Fixrobert

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.

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.

movie-ratio-adaptation-pixelated-2

the technique i worked out for mixing videos of differing aspect ratios is to simply overlay 2 (or however many channels you wish) jit.gl.videoplanes. you can composite using simply blend_enable and selecting your blendmode. i use slabs on the video inputs to control crossfading, etc.

this also allows you to easily change the aspect ratio of the output window, and have the videoplanes scale appropriately.

take a look at the example below, and let me know if anything is not clear.

I have allready experiment that solution (the fade can be done with the alpha value of the color argument of the videoplane too) and it solves the background issue (just have to apply the background color to the erase_color argument of renderer).

But I was looking for a solution that use a jit.gl.slab to operate the mix to eventually offer alternative mixing process (like wipe for exemple) via a menu (in the video processing system way).

And I still need to know why the scrdim / dstdim solution operate a pixelation (a loss of definition !).

Movie ratio adaptation pixelated