NOX Renderer

Discuss stuff not about Indigo.
User avatar
CoolColJ
Posts: 1738
Joined: Mon Jun 25, 2007 1:47 pm

Re: NOX Renderer

Post by CoolColJ » Thu May 10, 2012 2:17 am

lycium wrote:
CoolColJ wrote:And interestingly in the Post section, you can turn on/off direct lit and/or GI parts of the image, and even boost or reduce the GI :o
Yet another renderer calling itself unbiased while adding stuff like this :lol:

It really is too bad that users don't generally test if their software really is unbiased...
I'm pretty sure it's unbiased - it looks it, with the way it progressively renders :)
But I'm not sure how they are doing this - path tracing is unbiased right?

User avatar
Zom-B
1st Place 100
Posts: 4700
Joined: Tue Jul 04, 2006 4:18 pm
Location: ´'`\_(ò_Ó)_/´'`
Contact:

Re: NOX Renderer

Post by Zom-B » Thu May 10, 2012 2:22 am

CoolColJ wrote:path tracing is unbiased right?
path tracing is the general methode to solve the light calculation, but its like a car... there are many different that can be used
http://www.indigorenderer.com/what-unbiased-ray-tracer
polygonmanufaktur.de

User avatar
CoolColJ
Posts: 1738
Joined: Mon Jun 25, 2007 1:47 pm

Re: NOX Renderer

Post by CoolColJ » Thu May 10, 2012 2:40 am

In any case, I suppose they could just put the direct and indirect lit parts of the image on separate layers, which would allow you to turn them on/off or change the brightness etc.

Which is what unbiased renderers are doing now with lights

User avatar
galinette
1st Place Winner
Posts: 923
Joined: Sat Jan 09, 2010 1:39 am
Location: Nantes, France
Contact:

Re: NOX Renderer

Post by galinette » Thu May 10, 2012 2:50 am

lycium wrote:
CoolColJ wrote:And interestingly in the Post section, you can turn on/off direct lit and/or GI parts of the image, and even boost or reduce the GI :o
Yet another renderer calling itself unbiased while adding stuff like this :lol:
It really is too bad that users don't generally test if their software really is unbiased...
Yes but, in the CG industry, many people see unbiased renderers as very uncreative... Which led me to the conclusion that creativity is biased.
Having a renderer with both "unbiased like" quality and "biased like" flexibility is still what most people expect from the perfect renderer, even if it sounds strange. Otherwise, you are more targeting optical simulation audience than CG.

Etienne
Eclat-Digital Research
http://www.eclat-digital.com

User avatar
lycium
Posts: 1216
Joined: Wed Sep 12, 2007 7:46 am
Location: Leipzig, Germany
Contact:

Re: NOX Renderer

Post by lycium » Thu May 10, 2012 2:54 am

CoolColJ wrote:I'm pretty sure it's unbiased - it looks it, with the way it progressively renders :)
So if I took a screenshot of a Tetris game, and slowly displayed it with less and less noise, it's "unbiased"?
CoolColJ wrote:But I'm not sure how they are doing this - path tracing is unbiased right?
Not necessarily. It has to converge to a correct solution (of the rendering equation) and do so in a way that is free of systematic error; and you might be surprised to know how many rendering systems fail this basic test.

With Indigo for example, the results of single and bi-directional path tracing, with and without MLT, converge to the same result (not exactly of course, given finite precision on computers and very different computations, but to a very small tolerance). However, how many "unbiased" rendering systems will produce the same result in their render modes, with the path length limit set to just 8 bounces, with PMC stuff enabled and disabled, etc?

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: NOX Renderer

Post by CTZn » Thu May 10, 2012 3:28 am

galinette wrote:Which led me to the conclusion that creativity is biased.
As a skill it's hardly photorealistic ! Imagination is not the real world for sure as it is only a potentializing space :)

You nailed down what CG users expect from a biased renderer. But it's still a very demanding (evil) task to get vibrant images with one I believe. It's not only about technical considerations with unbiased, there's something to the eye simply.
obsolete asset

User avatar
CoolColJ
Posts: 1738
Joined: Mon Jun 25, 2007 1:47 pm

Re: NOX Renderer

Post by CoolColJ » Thu May 10, 2012 10:47 am

lycium wrote:
CoolColJ wrote:I'm pretty sure it's unbiased - it looks it, with the way it progressively renders :)
So if I took a screenshot of a Tetris game, and slowly displayed it with less and less noise, it's "unbiased"?
CoolColJ wrote:But I'm not sure how they are doing this - path tracing is unbiased right?
Not necessarily. It has to converge to a correct solution (of the rendering equation) and do so in a way that is free of systematic error; and you might be surprised to know how many rendering systems fail this basic test.

With Indigo for example, the results of single and bi-directional path tracing, with and without MLT, converge to the same result (not exactly of course, given finite precision on computers and very different computations, but to a very small tolerance). However, how many "unbiased" rendering systems will produce the same result in their render modes, with the path length limit set to just 8 bounces, with PMC stuff enabled and disabled, etc?
The options I described only apply to the MLT implementation in NOX, the path tracing option doesn't allow for this

edit - no in the most recent version you can do it with PT as well, but the results converged like an unbiased renderer
Last edited by CoolColJ on Thu May 10, 2012 11:03 am, edited 1 time in total.

User avatar
lycium
Posts: 1216
Joined: Wed Sep 12, 2007 7:46 am
Location: Leipzig, Germany
Contact:

Re: NOX Renderer

Post by lycium » Thu May 10, 2012 10:52 am

Regardless of how it is done in NOX, having some grain and being progressive doesn't mean a rendering algorithm is unbiased. Also, I'm not saying unbiased rendering is the only or "best" way (if such a thing even exists), but of course we love the quality and correctness :)

Post Reply
23 posts

Who is online

Users browsing this forum: Majestic-12 [Bot] and 20 guests