Cornell-box test scene need someone to render it ...

General questions about Indigo, the scene format, rendering etc...
Stromberg
Posts: 693
Joined: Mon Mar 02, 2009 5:01 am
Location: Norway
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by Stromberg » Thu Oct 22, 2009 1:11 pm

Psor: Just to say the "Why do you care" was not ment as rude :)
And the other thing is, you are completly right, i think the same way aswell i use the render engine i feel fits me best :)
Planning to buy maxwell soon tough, but will defently not stop using indigo, i must have the biggest "wip" topic here :P
And the indigo renders posted in the maxwell forum, dosnt serve indigo's right, luckly for me i am a multi fan boy, i like all unbiased engines :D

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

Re: Cornell-box test scene need someone to render it ...

Post by lycium » Thu Oct 22, 2009 1:22 pm

psor, i'm not trying to "start something" with you either, and i think we're saying the same thing - such comparisons will always be clueless sausage fests (the mac vs pc debate is the same). too much bias in discussion of unbiased renderers ;)

User avatar
psor
1st Place Winner
Posts: 1295
Joined: Sun Jun 25, 2006 1:25 am
Location: Berlin
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by psor » Thu Oct 22, 2009 1:27 pm

:D :D :D

Lets do something else guys! :wink:





take care
psor
"The sleeper must awaken"

Stromberg
Posts: 693
Joined: Mon Mar 02, 2009 5:01 am
Location: Norway
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by Stromberg » Thu Oct 22, 2009 1:36 pm

Stop talking and start rendering :wink:

User avatar
psor
1st Place Winner
Posts: 1295
Joined: Sun Jun 25, 2006 1:25 am
Location: Berlin
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by psor » Thu Oct 22, 2009 1:42 pm

Stromberg wrote:Stop talking and start rendering :wink:
Quoted for agreement. Fortunately I'm already playing with the material editor for some time
and since I'm using an unbiased renderer I've some spare time that I spend doing something
else ie reading topics. :lol: :wink:





take care
psor
"The sleeper must awaken"

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by suvakas » Thu Oct 22, 2009 9:17 pm

Tested the scene too.
Bidir MLT, a bit over 6000 spp. I see no dark edges?
Attachments
cornell_reflective.png
cornell_reflective.png (559.45 KiB) Viewed 3595 times

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

Re: Cornell-box test scene need someone to render it ...

Post by Zom-B » Thu Oct 22, 2009 10:10 pm

suvakas wrote:I see no dark edges?
The reason why some edges appear in the maxwell forum test is simply by using a "lowPoly" sphere and not the sphere primitive for a perfect round shape...
polygonmanufaktur.de

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by suvakas » Thu Oct 22, 2009 10:12 pm

Yes, that's probably the reason.

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

Re: Cornell-box test scene need someone to render it ...

Post by CTZn » Fri Oct 23, 2009 5:01 am

Even a cube can be subdivided :roll:
obsolete asset

User avatar
pixie
Indigo 100
Posts: 2332
Joined: Sat Dec 29, 2007 4:54 am
Location: Away from paradise
3D Software: Cinema 4D
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by pixie » Fri Oct 23, 2009 8:52 am

hybrid.jpg
hybrid 6300spp
hybrid.jpg (60.1 KiB) Viewed 3487 times

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by suvakas » Fri Oct 23, 2009 9:01 am

I also rendered one with the latest Luxrender v0.6 Final.
It took 11 hours on 8 cores and reached 4600 S/px using metropolis sampler + bidir. And it's still not clean. I must admit, that I was expecting a bit better results from Lux. It also has this black ring around the glass sphere.

Ps. My Indigo image took 9 hours on 4 cores to reach 6000 spp.

Here's the Lux image:
Attachments
Lux_cornell.png
Lux_cornell.png (555.21 KiB) Viewed 3482 times

User avatar
psor
1st Place Winner
Posts: 1295
Joined: Sun Jun 25, 2006 1:25 am
Location: Berlin
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by psor » Fri Oct 23, 2009 9:06 am

Ufff, ... nine hours is still a lot! Quite more then I did expect. What settings did you use
for the lux rendering? I mean how many bounces etc. for bidir and what kinda MLT settings.




take care
psor
"The sleeper must awaken"

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by suvakas » Fri Oct 23, 2009 9:27 am

psor wrote:Ufff, ... nine hours is still a lot! Quite more then I did expect. What settings did you use
for the lux rendering? I mean how many bounces etc. for bidir and what kinda MLT settings.
I have no idea when the Indigo image was clean. I just ran it over night and it showed 9 hours when I closed it. Usually 3000 spp should be enough.

Here are my Lux settings:

Code: Select all

PixelFilter "mitchell"
   "float B" [0.750000]
   "float C" [0.125000]

Sampler "metropolis"
   "float largemutationprob" [0.400000]

SurfaceIntegrator "bidirectional"
   "integer eyedepth" [16]
   "integer lightdepth" [16]

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

Re: Cornell-box test scene need someone to render it ...

Post by CTZn » Sat Oct 24, 2009 4:32 am

How much does the depth settings impact lux ? I'm asking because at 2 pow 16 (65k) it is much more than Indigo default (10k iirc). If I'm understanding them correctly, 13 or 14 would be closer.

Also I don't know how Indigo is managing the one parameter between the two directions, ie dunno if it is per sampling direction (for each of light>eye and eye>light) or for both alltogether.
obsolete asset

User avatar
jotero
Posts: 28
Joined: Wed Aug 16, 2006 4:14 am
Location: Hannover Germany
Contact:

Re: Cornell-box test scene need someone to render it ...

Post by jotero » Sat Oct 24, 2009 8:09 pm

hello all :)

Here is my final version of the Cornellbox 2! "MLT Bidirectional"
Image
The time is not so important to me.

ciao
torolf
[ Jotero Gallery ]
::everything my models (mathe and 3d-scan) not for Commercial use!::

Post Reply
48 posts

Who is online

Users browsing this forum: Majestic-12 [Bot] and 17 guests