[vorbis-dev] video codec

[Steve Underwood]

Gregory Maxwell wrote:
> 
> On Thu, Feb 08, 2001 at 10:12:43AM -0800, Ralph Giles wrote:
> > On Thu, 8 Feb 2001, Steve Underwood wrote:
> >
> > > mini-DV camcorders are capable of quite excellent results. This may not
> > > be the ideal test material, though. Its just too clean. Many compression
> > > schemes which will do very well on this, may do rather poorly on more
> > > typcial somewhat noisy but uncompressed video. Use a mini-DV handycam in
> > > low light, and make sure there is plenty of noise to really exercise the
> > > codec.
> >
> > One imagines running the source through a noise "filter" would be a better
> > way of exercising this. Likewise with filmic grain, scratches and dust.
> >
> > An unrealated question: what's a good license for any test clips we
> > produce. BSD? Public Domain? "non-commercial use" seems popular elsewhere.
> 
> Actually, that kind of noise (hi-gain on CCD in lowlight, noise 'filter',
> etc), isn't anywhere the kind of issue for wavelet thresholding coders that
> is is for DCT coders so it's perfectly reasonable to use non-noisy test
> sets.
> 
> Since we're looking at 3d transform coders, it would probably be more
> intesting to have footage with segments of smooth motion combined with
> sudden scene changes..

This sounds like focusing the testing on what you expect to happen,
rather than keeping an open mind about unexpected quirky behaviour. Not
a good idea. Nasty realities often bite suprisingly late in the day when
you take that approach.

Regards,
Steve

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.