LuxRender 0.9-dev Linux build of 04-02-2012

Weekly builds for testing and use between releases.

Moderator: coordinators

Forum rules
Please read the information / sticky post for some basic information regarding these builds/support.

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby LadeHeria » Fri Feb 10, 2012 11:37 am

When I say don't work, it means that rendering don't start.
The image in Luxrender is black and all the counts are to 0. My CPU/GPU is also idle-ing.


LadeHeria
LadeHeria
 
Posts: 129
Joined: Fri Aug 27, 2010 8:59 am

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby SATtva » Fri Feb 10, 2012 11:55 am

Is there anything in the Lux log?
Btw, do you observe this behavior with latest builds only? What about the previous ones?
Linux builds packager
聞くのは一時の恥、聞かぬのは一生の恥
User avatar
SATtva
Developer
 
Posts: 5487
Joined: Tue Apr 07, 2009 12:19 pm
Location: from Siberia with love

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby rendermagic » Fri Feb 10, 2012 12:31 pm

I finally had a chance to run a comparison test between the NoOpenCL build and the OpenCL build. It appears that the GPU does add more samples per pixel, but by only a 34% speed increase. In both cases, I am using path tracing surfaceintegrator and the metropolis sampler.

Although the GPU is definitely adding much greater number of 'contributions per second' and the efficiency is much higher, it doesn't appear that as much is impacting the number of samples per pixel.

