RE: YUV format

>For example, 16 bits -> 4 bits (4:1) reduces 65536 values to 16.
>that's 1/4000th the resolution, not 1/4.

That's probably the stupidest thing I've ever heard you say. "bits" do not 
correspond to "resolution" like that. Reducing resolution by 1/4 already 
reduces four 8-bit values to one, a reduction of 1:16.8mln by your accounting, 
though clearly the amount of "information" being thrown away is not that 
great. Assuming a Haar wavelet transform, a 1:4 resolution reduction is 
exactly equivalent to throwing away the lowest level of the coefficient tree. 
I believe it is commonly accepted that reducing the resolution of images 
before wavelet compression makes no sense.

This is not true for other transforms, in particular the DCT, which is the 
basis of previous video codecs, which is why many formats do reduce the 
chrominance resolution, and why supporting it for compatibility purposes is a 
good idea. The justification is that the human visual system perceives color 
at a lower spatial resolution than luminance. It is just happenstance that the 
reduction of bitrate for wavelets causes increased blurring.

--- >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 '[email protected]'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.