OTOY Forums

Proupin wrote:"Vray RT is unbiased so it will not use any of the render engines available such as irradiance map or light cache, if these are set, Vray RT will simply ignore them." source. They're probably lying though... and it's not only GPU accelerated, you can actually choose CPU as well, haven't you heard?

Maxwell is based on MLT, not some generic "gi algorithm" (source), and it's the same method used by Kerkythea, Indigo and LuxRender. Maxwell has passes, then the others could have them too... I'm guessing GPU/Cuda/Vram limitations make this hard or something. Maybe RS can somehow specify what they think will be supported in the future regarding passes...

I agree progressive rendering is just 'the way' unbiased rendering methods get to their respective solutions, it is not the reason, but they all share a progressive render nonetheless in orther to be interpolation (bias) free. That's why imo directlighting rendered this way is unbiased. there are just no bias errors to be found.

There is this question in the fryrender faq: fryrender is an unbiased render engine. What does that mean? too long to paste here, it states that with unbiased methods there are no artifacts caused by interpolation of light samples, like the ones irradiance maps (biased) produce. There is also a similar explanation in wikipedia, and there is a link to a small article which is pretty clarifying... no trace of 'randomly shot rays' or other properties you seem to give to the unbiased concept... do you have any sources for what you're stating? glad to learn from you guys.

normally i do also learn sthg. new while having such a discussion, and did quite a bit, but have to give up on it now.

i misquoted vray rt as comlpetely biased, because this are no longer separate products, what means biased and unbiased are just options. i have the same sources like you, all over the net. they explain raytracing, pathtracing, metropolis light transport, monte carlo methods and how they work together, and what they mean. if you don't think that any of this has to do with the calculation of rays, i can't help

i still haven't gathered much informations about unbiased renderers, that can render passes, especially ambient occlusion, refraction, reflection, shadows (what was all on the wishlist of the posters here). other passes like depth, material characteristics, etc. are pretty straight forward and imo method independent.

only exeption to all that is maxwell, providing shadow and even reflection channels. i don't know how it works, but at least i have found a post quoting a method to fake a shadow pass: http://www.lux-render.net/forum/viewtop ... 738#p23323 (by the way, the luxrender wiki has also nice explanations about unbiased rendering; i even think i have read the word "ray" somewhere). good advantage for maxwell, but even there the reflection pass (right from the maxwell documentation) doesn't look quite like i imagine a reflection pass should look like.

imo the whole discussion was also a bit off, because of a diffuse use of terms (also from my side).

so the point is, octane is unbiased, it uses pathtracing and "population monte carlo" alogrithms, and it is a gpu only renderer. and thus not compareable to renderers that are cpu based, because the use of the gpu for calculations implies specific limitations. and this means not every method available with cpu based calculations can be ported to a gpu. octane was even the first one to implement something mlt "alike" (=pmc) in addition to common pathtracing that most other gpu accelerated renderes still use.

so, if you find gpu based unbiased renderes, which provide several passes, just post'em here - would be of course interesting how they do it...

this makes sense. Really the only reason break out passes are used in production is because traditional cpu renders take so long that you need passes to dial in the look you need, so that you avoid having to re render. I think using something like octane just requires a shift in the way you approach a production. Any plans to support alembic for the stand alone version?

t_3 wrote:... octane was even the first one to implement something mlt "alike" (=pmc) ...

Since you frequent luxrender forums you should already know that SLG had real MLT implemented on 24 april 2011 and the first test release of octane with the new kernel was online on 03 june 2011. Hardly "first" I must say...

vimaxus wrote:

t_3 wrote:... octane was even the first one to implement something mlt "alike" (=pmc) ...

Since you frequent luxrender forums you should already know that SLG had real MLT implemented on 24 april 2011 and the first test release of octane with the new kernel was online on 03 june 2011. Hardly "first" I must say...

uh - oh :/
is it okay to say "the first commercial..." then?

mkirylo wrote:this makes sense. Really the only reason break out passes are used in production is because traditional cpu renders take so long that you need passes to dial in the look you need, so that you avoid having to re render. I think using something like octane just requires a shift in the way you approach a production. Any plans to support alembic for the stand alone version?

I don't think lack of quality is the reason at all for traditional renderers to have passes, it's just another tool that can speed up your work, not to mention enhance a raw render big time, to a point impossible to achieve by any renderer alone. If it saves time, it's useful to have, imo. It'd be like saying we don't need to .zip big files over the Internet since tranfer speeds got much faster.

Try to be positive is all I'm saying, having more features is always good, not a single feature is going to make OR less unbiased or less powerful, don't use passes if you don't want to but it might be useful to others (I personally do not use them that often). Octane is fast but not that fast that you can always afford a re-render... specially without the upcoming plugins which will really start to make this software epic, FINALLY!

just to say Thea render have 5 render passes in unbiased mode (TR1-TR2):
normal, depth, alpha, object id, material id
(need to save as thea picture to have layers added to the color one)
and 7 more in biased mode (BSD):
direct, AO, GI, SSS, reflection, transparence, irradiance

ycarry wrote:just to say Thea render have 5 render passes in unbiased mode (TR1-TR2):
normal, depth, alpha, object id, material id...

That you can render faster with you 3D-app and maybe parallel when the unbiased renderer is working...

face