Code: Select all
CPU Only
   [Lux 2012-Feb-10 10:41:56 INFO : 0] 00:00:05 - 16T: 0.60 S/p  126.14 kS/s  184% Eff  231.91 kC/s - 0% Complete (S/Px) - 02:37:43 Remaining
   [Lux 2012-Feb-10 10:42:01 INFO : 0] 00:00:10 - 16T: 1.20 S/p  133.33 kS/s  184% Eff  245.59 kC/s - 0% Complete (S/Px) - 02:29:07 Remaining
   [Lux 2012-Feb-10 10:42:06 INFO : 0] 00:00:15 - 16T: 1.81 S/p  135.89 kS/s  184% Eff  250.34 kC/s - 0% Complete (S/Px) - 02:26:13 Remaining
   [Lux 2012-Feb-10 10:42:11 INFO : 0] 00:00:20 - 16T: 2.40 S/p  136.69 kS/s  184% Eff  251.80 kC/s - 0% Complete (S/Px) - 02:25:17 Remaining
   [Lux 2012-Feb-10 10:42:16 INFO : 0] 00:00:25 - 16T: 3.00 S/p  137.14 kS/s  184% Eff  252.70 kC/s - 0% Complete (S/Px) - 02:24:43 Remaining
   [Lux 2012-Feb-10 10:42:21 INFO : 0] 00:00:30 - 16T: 3.60 S/p  137.50 kS/s  184% Eff  253.36 kC/s - 0% Complete (S/Px) - 02:24:15 Remaining
   [Lux 2012-Feb-10 10:42:26 INFO : 0] 00:00:35 - 16T: 4.20 S/p  137.77 kS/s  184% Eff  253.82 kC/s - 0% Complete (S/Px) - 02:23:53 Remaining
   [Lux 2012-Feb-10 10:42:31 INFO : 0] 00:00:40 - 16T: 4.80 S/p  138.15 kS/s  184% Eff  254.54 kC/s - 0% Complete (S/Px) - 02:23:25 Remaining
   [Lux 2012-Feb-10 10:42:36 INFO : 0] 00:00:45 - 16T: 5.39 S/p  138.21 kS/s  184% Eff  254.56 kC/s - 1% Complete (S/Px) - 02:23:16 Remaining
   [Lux 2012-Feb-10 10:42:41 INFO : 0] 00:00:50 - 16T: 6.00 S/p  138.47 kS/s  184% Eff  255.04 kC/s - 1% Complete (S/Px) - 02:22:54 Remaining
   [Lux 2012-Feb-10 10:42:46 INFO : 0] 00:00:55 - 16T: 6.60 S/p  138.56 kS/s  184% Eff  255.22 kC/s - 1% Complete (S/Px) - 02:22:44 Remaining
   [Lux 2012-Feb-10 10:42:51 INFO : 0] 00:01:00 - 16T: 7.20 S/p  138.68 kS/s  184% Eff  255.44 kC/s - 1% Complete (S/Px) - 02:22:32 Remaining
   [Lux 2012-Feb-10 10:42:56 INFO : 0] 00:01:05 - 16T: 7.79 S/p  138.70 kS/s  184% Eff  255.47 kC/s - 1% Complete (S/Px) - 02:22:25 Remaining
   [Lux 2012-Feb-10 10:43:01 INFO : 0] 00:01:10 - 16T: 8.40 S/p  138.86 kS/s  184% Eff  255.76 kC/s - 1% Complete (S/Px) - 02:22:10 Remaining
   [Lux 2012-Feb-10 10:43:06 INFO : 0] 00:01:15 - 16T: 8.99 S/p  138.87 kS/s  184% Eff  255.77 kC/s - 1% Complete (S/Px) - 02:22:04 Remaining
   [Lux 2012-Feb-10 10:43:11 INFO : 0] 00:01:20 - 16T: 9.59 S/p  138.95 kS/s  184% Eff  255.90 kC/s - 1% Complete (S/Px) - 02:21:55 Remaining
   [Lux 2012-Feb-10 10:43:16 INFO : 0] 00:01:25 - 16T: 10.19 S/p  138.91 kS/s  184% Eff  255.81 kC/s - 1% Complete (S/Px) - 02:21:52 Remaining
   [Lux 2012-Feb-10 10:43:21 INFO : 0] 00:01:30 - 16T: 10.79 S/p  139.02 kS/s  184% Eff  256.00 kC/s - 1% Complete (S/Px) - 02:21:40 Remaining
   [Lux 2012-Feb-10 10:43:26 INFO : 0] 00:01:35 - 16T: 11.39 S/p  139.10 kS/s  184% Eff  256.12 kC/s - 1% Complete (S/Px) - 02:21:31 Remaining
   [Lux 2012-Feb-10 10:43:31 INFO : 0] 00:01:40 - 16T: 11.99 S/p  139.14 kS/s  184% Eff  256.20 kC/s - 1% Complete (S/Px) - 02:21:23 Remaining
   [Lux 2012-Feb-10 10:43:36 INFO : 0] 00:01:45 - 16T: 12.59 S/p  139.14 kS/s  184% Eff  256.19 kC/s - 1% Complete (S/Px) - 02:21:18 Remaining
   [Lux 2012-Feb-10 10:43:41 INFO : 0] 00:01:50 - 16T: 13.19 S/p  139.21 kS/s  184% Eff  256.32 kC/s - 1% Complete (S/Px) - 02:21:09 Remaining
   [Lux 2012-Feb-10 10:43:46 INFO : 0] 00:01:55 - 16T: 13.79 S/p  139.26 kS/s  184% Eff  256.40 kC/s - 1% Complete (S/Px) - 02:21:01 Remaining
   [Lux 2012-Feb-10 10:43:51 INFO : 0] 00:02:00 - 16T: 14.39 S/p  139.22 kS/s  184% Eff  256.33 kC/s - 1% Complete (S/Px) - 02:20:58 Remaining
   [Lux 2012-Feb-10 10:43:56 INFO : 0] 00:02:05 - 16T: 14.98 S/p  139.18 kS/s  184% Eff  256.24 kC/s - 1% Complete (S/Px) - 02:20:56 Remaining
   [Lux 2012-Feb-10 10:44:01 INFO : 0] 00:02:10 - 16T: 15.59 S/p  139.28 kS/s  184% Eff  256.43 kC/s - 2% Complete (S/Px) - 02:20:44 Remaining
   [Lux 2012-Feb-10 10:44:06 INFO : 0] 00:02:15 - 16T: 16.18 S/p  139.29 kS/s  184% Eff  256.45 kC/s - 2% Complete (S/Px) - 02:20:39 Remaining
   [Lux 2012-Feb-10 10:44:11 INFO : 0] 00:02:20 - 16T: 16.78 S/p  139.28 kS/s  184% Eff  256.40 kC/s - 2% Complete (S/Px) - 02:20:35 Remaining
   [Lux 2012-Feb-10 10:44:16 INFO : 0] 00:02:25 - 16T: 17.38 S/p  139.30 kS/s  184% Eff  256.46 kC/s - 2% Complete (S/Px) - 02:20:28 Remaining
   [Lux 2012-Feb-10 10:44:21 INFO : 0] 00:02:30 - 16T: 17.98 S/p  139.30 kS/s  184% Eff  256.42 kC/s - 2% Complete (S/Px) - 02:20:23 Remaining
   [Lux 2012-Feb-10 10:44:26 INFO : 0] 00:02:35 - 16T: 18.58 S/p  139.31 kS/s  184% Eff  256.44 kC/s - 2% Complete (S/Px) - 02:20:18 Remaining
   [Lux 2012-Feb-10 10:44:31 INFO : 0] 00:02:40 - 16T: 19.17 S/p  139.29 kS/s  184% Eff  256.41 kC/s - 2% Complete (S/Px) - 02:20:14 Remaining

