 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hello,
I have a problem with max_trace_level in 3.6.1
I'm trying to render some glass superellipsoids (thanks Tek) for my
2-monitor background. Having set max_trace_level to 128, 3.6.1 goes only
up to 66 but has black areas in the image. With 3.5.0 this is not the
case. Whats causing this behaviour? (Images in p.b.i)
Thanks for your help, Manuel
#########################################################################
Technical stuff:
3.6.1 - Versions tried:
self compiled vanilla 3.6.1 on Linux
official 3.6.1.icl8.win32 on Windows (same behaviour - render cancelled)
3.5.0 - Version:
unofficial Debian compiled 3.5.0c-10
commandline: +h1024 +w2560 +j0 +a0.2
Statistics:
3.6.1
Render Statistics
Image Resolution 2580 x 1024
----------------------------------------------------------------------------
Pixels: 2644500 Samples: 3483532 Smpls/Pxl: 1.32
Rays: 17380972 Saved: 3952513 Max Level: 66/128
----------------------------------------------------------------------------
Ray->Shape Intersection Tests Succeeded Percentage
----------------------------------------------------------------------------
CSG Intersection 24819711 16163129 65.12
Plane 29788868 9621188 32.30
Superellipsoid 49639422 23849714 48.05
Bounding Box 84535705 35236525 41.68
Light Buffer 135580214 60704161 44.77
Vista Buffer 14370337 8982801 62.51
----------------------------------------------------------------------------
Calls to Noise: 110332128 Calls to DNoise: 10
----------------------------------------------------------------------------
Shadow Ray Tests: 16360409 Succeeded: 8783399
Reflected Rays: 7375767 Total Internal: 1127457
Refracted Rays: 6248292
Transmitted Rays: 18
----------------------------------------------------------------------------
Number of photons shot: 273363
Surface photons stored: 1020643
Priority queue insert: 940536910
Priority queue remove: 830910106
Gather function called: 10731591
----------------------------------------------------------------------------
Smallest Alloc: 9 bytes
Largest Alloc: 10567688 bytes
Peak memory used: 32120701 bytes
Total Scene Processing Times
Parse Time: 0 hours 0 minutes 1 seconds (1 seconds)
Photon Time: 0 hours 0 minutes 50 seconds (50 seconds)
Render Time: 0 hours 32 minutes 38 seconds (1958 seconds)
Total Time: 0 hours 33 minutes 29 seconds (2009 seconds)
3.5.0
Statistics for glass.pov, Resolution 2580 x 1024
----------------------------------------------------------------------------
Pixels: 2644500 Samples: 3437940 Smpls/Pxl: 1.30
Rays: 27891991 Saved: 4936000 Max Level: 128/128
----------------------------------------------------------------------------
Ray->Shape Intersection Tests Succeeded Percentage
----------------------------------------------------------------------------
CSG Intersection 45222266 30772787 68.05
Plane 50178290 14187631 28.27
Superellipsoid 90444532 43215989 47.78
Bounding Box 162193769 66499739 41.00
Light Buffer 232117316 103676853 44.67
Vista Buffer 14217401 8874457 62.42
----------------------------------------------------------------------------
Calls to Noise: 125302464 Calls to DNoise: 10
----------------------------------------------------------------------------
Shadow Ray Tests: 27223387 Succeeded: 15209464
Reflected Rays: 15557468 Total Internal: 6934248
Refracted Rays: 8623166
Transmitted Rays: 54
Number of photons shot: 273363
Surface photons stored: 1672297
Priority queue insert: 3269277794
Priority queue remove: 1490616654
Gather function called: 18885402
----------------------------------------------------------------------------
Smallest Alloc: 10 bytes Largest: 327688
Peak memory used: 34788016 bytes
----------------------------------------------------------------------------
Time For Photon: 0 hours 2 minutes 30.0 seconds (150 seconds)
Time For Trace: 1 hours 56 minutes 0.0 seconds (6960 seconds)
Total Time: 1 hours 58 minutes 30.0 seconds (7110 seconds)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> I have a problem with max_trace_level in 3.6.1
> I'm trying to render some glass superellipsoids (thanks Tek) for my
> 2-monitor background. Having set max_trace_level to 128, 3.6.1 goes only
> up to 66 but has black areas in the image. With 3.5.0 this is not the
> case. Whats causing this behaviour? (Images in p.b.i)
Hmm, IIRC, some adc_bailout bug was fixed in 3.6; does setting adc_bailout
to 0 fix the problem?
- Slime
[ http://www.slimeland.com/ ]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Slime wrote:
>
> Hmm, IIRC, some adc_bailout bug was fixed in 3.6; does setting adc_bailout
> to 0 fix the problem?
not really. lowering photon spacing to 0.1 brings photon time
for 3.5 to 18s, for 3.6.1 to 6s.
Adding adc_bailout 0 increases photon time by factors for
both versions (canceled after 5 minutes each). I'll try a
3.6.1 render with that settings, but don't know how long
it will take. (So this is not a solution....)
Manuel
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Adding adc_bailout 0 increases photon time
Note I was talking about the *scene* adc_bailout, not the *photon*
adc_bailout. That is,
global_settings {
adc_bailout 0
}
(It sounded from your response that you thought I was referring to
photons{adc_bailout 0}.)
- Slime
[ http://www.slimeland.com/ ]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Slime wrote:
>>Adding adc_bailout 0 increases photon time
>
>
> Note I was talking about the *scene* adc_bailout, not the *photon*
> adc_bailout. That is,
>
> global_settings {
> adc_bailout 0
> }
Yes, that's what I've done. I did a fast test to see if it helps the
problem. (Smaller resolution, rendering only part of the image and
decreased photon spacing). But adding global adc_bailout 0 makes photon
time go up (btw. I canceled the test after 2 1/2 hour photon time,
compared to 6sec without adc_bailout 0).
Manuel
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Manuel Kasten nous apporta ses lumieres en ce 2005-07-06 05:28:
> Slime wrote:
>
>>> Adding adc_bailout 0 increases photon time
>>
>>
>>
>> Note I was talking about the *scene* adc_bailout, not the *photon*
>> adc_bailout. That is,
>>
>> global_settings {
>> adc_bailout 0
>> }
>
>
> Yes, that's what I've done. I did a fast test to see if it helps the
> problem. (Smaller resolution, rendering only part of the image and
> decreased photon spacing). But adding global adc_bailout 0 makes photon
> time go up (btw. I canceled the test after 2 1/2 hour photon time,
> compared to 6sec without adc_bailout 0).
>
> Manuel
Photons will use the global_settings value for adc_bailout unless another one is set
in the photons
block. The same is true for the max_trace_level.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
