I've been searching and searching again but there's definitely something I misanderstand about movie ratio with Jitter.
I want to mix different video sources (movie, camera, colored background) with different ratio,
and make every sources appear with full definition, maximum size and original ratio in full screen.
when displaying only one movie via a jit.gl.videoplane, I've got a nice mathematical process (thank's E. Jourdain at Ircam) witch is quite similar than Vade's one to tune the scale argument of videoplane with screen ratio and original video ratio, but of course, it can't work when mixing two sources whith different ratio...
I've been trying to play with srcdim and dstdim of a matrix inserted between jit.qt.movie and the jit.gl.slab who mixes the two sources.
it works for ratio but displays movies with the wrong definition, (got a pixelation effect - witch disappears when matrix is in adapt 1, but it kills the ratio adaptation solution).
I realy don't understand why a HD movie (1920x1080) appears pixelated in a 1400x900 screen when using source and destination dimension to display it with its own ratio.
second problem : when displaying a video source with its original ratio, we've got those two horizontal (or vertical) black bands.
using the yuyv conversion, they are green. but I want them in the same color than the background I've created with a swatch objetct - so I can mix the two sources softly.
once again, with only one movie to display, it's easy to appli that particular background color to the erase_color attribute of the jit.gl.render.
but, when mixing two different sources, it must be applied before the jit.gl.slab who mixes the sources.