Is the YCbCr colour space still used?

Yes, YCbCr is still used everywhere in end-user oriented video and image formats except lossless image compression modes like in TIFF, PNG, WebP lossless, and JPEG-XL lossless.

All lossy video formats including H.264 (AVC), H.265 (HEVC), H.266 (VVC), AV1 and VP9, use YCbCr color space with subsampling 4:2:0 or 4:2:2. VP8 supports only 4:2:0, and 4:2:2 support in H.264 is problematic due to most decoders not having support for the revisions.

The lossless modes of H.264, H.265 and H.266 also use YCbCr but without chroma subsampling (4:4:4). Same with AV1's and VP9's lossless modes.

The image formats JPEG, AVIF, HEIF / HEIC, and JPEG-XL (lossy mode) all use YCbCr with no subsampling (4:4:4) or 4:2:2 or 4:2:0 subsampling. WebP supports only 4:2:0 since it's based on VP8.

AVIF is a container format that uses the AV1 image file format (optimized for images compared to AV1). It supports video / AV1 image file sequences, but only i-frames.

HEIF is a container format that supports AV1 image file format, HEVC, VVC, AVC, JPEG, JPEG2000, AV1 lossless, HEVC lossless, and various raw formats. Although it's mainly intended for images, it also supports video (including p-frames - limited inter-prediction).

WilyKit said:

Yes, YCbCr is still used everywhere in end-user oriented video and image formats except lossless image compression modes like in TIFF, PNG, WebP lossless, and JPEG-XL lossless.

All lossy video formats including H.264 (AVC), H.265 (HEVC), H.266 (VVC), AV1 and VP9, use YCbCr color space with subsampling 4:2:0 or 4:2:2. VP8 supports only 4:2:0, and 4:2:2 support in H.264 is problematic due to most decoders not having support for the revisions.

The lossless modes of H.264, H.265 and H.266 also use YCbCr but without chroma subsampling (4:4:4). Same with AV1's and VP9's lossless modes.

The image formats JPEG, AVIF, HEIF / HEIC, and JPEG-XL (lossy mode) all use YCbCr with no subsampling (4:4:4) or 4:2:2 or 4:2:0 subsampling. WebP supports only 4:2:0 since it's based on VP8.

AVIF is a container format that uses the AV1 image file format (optimized for images compared to AV1). It supports video / AV1 image file sequences, but only i-frames.

HEIF is a container format that supports AV1 image file format, HEVC, VVC, AVC, JPEG, JPEG2000, AV1 lossless, HEVC lossless, and various raw formats. Although it's mainly intended for images, it also supports video (including p-frames - limited inter-prediction).

Click to expand...

Oh, WilyKit thank you very much for such this exhaustive explanation. Yes, I am seeing this colour subsampling in the spec of all modern cameras, so I intuitively knew that YCbCr is still there. Bit still I asked . O.K. so then all the TV receivers and PC' video cards have the decoders, which convert those 3 YCbCr (or only 2 in the case of lossless modes of H.264, H.265 and H.26 ) signals (as time functions) into a stream of bitmap data to feed the individual LCD/LED element to generate a stream of images.

WilyKit, while I have you "on wire" (and as you are definitely an EXPERT), can you please explain another problem with the interlaced video. I have an old Sony camcorder recording FHD video at 60i. When I watch my videos on TV, even modern OLED one, there are no artifacts present, as if they worked like the old TVs showing two interlaced half-frames. But watching in on PC causes problems. Those video players (VLC, Home Media) and mainly the video cards apparently can't work with the half-frames and their de-interlacing methods are pretty poor. Is there some solution (for example a special video card, similar to what TVs use) how to watch such 60i video on PC? And I really appreciate your reply!

--
Regards,
Peter

flektogon said:

WilyKit said:

Yes, YCbCr is still used everywhere in end-user oriented video and image formats except lossless image compression modes like in TIFF, PNG, WebP lossless, and JPEG-XL lossless.

All lossy video formats including H.264 (AVC), H.265 (HEVC), H.266 (VVC), AV1 and VP9, use YCbCr color space with subsampling 4:2:0 or 4:2:2. VP8 supports only 4:2:0, and 4:2:2 support in H.264 is problematic due to most decoders not having support for the revisions.

The lossless modes of H.264, H.265 and H.266 also use YCbCr but without chroma subsampling (4:4:4). Same with AV1's and VP9's lossless modes.

The image formats JPEG, AVIF, HEIF / HEIC, and JPEG-XL (lossy mode) all use YCbCr with no subsampling (4:4:4) or 4:2:2 or 4:2:0 subsampling. WebP supports only 4:2:0 since it's based on VP8.

AVIF is a container format that uses the AV1 image file format (optimized for images compared to AV1). It supports video / AV1 image file sequences, but only i-frames.

HEIF is a container format that supports AV1 image file format, HEVC, VVC, AVC, JPEG, JPEG2000, AV1 lossless, HEVC lossless, and various raw formats. Although it's mainly intended for images, it also supports video (including p-frames - limited inter-prediction).

Click to expand...

Oh, WilyKit thank you very much for such this exhaustive explanation. Yes, I am seeing this colour subsampling in the spec of all modern cameras, so I intuitively knew that YCbCr is still there. Bit still I asked . O.K. so then all the TV receivers and PC' video cards have the decoders, which convert those 3 YCbCr (or only 2 in the case of lossless modes of H.264, H.265 and H.26 ) signals (as time functions) into a stream of bitmap data to feed the individual LCD/LED element to generate a stream of images.

WilyKit, while I have you "on wire" (and as you are definitely an EXPERT), can you please explain another problem with the interlaced video. I have an old Sony camcorder recording FHD video at 60i. When I watch my videos on TV, even modern OLED one, there are no artifacts present, as if they worked like the old TVs showing two interlaced half-frames. But watching in on PC causes problems. Those video players (VLC, Home Media) and mainly the video cards apparently can't work with the half-frames and their de-interlacing methods are pretty poor. Is there some solution (for example a special video card, similar to what TVs use) how to watch such 60i video on PC? And I really appreciate your reply!

Click to expand...

No problem. TVs by Sony, LG, Samsung, Panasonic etc. indeed have advanced deinterlacing algorithms which are usually much better than the deinterlacers used in video player software for computers, smartphones and tablets. Of course that's not surprising, because deinterlacing quality has always been crucial for TVs, so serious developments were made for this. What's very disappointing is that none of the major GPU manufacturers, including NVIDIA, AMD, Intel, Qualcomm, and Apple, have integrated high quality deinterlacing processes in their GPUs. They're all too basic, with poor motion analysis if any.

In VLC you can use "yadif (2x)" which avoids blending frames and thus reducing the frame rate, and the quality can be similar sometimes, but often it can be much worse, depending on the interlacing process. And interlacing is horrible in general, it's a terrifying rabbit hole to get into, especially since there are "mixed" interlacing types.

In order to get better deinterlacing quality in software, better motion-adaptive deinterlacing is needed, but there's very little support for this in media players. TDeint (an avisynth script) can do high quality deinterlacing using 4 or 5 field motion checks, especially with the "slow" mode which has better accuracy. QTGMC is widely regarded as the best software deinterlacer (although TDeint is better in some cases) and it uses a wide range of processes to achieve "smooth" results, but it's also based on avisynth and I'm not sure any media player supports it. Basically you'd have to transcode a video with this deinterlacing already applied.

Bottom line: sadly, interlacing is hell , always has been and always will be, so whenever possible, it's best to avoid it. In your case of course that means using a newer camcorder.