Hybrid
   [Lux 2012-Feb-10 10:45:31 INFO : 0] 00:00:06 - 16T: 0.00 S/p  0.00 S/s  329% Eff  0.00 C/s - GPU Load: 26%
   [Lux 2012-Feb-10 10:45:36 INFO : 0] 00:00:11 - 16T: 0.91 S/p  93.38 kS/s  300% Eff  279.68 kC/s - 0% Complete (S/Px) - 03:32:59 Remaining - GPU Load: 54%
   [Lux 2012-Feb-10 10:45:41 INFO : 0] 00:00:16 - 16T: 1.73 S/p  123.65 kS/s  299% Eff  369.38 kC/s - 0% Complete (S/Px) - 02:40:43 Remaining - GPU Load: 65%
   [Lux 2012-Feb-10 10:45:46 INFO : 0] 00:00:21 - 16T: 2.55 S/p  139.07 kS/s  299% Eff  415.60 kC/s - 0% Complete (S/Px) - 02:22:47 Remaining - GPU Load: 71%
   [Lux 2012-Feb-10 10:45:51 INFO : 0] 00:00:26 - 16T: 3.38 S/p  149.66 kS/s  299% Eff  446.95 kC/s - 0% Complete (S/Px) - 02:12:34 Remaining - GPU Load: 75%
   [Lux 2012-Feb-10 10:45:56 INFO : 0] 00:00:31 - 16T: 4.22 S/p  156.88 kS/s  298% Eff  468.00 kC/s - 0% Complete (S/Px) - 02:06:22 Remaining - GPU Load: 78%
   [Lux 2012-Feb-10 10:46:01 INFO : 0] 00:00:36 - 16T: 5.04 S/p  161.62 kS/s  298% Eff  481.79 kC/s - 0% Complete (S/Px) - 02:02:33 Remaining - GPU Load: 80%
   [Lux 2012-Feb-10 10:46:06 INFO : 0] 00:00:41 - 16T: 5.86 S/p  165.39 kS/s  298% Eff  492.78 kC/s - 1% Complete (S/Px) - 01:59:40 Remaining - GPU Load: 82%
   [Lux 2012-Feb-10 10:46:11 INFO : 0] 00:00:46 - 16T: 6.70 S/p  168.67 kS/s  298% Eff  502.44 kC/s - 1% Complete (S/Px) - 01:57:14 Remaining - GPU Load: 83%
   [Lux 2012-Feb-10 10:46:16 INFO : 0] 00:00:51 - 16T: 7.54 S/p  171.16 kS/s  298% Eff  509.91 kC/s - 1% Complete (S/Px) - 01:55:27 Remaining - GPU Load: 84%
   [Lux 2012-Feb-10 10:46:21 INFO : 0] 00:00:56 - 16T: 8.37 S/p  173.29 kS/s  298% Eff  516.38 kC/s - 1% Complete (S/Px) - 01:53:56 Remaining - GPU Load: 85%
   [Lux 2012-Feb-10 10:46:26 INFO : 0] 00:01:01 - 16T: 9.20 S/p  174.86 kS/s  298% Eff  520.96 kC/s - 1% Complete (S/Px) - 01:52:49 Remaining - GPU Load: 85%
   [Lux 2012-Feb-10 10:46:31 INFO : 0] 00:01:06 - 16T: 10.04 S/p  176.43 kS/s  298% Eff  525.37 kC/s - 1% Complete (S/Px) - 01:51:43 Remaining - GPU Load: 86%
   [Lux 2012-Feb-10 10:46:36 INFO : 0] 00:01:11 - 16T: 10.87 S/p  177.65 kS/s  298% Eff  528.79 kC/s - 1% Complete (S/Px) - 01:50:51 Remaining - GPU Load: 87%
   [Lux 2012-Feb-10 10:46:41 INFO : 0] 00:01:16 - 16T: 11.70 S/p  178.77 kS/s  298% Eff  532.00 kC/s - 1% Complete (S/Px) - 01:50:04 Remaining - GPU Load: 87%
   [Lux 2012-Feb-10 10:46:46 INFO : 0] 00:01:21 - 16T: 12.53 S/p  179.65 kS/s  298% Eff  534.56 kC/s - 1% Complete (S/Px) - 01:49:27 Remaining - GPU Load: 88%
