I can reconfirm what you are seeing (9400M/9600M GT combo), Chris. I'm attaching an additional version that sees to show a billboard shadow in the precise opposite place on the sphere from where one would expect it to be - light source is x>0, which would make me expect to see the shadow appear on the left side of the sphere. Instead, the shadow appears on the right side, just about exactly opposite of where it should be.
Also, Tom, using a sprite isn't really any kind of workaround for use with GLSL given the limited vertices, and corresponding lack of resolution control; even given that, one can still see it "do" the same error as it makes the topmost part of the rotation, and the bottom right corner will disappear before it should (if you mimic the setting of the GLSL grid as closely as possible, by making the w/h 2 QC Units, and placing it at z=-1. Best regards, George Toledo On Sat, Sep 5, 2009 at 3:30 AM, Tom Beddard <[email protected]> wrote: > In my own experiments with the GLSL grid patch I've been seeing lots of > visual artefacts and quite a few complete system lockups requiring hard > reboot (iMac with nvidia GeForce 8800 GS gfx card). It does seem > intermittent though with the script working fine for a while and then not at > all even without any code changes. > > At first I thought it was some odd issues with my fragment shader but > switching out the GLSL grid patch for a standard sprite seems patch seems to > have sorted my problem for the moment. > I'll post an example when I've investigated it a little further. > > > > -- > Tom Beddard > http://www.subblue.com > > > > > On 5 Sep 2009, at 04:39, Christopher Wright wrote: > > Spent some time with the GLSL shader today, had some annoying problems pop >> up. >> >> This composition demonstrates 3 such problems, described in Root-level >> notes. >> >> <many issues.qtz> >> >> >> Issue 1: The GLSL Grid is incorrectly assumed to always be from -width/2, >> -height/2 to +width/2,+height/2 (z=0). Because of this, it is incorrectly >> not rendered when QC deems it to be off screen. However, Vertex Shaders can >> deform vertices such that the grid _is_ on screen, even when its original >> vertices aren't. >> >> Issue 2: The billboard patch casts an incorrect shadow on the sphere -- >> I'm not entirely sure what's going on there, and will probably need to >> investigate this further. >> >> Issue 3: The lighting patch renders an impostor-shadow where the GLSL >> grid's original (non-deformed) vertices are. However, if the vertices are >> deformed, the shadow is still rendered in space where the grid was. In >> comical cases (i.e. output vertex = input vertex - 2 on the z axis), the >> shadow can even end up being rendered in space. (In this example, the grid >> is rippled with sine waves, yet the shadow remains planar, which is wrong.) >> >> Can anyone confirm/flesh out these (for a more detailed bug report/bug >> report set)? >> >> -- >> [ christopher wright ] >> [email protected] >> http://kineme.net/ >> >> _______________________________________________ >> Do not post admin requests to the list. They will be ignored. >> Quartzcomposer-dev mailing list ([email protected]) >> Help/Unsubscribe/Update your Subscription: >> >> http://lists.apple.com/mailman/options/quartzcomposer-dev/tom%40subblue.com >> >> This email sent to [email protected] >> > > > > > > _______________________________________________ > Do not post admin requests to the list. They will be ignored. > Quartzcomposer-dev mailing list ([email protected]) > Help/Unsubscribe/Update your Subscription: > > http://lists.apple.com/mailman/options/quartzcomposer-dev/gtoledo3%40gmail.com > > This email sent to [email protected] >
many issues 2.qtz
Description: Binary data
_______________________________________________ Do not post admin requests to the list. They will be ignored. Quartzcomposer-dev mailing list ([email protected]) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/quartzcomposer-dev/archive%40mail-archive.com This email sent to [email protected]
