vincent Rioux wrote:
i am not sure whether i fully understand what you mean.
i agree with you, my description is not crystal clear. imagine that there is a video in the background (fullscreen). progressively some small recatngular parts of the video are transformed (for example gets darker) and move - in order to do that i apply another gem layer on top of the video layer and insert several rectangles objects with alpha channel.
thanks. what you describe canNOT be recorded directly with [pix_record] (you already know that).
[pix_record] will _only_ record pixes. (e.g. you have to transform your "rectangles" into pix-domain; just using [rectangle] won't do)
ok, how do one make this kind of transformation?
[pix_snap] transforms the rendering buffer (what has been rendered so far in the render-cycle!) into a pix-buffer.
e.g.
[pix_video] | [pix_gain] | [pix_record] | [pix_invert]
will record a video with [pix_gain] applied, but [pix_invert] is NOT applied.
ok, i think i got this one but what about pix_mix?
[pix_mix] is harder to draw in ascii. apart from that you shoud insert it instead of [pix_gain] to have an effect.
mfg.adsr IOhannes