[mythtv-users] MonoChrome (B+W) Recordings

[Brian Wood]

On May 3, 2006, at 8:21 AM, Dan Seddon wrote:

> Hi Brian
> Nice idea - but I think this kind of thing is already fairly well  
> dealt
> with in the video compression. As far as I'm aware (and I'm no expert,
> but I think I'm right on this) video is stored in "yuv" format where
> "y" is the luma and u&v are kinds of difference channels for 2 of  
> the 3
> colours (where the 3rd can be worked out from the first two -  
> giving the
> full chroma information). In 4:2:2 video, the chroma is stored at half
> the resolution of the luma, partly because the eye is less  
> sensitive to
> it and colour does tend to vary less than luminance - so adding colour
> already only doubles the amount of data, rather than triple it.  
> Then in
> digital video this information is compressed which may mean  
> information
> which is repeated is saved in "short hand" or information which is  
> less
> significant is thrown away. In the first case, the chroma information
> tends to compress better than the luma as is varies less (a red apple
> may only have a small variation of red in it, but a lot of tonal
> changes) - which means the chroma might only make up a small  
> fraction of
> the data in the first place, but in the case of BW this would  
> obviously
> be shrunk even further as you'd hope these channels would be  
> practically
> blank.

Good description of video compression, and the reason B+W doesn't  
take up the space it does in analog video, where, if nothing else,  
you are transmitting a color burst on every line.

FCC rules used to (and may still) require stations to kill the color  
burst during B+W programming, to prevent "colored" snow among other  
reasons, the "color killer" in the set will turn off the chroma  
circuits if the burst falls below a certain threshold (a common  
problem was forgetting to turn it on again for commercial breaks that  
were in color).

What you say about compression is true enough for perfect or very  
high quality video, but if the video contains any significant noise  
this will be interpreted at least partly as color information by the  
compression algorithm. What I'd like to see is a way to tell the  
compressor/encoder that there is no intentional color information in  
the signal, and to not waste any bandwidth in trying to transmit any.  
It might not be a whole lot of saved space, but I'd guess that for  
video captured from analog cable of off-air it might be as much as  
5%, certainly worth saving if it was a simple software switch to  
implement it.

Thanks for the interest.