|
|
Posted by davidrobinson@postmaster.co.uk on 10/10/06 12:18
Roderick Stewart wrote:
> Filtering the RGB outputs of the decoders must be tricky, as the
> subcarriers can vary over a range of frequencies. You can use a notch
> filter in PAL and NTSC, and only see spurious components outside its
> range on transitions between coloured areas, but with SECAM, the more
> brightly coloured areas will have subcarrier of various frequencies
> (depending on colour) all over them. To remove this, a bandpass filter
> must be used, which will have a worse effect on luminance.
It's not that clear cut. Even an amplitude modulated signal (e.g. PAL
chroma) contains frequency components other than the subcarrier
frequency. You can't get clean luma from PAL with a simple notch filter
- the spectra is actually very complex.
You can see the effect of a simple notch filter in this paper from
S&W...
http://www.snellwilcox.com/community/knowledge_center/engineering_guides/compositedecoder_wp.pdf
The section on composite decoding in here is good...
http://www.snellwilcox.com/community/knowledge_center/engineering_guides/estandard.pdf
In fact the whole of S&W's knowledge centre is great!
http://www.snellwilcox.com/community/knowledge_center/engineering_guides/
Sorry this has nothing to do with SECAM. The best link (pembers) has
already been given at the top of the thread!
Cheers,
David.
Navigation:
[Reply to this message]
|