-system is horrible expensive and only CPU-based, therefore extremely slow, Octane costs a fraction and offers light-speed with a 100% NVIDIA-GPU-based technology.
Don't get me wrong, I really liked Maxwell so far, but the last upgrade-process and the horrible fee I payed for my render-nodes in addition with the slow CPU-based technology makes the decision for a change very easy in case octane holds lately with the 1.x-versions what they promise now with the beta.
User avatar
By MetinSeven_com
#325914
Let's not forget Mental Images Iray, which will probably soon be available to Autodesk subscription customers and be included alongside Mental Ray with the next major releases of Max, Maya and XSI.

Next Limit has indeed some catching up to do, in order to keep up with the growing competition. They will undoubtedly realize that and already working on it, but a little more communication about Maxwell development with their much-paying customers would definitely be an improvement.
By JTB
#325953
MetinSeven_com wrote:Let's not forget Mental Images Iray, which will probably soon be available to Autodesk subscription customers and be included alongside Mental Ray with the next major releases of Max, Maya and XSI.
REALLY ???

http://www.youtube.com/watch?v=z5WyOQe4AHc

CGarchitect was provided access to the online iray demo.
We recorded this screencast to demonstrate the rendering speed and quality.
This demo uses 15 Tesla GPUs :lol: :lol: :lol: . The demo was run from from the Firefox web browser.
By RK_art
#326008
I have to correct my post above.
Maxwell isn't so much slower than Octane according to my first comparison tests.

Here are my results:

Maxwell uses my quadcore 2,66 Ghz with 4 GB RAM, Octane used my 2 NVIDIA 8700m GT with a sum of 64 cores and 2 x 512 MB VRAM.

All texture is simple diffuse material, no SSS (Maxwell) etc.

1) DayLight-System, 3 minutes rendering:

Maxwell:
Image

Octane:
Image


2) HDRI (same for both), 5 minutes rendering:

Maxwell:

Image

Octane:

Image


There is obviously no big difference in the render-qualities.

But btw., beside that Maxwell is surely much more evolved than currently Octane, the price-render-quality-relationsship is clearly on Octane's site. So my second statement from above persists, if Octane evolves in the way I hope and keeps such a good price, hard times are rising for NL's high-(update)-price-politic.
By JTB
#326011
I think a more appropriate subject should be chosen...

So, is this a known issue?

Thanks a lot for your response, I will update and […]

did you tried luxCore?