Chrominance subsampling for MPEG-2 interlaced encodings

[Cary Knoop]

Interlaced MPEG-2 allows the chrominance samples to be aligned temporally to the lines or, less common, to the first field of the frame.

Is there an easy way to determine which method was used during encoding?

[manolito]

Are you referring to field based encoding of interlaced sources? The only encoder which supports this AFAIK is Procoder / Carbon Coder. This option complies with the MPEG2 specs, but not with the DVD specs, so I never used it.

From the ECLPro readme for DVD Rebuilder:
Alsaysframe = 0:Use field base encoding for interlaced sources (better) - 1(default):Always use frame mode (more compatible) - Please use frame mode in DVD-RB else you get errors in rebuilding.

So I do not have an answer to your question how you can determine which mode was used during encoding.

[kolak]

Interlaced MPEG-2 allows the chrominance samples to be aligned temporally to the lines or, less common, to the first field of the frame.

Is there an easy way to determine which method was used during encoding?

Maybe by converting it (eg. upscaling) and forcing to use different chroma sampling position. Wrong ones may introduce chroma ghosting. Not sure. Maybe this info is also somewhere in mpeg2 headers.

[Emulgator]

temporally aligned ? I thought spatially aligned, and both fields coded as one frame,
so chroma will be temporally smeared.
Horizontal axis: MPEG-2: aligned on even columns, (col 0,2,4 = even),
(MPEG-1: would be aligned in-between neigbored columns 0,1 and 2,3 etc.
Vertical axis: MPEG-1 and MPEG-2: aligned midways in-between two lines.

http://www.cs.columbia.edu/~delbert/docs/Dueck%20--%20MPEG-2%20Video%20Transcoding.pdf
https://poynton.ca/PDFs/Chroma_subsampling_notation.pdf
https://pdfs.semanticscholar.org/9ae4/4ca7eb8be9b9bdf3ed54f8e1b5874539ff3d.pdf


http://dvd.sourceforge.net/dvdinfo/mpeghdrs.html
Extension header.

Sequence_Extension Byte5 Bit1, Bit2
chroma_format

GoTo H.262 Spec:

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=6&cad=rja&uact=8&ved=2ahUKEwjNwMuauuPlAhXJLVAKHfLoD68QFjAFegQIBhAC&url=https%3A%2F%2Fwww.itu.int%2Frec%2Fdologin_pub.asp%3Flang%3De%26id%3DT-REC-H.262-200002-S!!PDF-E%26type%3Ditems&usg=AOvVaw2JT9nhbWA8uq_RQBzRwyCs

page 54/220 Table 6-5 chroma_format

00 = Reserved
01 = 4:2:0
10 = 4:2:2
11 = 4:4:4

Ha, found something more:
Picture_Coding_Extension Byte7 Bit0
chroma_420_type

page 63/220

"chroma_420_type – If chroma_format is "4:2:0", the value of chroma_420_type shall be the same as progressive_frame; else chroma_420_type has no meaning and shall be equal to zero."