Maxwell Render | Official public forum

Maybe I've overlooked this but I haven't seen or heard anything about where Maxwell is going in terms of development. I've seen a video about GPU rendering a few years back but never heard anything about that since then. Where is Maxwell Render going with new features and improvements? I would very much like to get an idea of what the dev team is thinking about and what will be added in the future.

+1 for more active devs here at forum...

"If all goes as planned, 2016 is going to be a very exciting year."
Juan Cañada - interview Advancing the future

That's all we know

To be honest after buyin the educational i am a bit clueless on whats coming next, i know they cant say much but.. my major issue so far being the speed. With so many alternatives on the market i am just a bit worried.
Also forum look a bit desert at a times, i might be wrong but thats the first impression.

T0M0 wrote:+1 for more active devs here at forum...

"If all goes as planned, 2016 is going to be a very exciting year."
Juan Cañada - interview Advancing the future

That's all we know

Great interview, the questions often have more lines than the answers..

In my opinion Next Limit should keep it's costumers up to date about development, at the very least to give them an idea what kind of hardware to upgrade to, to be ready for the future.

Max wrote:To be honest after buyin the educational i am a bit clueless on whats coming next, i know they cant say much but.. my major issue so far being the speed. With so many alternatives on the market i am just a bit worried.
Also forum look a bit desert at a times, i might be wrong but thats the first impression.

so much this.

competition like corona for example is just crazy good right now.
maybe because its young - but still....

im still really interested in whats to come though.

clemens_at wrote:

Max wrote:To be honest after buyin the educational i am a bit clueless on whats coming next, i know they cant say much but.. my major issue so far being the speed. With so many alternatives on the market i am just a bit worried.
Also forum look a bit desert at a times, i might be wrong but thats the first impression.

so much this.

competition like corona for example is just crazy good right now.
maybe because its young - but still....

im still really interested in whats to come though.

i was thinking more about Redshift, Octane, gpu based. Also yeah corona is crazy fast.
I am converting some scenes from maxwell to redshift and rendering them, and frankly i dont see much difference in quality that justify the insane amount of hours to render that.
A scene maxwell rendered in 4 hours on my i7990x, got rendered in 7 minutes on Redshift. I know apples and oranges, but i think the speed of maxwell been overlooked for quite some time now.
We need a huge boost, cant expect everyone to sink money into renderfarms.

Im on hold with the educational before i commit to this render Engine because at the moment is completely unusable for my freelance works, due to the crazy render times.

I agree, the only thing keeping me from moving on to v.3 from 2.7, is the fact that speed remains the main issue with maxwell....

Yes, at least some rough roadmap would be helpful - especially regarding GPU.
I need to upgrade my hardware and it would be good to know if i need more CPU cores in the future or some bigger GPUs.

The be honest, I think the 100% physically correct simulation of light is a very nice feature to have. But at the moment it is not being brought as a choice, you have to use it or don't use Maxwell at all. And this is were Maxwell's problem lies and gives other render engines a big advantage. If I had to guess then 90% of the Maxwell users are using Maxwell because of the great quality and ease of use. That unparalleled quality comes with the price of render time. A vast(!) amount of this goes into simulating the light just as it would in the real world, but 80 or maybe 90% of Maxwell's users do not see the difference in that small part of increased quality. Comparison renders have been done lots of times and I often have to look very hard to see the difference myself, most of the time it's after one of the Next Limit team members point out where to look before I see the difference in quality.

So this is what Corona Render is doing so well. They make a render engine that puts out great quality, just like Maxwell, and leave out the calculations you don't really need. This is what MOST people want, most of the people in the CG community want great quality in a normal time frame. They care much less about the render being a 100% physically accurate. And why would you, if even most of us don't see the difference then the client will never notice it. There have been many discussions about this and Maxwell is very clear about sticking with the 100% accurate simulation of light. This is their choice and I respect that but I ask myself what would happen if you would give the users a choice. When you start Maxwell you need to chose if you are going to take the physically correct road and render 10x longer or the somewhat less accurate road (like Corona does) and render very fast. I think that most users would permanently switch to the latter, and stay in that mode forever. Why? Because the quality of what Corona puts out is good enough for 99% of the CG jobs. There is a very limited number of people who produce images that are truly photo realistic (for example Mike Verta) and make use of the 100% physically correct feature.

I don't know what is coming up for Maxwell this year but I hope it is something like GPU rendering or some other speed improvement because we as users need it, and Maxwell needs it even more.

Tok_Tok wrote: There have been many discussions about this and Maxwell is very clear about sticking with the 100% accurate simulation of light.

The absolute physical correctness of Maxwell is only an illusion because there are these fundamental limitations regarding some things like dielectrics that make it impossible to use them in real life. And i think speaking about physical correctness regarding scenes that use AGS or other fakery is self-deception. In my opinion it would be much more useful to have a real looking biased (material) solution for these special cases that you can use "at your own risk" ...but yes, this has been discussed for many years now.

Corona engine is fast because developers are communicating with the community (also have known professionals & artists corresponding), have fun and are also working hard, implementing from the latest advances in CG area (BDPT VCM, denoise, CPU tech...), have road map planned out, beta testing daily builds... basically the board and forums are alive, fit and healthy! As are many other competitive company boards (RedShift, SideFX, Blender, ChaosG., Foundry, Pixar, Otoy...)

Am wondering as momentarily it looks as if MaxwellR of NextLimit took the Fryrender/Arion path?! Silence for the cure? Customer base sufficient (SketchUp, Rhino users mainly (?) rarely anyone animating? Or maybe orientation tends to tilt toward fluids for now.

Hope things will soon clear out. Half a year has almost pass, yet no excitement...

... tho Thanks, nonetheless.

I would love to come back to maxwell but since there is no blender support, I´m moving on with luxrender, that has an excellent in house integration into blender..

T0M0 wrote:+1 for more active devs here at forum...

"If all goes as planned, 2016 is going to be a very exciting year."

That's all we know

You mean, the text overflowing modal alerts and windows will receive line-breaks so it can be read in its entirety?

I've been using maxwell since I think 2008. it was my secret weapon for ultimate results. but in the last 3 years i can see the competition getting very close. i didn't upgrade to v.3 as i couldn't find enough reasons for this over 2.7 (at least for me...). i said to myself, gpu rendering starts to catch up, let's wait a bit. but until now, all I see is fast and pretty damn good results, but with others running GPU rendering....I still wait for the weekend for rendering. so I think we all wanna know what is planned. should I expect a maxwell GPU rendering release soon, or prepare myself to work in different ways? as for today, maxwell is still the best imho, but it's no secret weapon anymore as others are very good and much faster...

choo-chee wrote:I've been using maxwell since I think 2008. it was my secret weapon for ultimate results. but in the last 3 years i can see the competition getting very close. i didn't upgrade to v.3 as i couldn't find enough reasons for this over 2.7 (at least for me...). i said to myself, gpu rendering starts to catch up, let's wait a bit. but until now, all I see is fast and pretty damn good results, but with others running GPU rendering....I still wait for the weekend for rendering. so I think we all wanna know what is planned. should I expect a maxwell GPU rendering release soon, or prepare myself to work in different ways? as for today, maxwell is still the best imho, but it's no secret weapon anymore as others are very good and much faster...

Totally agree