I have a strange problem with my IRTC entry when I turn on radiosity. This
was working OK so it seems to be something I have added to the image.
Problem is the image is nearly 12,000 lines of SDL and I have added a fair
bit since last generating the radiosity data. I am not really sure what has
been added since I last switched on radiosity. I am trying to eliminate
objects and the radiosity settings but have not yet found out what causes
it. Thought I would see if anyone has seen this problem before?

This happens with POV 3.6 and megapov 1.2

My Radiosity settings are

global_settings{
 assumed_gamma 1.0
 ambient_light 0.0
 max_trace_level 15
 #if (use_rad)
 radiosity{
  // save settings: modified Outdoor settings from rad_def.inc
  #if (use_rad=2)
    pretrace_start 0.08
    pretrace_end   1/image_width
    count 200
    nearest_count 20
    error_bound 0.1
    recursion_limit 2
    low_error_factor .9
    gray_threshold 0
    minimum_reuse 0.015
    brightness 1
    adc_bailout 0.01/2
    normal on
    save_file "oldtrain.rad"

  #else
    // load settings
    pretrace_start 1 pretrace_end 1
    load_file "oldtrain.rad"
    error_bound 0.1    // it's important to specify the same value used to
save data
    recursion_limit 2  // also the same if specified
    always_sample off  // no more samples: specially useful with focal blur
    normal on
  #end
 }
 #end
}

For megapov I used:

 radiosity{
  // save settings: modified Outdoor settings from rad_def.inc
  #if (use_rad=2)
    pretrace_start 1
    pretrace_end  PRETRACE_END //1/image_width

    count 200
    nearest_count 10
    //error_bound 0.2
    error_bound { 0.2 adaptive 2.0, 20 }

    low_error_factor .8
    gray_threshold 0
    recursion_limit 2
    brightness 0.8
    adc_bailout 0.01/2
    normal on
    randomize on
    save_file "oldtrain.rad"
  #end
  #if(use_rad=1)
    // load settings
    pretrace_start 1 pretrace_end 1
    load_file "oldtrain.rad"
//    error_bound 0.07    // it's important to specify the same value used
to save data
    error_bound { 0.2 adaptive 2.0, 20 }
    recursion_limit 2  // also the same if specified
    always_sample off  // no more samples: specially useful with focal blur
    normal on
    randomize on
  #end
 }
 #end
}

If I use the default (low level) radiosity settings (when just inserting
radiosity from the menu I don't seem to get the problem.)

Also if I don't render all objects I don't seem to get the problem but I
have not yet found the culprit (either setting or object).

The image to the left has radiosity turned on the one to the right has it
turned off..

Help..

Sean

Post a reply to this message

Attachments:
Download 'radproblem.jpg' (43 KB)

Preview of image 'radproblem.jpg'

OK, I think I have found the problem, I think it is the "normal on" in the
radiosity settings, the problem only seems to show up with fairly high
count values though and I think it is because some of my normals are set
very high. Never had this problem before but I am fairly sure it is to do
with this. I changed to using normals as I was using isosurfaces before but
the render times were too slow for this IRTC round deadline.

Sean

Post a reply to this message

"s.day" <s.d### [at] uelacuk> wrote:
> OK, I think I have found the problem, I think it is the "normal on" in the
> radiosity settings, the problem only seems to show up with fairly high
> count values though and I think it is because some of my normals are set
> very high. Never had this problem before but I am fairly sure it is to do
> with this. I changed to using normals as I was using isosurfaces before but
> the render times were too slow for this IRTC round deadline.
>
> Sean

was experiencing same problem with an outdoor scene. Will try turning off
normal.
Thanks
Joanne

Post a reply to this message

Joanne Simpson nous apporta ses lumieres en ce 27/08/2006 21:46:
> "s.day" <s.d### [at] uelacuk> wrote:
>> OK, I think I have found the problem, I think it is the "normal on" in the
>> radiosity settings, the problem only seems to show up with fairly high
>> count values though and I think it is because some of my normals are set
>> very high. Never had this problem before but I am fairly sure it is to do
>> with this. I changed to using normals as I was using isosurfaces before but
>> the render times were too slow for this IRTC round deadline.
>>
>> Sean
> 
> was experiencing same problem with an outdoor scene. Will try turning off
> normal.
> Thanks
> Joanne
> 
Only set normal on if you REALY need to as it makes radiosity calculations a lot 
more complex, resulting in greatly increased render times.

-- 
Alain
-------------------------------------------------
Don't waste your time on a man/woman, who isn't
willing to waste their time on you.

Post a reply to this message