Maxwell Render | Official public forum

...a faster engine" is really needed at least four times the actual speed...
even a biased version of the engine would be welcome if the speed is there...

why?

...just buy vray if that is what you want....

cheers
stefan

or final render or mental ray or C4D advanced render or modo or brazil or....

I guess what thomas means is that there are customers that don't need or understand Maxwell's quality. Instead of delivering high quality images to someone that won't appreciate them and having problems with deadlines, we could use a typical (like vray,mental ray etc...)engine with less accuracy but the client will be happy to have them in a short time.
This way we won't have to learn two apps,we will use one material library, we will always work with the same program and choose the quality of our results and most important we will save money because let's face it, we all have wished for faster results

I know that we will have improvements with RS2 but even then, a biased RS2 will be a nice tool.

dont worry illab, as soon you 'll make it available i'll buy vrayforc4d,
but what is sure, is that i use maxwell since the begining, and cannot use it for work, because theres too much update ( daily ones) and not enough time to wait the render to complete, and i dont want to hear that i could buy another computer, i already did

mihai i' already got the advanced renderer, but it s also a slow interior renderer ( at least its GI is quite slow )

and yes client understand the qualities of maxwell, and the multilight fonction, but if i've got an update to do, it still will take 4 to 8 times the time i've got...
we should be able to render a 2k interior in 6-8 hours,
hope we'll get a faster engine soon, cause i like the engine and wish to stick with it

thomas,

i just think it makes no sense to push maxwell in a direction where already a lot of renderers are... there is plenty of choice, maybe not for cinema, but soon;-)

NL should concentrate their enrgy to make MW as good as possible in the way it is,a physical correct renderengine. this makes them unique. for non physical correct-biased engines i think final render, MR and vray are the specialists here....there is no sense in copying vray into maxwell. that just makes the engine a mixed thing that probably isnt as good as vray anyways in that field.

maxwell is best and development leader in the new unbiased market and i really think they should concentrate on this.

those mixed "it can do all" things are mostly not good...

...and the quality of maxwell AND the speed of vray isnt realistic i think.

cheers
Stefan

...and the quality of maxwell AND the speed of vray isnt realistic i think.

why not ? sure i like maxwell and its quality, but its biggest flaw is rendertime, i only wish it would be faster ( lets be specific, mostly for interior, because it can be quite production friendly in exterior shot ) compare it to C4d advanced renderer, you can control a bit the Gi or run in stochastic mode ( wich is more brute force )

anyway, we'll see whats coming in the next months...

cheers

i m agree with a thomas . the speed is crutial question .....

my question the next update of mr have a good optimisation of time ?

maxwell is best and development leader in the new unbiased market and i really think they should concentrate on this.

i understand the choice of nl

but

i think if the user can choice unbiased solution or biased solution is very nice solution for professionnal work.

thomas lacroix wrote: and yes client understand the qualities of maxwell, and the multilight fonction, but if i've got an update to do, it still will take 4 to 8 times the time i've got...
we should be able to render a 2k interior in 6-8 hours,

Usually I have to make 3-4 updates. But I don't make the updates at full rez...why should I? It's only the final image that needs to be high rez. If they want to make changes after that, well it's extra.

i think if the user can choice unbiased solution or biased solution is very nice solution for professionnal work.

My point exactly!

well biased solution_doesn not look so good, or needs other workflow= vray, MR,FR etc....this are the "volkswagen" of the renderers

unbiased-maxell-looks better in most cases, is physical correct, renders longer or needs more cpu power - this is the "expensive" ferrari;-)
good but needs a bit extra....

a ferrari that is cheap, real, fast, etc does not exist.

i cannot imagine that maxwell quality can be directly achieved in classic biased (fast) ways. maybe there is some room for optimizing maxwell, but i woud guess not very much.

my2cent
cheers
Stefan

ok ..............................but i go to work wtih ax citroen and if i want for the week end i rent a ferrari

so , no offence but ..............



ags is biase dialetric or unbiase dialectric ?

maybe :

nl could work for another stuff biases ? (as emiter for simulate ies for example )

i know all this mihai ( it depend of the client and job) and illab (you've seen the thread on the vray forum as well i suppose, the one with the car analogy? sure it s a quite good analogy, but when i bought maxwell i wasnt aware that it was a ferrari with a fiat 500 engine in it) ,well to end i was just wishing out loud that maxwell will be able to render faster and will allow me to only wait 10 Hours for a 2 k interior render with my current config , no more no less

cheers

I have to disagree with this request, it makes no sense to take Maxwell down to the same level as any other renderengine just to obtain speed. That would be like rippiing the carpet away from someone made outta glass standing on it. It'd come crashing down in pieces.. please don't ruin it!

If you want a fast cheating solution use another engine.

/ Max

I'm sorry for those who have problems with Maxwell's speed and need to use it for production but i'm all for keeping it unbiased and, if there's no way around it, as slow as it is right now.

With the new batch of 65nm processors we've seen a doubling in performance and the industry has long planned the path to further improvement (mutlicore, 45nm), so i see a good chance there that biased solutions gain popularity.

Btw, for simple studio setups like product shots Maxwell is imo faster than most biased solutions if you take blurry reflections and dof into account, things like architecture are another story.

Keep the research on biased solutions to the dozens of dev-teams who went into that direction for years, let NL do something else