David Koski schrieb:
> I have found an easier way.  There is a deinterlace effect called "Spacial 
> field swap" that does the trick for the most part.  Still, some sections have
>  jerky motion instead of the jaggies like a field order reversal has.  I 
> suspect the field order is not swapped within a frame but swapped between 
> frames....

so probably your source material contains maybe some sort of synchronisation
error or is skipping single frames?

Maybe I should add that in general, video/media formats are highly complex
beasts, and, still worse, are often implemented "slightly incorrect" by
the vendors of cameras or other media-producing gear. Thus it is common
habit for media decoders to have some built-in tolerance and heuristics
to compensate for some glitches. It figures, that any such automatic
compensation can itself yield other glitches or be unpredictable in
many ways. Usually it's quite difficult to do an actual diagnosis
in such cases.

> so instead of this:
> 
> 1a+1b, 2a+2b, 3a+3b, ...
> 
> ..it is this:
> 
> 0b+1a, 1b+2a, 2b+3a, ...

...which would be completely correct field sequence (because the numbering
with "1" or "2" is here just added for the sake of the argument.

Rather in this case I'd suspect that e.g. what you label here as "1b" actually
was originally created in the camera to be placed in the lower field, but
here does end up here in the upper field?
But, honestly, I don't know enough about the internals of any of these codecs
to tell you if that would even be possible somehow.

Greetings
Hermann V.




_______________________________________________
Cinelerra mailing list
[email protected]
https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra

Reply via email to