   [Lux 2012-Feb-10 10:46:51 INFO : 0] 00:01:26 - 16T: 13.38 S/p  180.64 kS/s  298% Eff  537.47 kC/s - 1% Complete (S/Px) - 01:48:45 Remaining - GPU Load: 88%
   [Lux 2012-Feb-10 10:46:56 INFO : 0] 00:01:31 - 16T: 14.21 S/p  181.39 kS/s  298% Eff  539.69 kC/s - 1% Complete (S/Px) - 01:48:13 Remaining - GPU Load: 88%
   [Lux 2012-Feb-10 10:47:01 INFO : 0] 00:01:36 - 16T: 15.05 S/p  182.11 kS/s  297% Eff  541.76 kC/s - 1% Complete (S/Px) - 01:47:42 Remaining - GPU Load: 89%
   [Lux 2012-Feb-10 10:47:06 INFO : 0] 00:01:41 - 16T: 15.88 S/p  182.76 kS/s  297% Eff  543.57 kC/s - 2% Complete (S/Px) - 01:47:13 Remaining - GPU Load: 89%
   [Lux 2012-Feb-10 10:47:11 INFO : 0] 00:01:46 - 16T: 16.71 S/p  183.27 kS/s  297% Eff  544.95 kC/s - 2% Complete (S/Px) - 01:46:50 Remaining - GPU Load: 89%
   [Lux 2012-Feb-10 10:47:16 INFO : 0] 00:01:51 - 16T: 17.55 S/p  183.82 kS/s  297% Eff  546.46 kC/s - 2% Complete (S/Px) - 01:46:26 Remaining - GPU Load: 89%
   [Lux 2012-Feb-10 10:47:21 INFO : 0] 00:01:56 - 16T: 18.39 S/p  184.35 kS/s  297% Eff  547.94 kC/s - 2% Complete (S/Px) - 01:46:02 Remaining - GPU Load: 90%
   [Lux 2012-Feb-10 10:47:26 INFO : 0] 00:02:01 - 16T: 19.23 S/p  184.80 kS/s  297% Eff  549.22 kC/s - 2% Complete (S/Px) - 01:45:41 Remaining - GPU Load: 90%
   [Lux 2012-Feb-10 10:47:31 INFO : 0] 00:02:06 - 16T: 20.07 S/p  185.23 kS/s  297% Eff  550.44 kC/s - 2% Complete (S/Px) - 01:45:21 Remaining - GPU Load: 90%
   [Lux 2012-Feb-10 10:47:36 INFO : 0] 00:02:11 - 16T: 20.90 S/p  185.53 kS/s  297% Eff  551.26 kC/s - 2% Complete (S/Px) - 01:45:06 Remaining - GPU Load: 90%
   [Lux 2012-Feb-10 10:47:41 INFO : 0] 00:02:16 - 16T: 21.74 S/p  185.92 kS/s  297% Eff  552.34 kC/s - 2% Complete (S/Px) - 01:44:47 Remaining - GPU Load: 90%
   [Lux 2012-Feb-10 10:47:46 INFO : 0] 00:02:21 - 16T: 22.58 S/p  186.26 kS/s  297% Eff  553.26 kC/s - 2% Complete (S/Px) - 01:44:31 Remaining - GPU Load: 91%
   [Lux 2012-Feb-10 10:47:51 INFO : 0] 00:02:26 - 16T: 23.41 S/p  186.53 kS/s  297% Eff  553.97 kC/s - 2% Complete (S/Px) - 01:44:17 Remaining - GPU Load: 91%
   [Lux 2012-Feb-10 10:47:56 INFO : 0] 00:02:31 - 16T: 24.25 S/p  186.87 kS/s  297% Eff  554.98 kC/s - 2% Complete (S/Px) - 01:44:00 Remaining - GPU Load: 91%
   [Lux 2012-Feb-10 10:48:01 INFO : 0] 00:02:36 - 16T: 25.09 S/p  187.15 kS/s  297% Eff  555.77 kC/s - 2% Complete (S/Px) - 01:43:45 Remaining - GPU Load: 91%
   [Lux 2012-Feb-10 10:48:06 INFO : 0] 00:02:41 - 16T: 25.93 S/p  187.43 kS/s  297% Eff  556.56 kC/s - 3% Complete (S/Px) - 01:43:31 Remaining - GPU Load: 91%
Render Magic
----------------
i7 950 - Not OC'd
24G DDR3 RAM
2 GTX 580s
rendermagic
Developer
 
