[FFmpeg-user] Chroma subsampling formats

[Mark Filipak]

On 15/07/2025 14.51, Paul B Mahol wrote:
> On Tue, Jul 15, 2025 at 6:00 PM Mark Filipak <markfilipak.imdb at gmail.com>
> wrote:
> 
>> Are these correct? Or, is there some official reference I can consult?
>>
> 
> Maybe correct for packed formats only.

Thanks for replying, Paul.

Ya know, I've tried to read H.262 & H.264 about pixel planes and it all seems so vague to me that I 
finally give up looking at words that just lack meaningful content. Maybe the answers were there and 
I just didn't see them but I get along much better with pictures of objects than with words about 
objects. Specifically, where are those planes? Perhaps "plane" is a metaphor for something. I just 
don't know.

>> Reference: https://en.wikipedia.org/wiki/Chroma_subsampling
> 
> Isn't this very obvious and straightforward read?

Not for me.

• J: horizontal sampling reference (width of the conceptual region). Usually, 4.
• a: number of chrominance samples (Cr, Cb) in the first row of J pixels.
• b: number of changes of chrominance samples (Cr, Cb) between first and second row of J pixels. b 
is usually either zero or equal to a (except in rare irregular cases like 4:4:1 and 4:2:1, which do 
not follow this convention).

The explanation of 'b' is vague. I think it's wrong -- the wrong 'method'. I think the authors made 
it up -- "number of changes" -- because it somehow makes sense to them. But "irregular cases like 
4:4:1 and 4:2:1" doesn't make sense, does it?

I think 'b' is simply the number of chroma samples in the lines of the second field, or it's '0' if 
the second field isn't stored because it's not needed because it will be copied from the first 
field. I don't think it has anything to do with "changes". The test is: Are my 4:4:1 and 4:2:1 
(below) correct?

 >>            Y element, pixels 0..3
 >>           /    Cb element, pixels 0..3
 >>          /    /    Cr element, pixels 0..3
 >>         /    /    /
 >> 4:4:1  ¤¤¤¤ ¤¤¤¤ ¤¤¤¤ —— lines in field 0
 >>        ¤¤¤¤ ¤••• ¤••• —— lines in field 1
 >>
 >> 4:2:1  ¤¤¤¤ ¤•¤• ¤•¤•
 >>        ¤¤¤¤ ¤••• ¤•••

> Also you forgot that chroma, when subsampled, can have different positions,
> top, top-left, center, ..

No. I didn't forget.

--Mark.