MLM Veraart (Veraart++at++fel.tno.nl)
Mon, 26 Apr 1999 14:47:37 +0200
The red square you see IS the shadow map that is generated in the first
pass of drawing phase. In the IrisGL performer 2.0 version you can see
the scene as viewed from the lightsource in a dark colour the rest of
the map was red.
With performer 2.2 the shadow map has changed to completely red. You
can get a glimps of what is the contents of the map by moving the
viewpoint such that the depthbuffer is not cleared anymore. Or just
disable the clearing of the depth buffer. Then playing with the
viewpoint you can force parts of the shadow map picture behind your
groundplane (or scene) and get a contour of what the light source is
getting from the scene. You can use this technique to see if you
effectively use the shadow map. Is the shadow frustum tight enough?
You might try to see if you are clearing the frame buffer by
means of an EarthSky or any other method.
Whilst on the subject of the red square.
If you combine stereo.c and shadow.c you will see the shadow map
persistent in one of the eye views. I think this is a bug.
If you disable the continous generation of shadowmaps
lsource.setMode(PFLS_FREEZE_SHADOWS, PF_ON) the red square disapears.
Mario Veraart
>
> >> 3. The project I'm working on has two main parts (almost totally
> >> independent), the computation of all the different parts of the
> >> simulation (dynamics, GUI, ...) & the visual renderer (pf based)
> >> representing visually the state of the simulation. We are able to
> >> run the simulation with or without the visuals.
> >> When I load my simulation with the visuals for the first time,
> >> everything is OK (the shadows are casted correctly on all my
> >> textured objects) but if I unload the visuals (only the visuals,
> >> the other part keeps running) and reload it some minutes after, the
> >> shadows will only appear on some textures & won`t appear anymore on
> >> other textures (none of my objects or lights have moved! so the
> >> scene should be the same!!)
> >> It is really strange because a shadow can completly cover an object
> >> before I kill my visuals, but when I re-run my visuals the same
> >> shadow appears on parts of the object and doesn`t appear on other
> >> parts of the same object (the only difference between the parts is
> >> the texture).
> >> HINT: If I wait about 30-40 minutes before I load my visuals again
> >> everything is OK.
>
> I am using shared memory (for synchronous data) to interface my pf app
> and the rest of the simulation: objects positions (bodies, lights,
> ...) are computed and then the shared memory is updated in real-time
> that way my pf app can read into the shared memory and update the
> visual with the actual data. I also send messages through messages
> queues (asynchronous data)
> I don`t know if my problem is due to a second initialisation of the
> pipes (the second time I load my pf app) ,maybe something stays
> somewhere in memory (???), I don`t know.
> It always appear on the same textures! (is it possible that the .pfb
> files containing my visual database are not properly loaded the second
> time (???) ...)
>
> Thank you, please keep sending your comments or suggestions,
> I will really appreciate ANY help
>
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
> -----------------------------------------------------------------------
> List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Mon Apr 26 1999 - 05:46:57 PDT