Posts: 141
Joined: Wed Mar 23, 2011 2:32 pm
Location: Leading edge of a photon (California USA)

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby SATtva » Fri Feb 10, 2012 12:50 pm

Samples per pixel is a low-level technical value, while contributions per second is the actual speed of convergence.
Linux builds packager
聞くのは一時の恥、聞かぬのは一生の恥
User avatar
SATtva
Developer
 
Posts: 5487
Joined: Tue Apr 07, 2009 12:19 pm
Location: from Siberia with love

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby rendermagic » Fri Feb 10, 2012 1:10 pm

SATtva wrote:Samples per pixel is a low-level technical value, while contributions per second is the actual speed of convergence.


That's kind of what I thought. So a good test would be to compare to renders based upon 'time' instead of 'samples per pixel', would you agree?

I'll run the same two tests again and compare the two resulting images. Since the OpenCL is reporting that it's contributing / converging 2.5 times more than the the NoOpenCL version, one can assume that if the NoOpenCL were allowed to run 2.5 times longer, the noise level in each would be the same. I'll let you know the results.
Render Magic
----------------
i7 950 - Not OC'd
24G DDR3 RAM
2 GTX 580s
rendermagic
Developer
 
Posts: 141
Joined: Wed Mar 23, 2011 2:32 pm
Location: Leading edge of a photon (California USA)

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby rendermagic » Fri Feb 10, 2012 1:39 pm

I just performed the tests. When I compared the two resulting images, it's clear that my hypotesis stated in my last message post was correct. When you compare the images, you can clearly see that the number of 'sample per pixel' is not accurate to measure render image quality.

It appears that the 'contributions per second' is far more accurate to use a measure of image / bake quality. Indeed, the 'samples per pixel' is a little misleading. Perhaps we would be using a different measuring stick...maybe 'contributions per pixel' :o
Render Magic
----------------
i7 950 - Not OC'd
24G DDR3 RAM
2 GTX 580s
rendermagic
Developer
 
Posts: 141
Joined: Wed Mar 23, 2011 2:32 pm
Location: Leading edge of a photon (California USA)

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby LadeHeria » Fri Feb 10, 2012 3:04 pm

SATtva wrote:Is there anything in the Lux log?
Btw, do you observe this behavior with latest builds only? What about the previous ones?


No, i don"t see anything in the Lux log :-(.
http://www.pasteall.org/pic/show.php?id=26246

My build don't work to :-(.


LadeHeria
LadeHeria
 
Posts: 129
Joined: Fri Aug 27, 2010 8:59 am

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby SATtva » Sat Feb 11, 2012 12:44 am

Still, your GPU is recognized correctly. Have you tried a more real-life scene instead of a default cube? You should note hybrid rendering is meaningful only with heavy scenes -- 100k polygons is the minimum.

Btw, what's your video driver version?
Linux builds packager
聞くのは一時の恥、聞かぬのは一生の恥
User avatar
SATtva
Developer
 
Posts: 5487
Joined: Tue Apr 07, 2009 12:19 pm
Location: from Siberia with love

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby LadeHeria » Sun Feb 12, 2012 10:49 am

I have the 290 driver. I tried with a 500k polygons scene, it doesn't work to.
I just have this error in the terminal.

http://www.pasteall.org/pic/show.php?id=26365

LadeHeria
LadeHeria
 
Posts: 129
Joined: Fri Aug 27, 2010 8:59 am

Re: LuxRender 0.9-dev Linux build of 04-02-2012

Postby Dade » Sun Feb 12, 2012 1:59 pm

LadeHeria wrote:I have got a stange error. OpenCL render don't work with Luxrender but it's work with SLG/Cycles/Luxmark.

Code: Select all
./luxrender: /usr/lib/nvidia-current-updates/libOpenCL.so.1: no version information available (required by ./luxrender)


I just see this error in my terminal. Any idea ?


This isn't really an error, more a warning you can ignore.

What GPU do you have ? This may be related to the increase of default QBVH stack size (http://src.luxrender.net/lux/rev/7a8001ac6714). Can you try to use the "accelerator.qbvh.stacksize.max" set to 24 (the old default value) ?
User avatar
Dade
Developer
 
Posts: 4795
Joined: Sat Apr 19, 2008 6:04 pm
Location: Italy

PreviousNext

Return to Weekly Testing Builds

Who is online

Users browsing this forum: No registered users and 1 guest