On 1/14/24 6:32 AM, Stefano Sabatini wrote:
> On date Monday 2024-01-08 09:53:37 -0500, Jeffrey Knockel wrote:
>> Previously the description was partially mistaken, explaining the format
>> as RGB 3:3:2, (msb)2R 3G 3B(lsb). While the RGB 3:3:2 part is correct,
>> the latter part should be: (msb)3R 3G 2B(lsb). The corresponding bit
>> masks are red: 0xE0, green: 0x1C, blue: 0x03.
>>
>> Signed-off-by: Jeffrey Knockel <j...@jeffreyknockel.com>
>> ---
>> libavutil/pixfmt.h | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/libavutil/pixfmt.h b/libavutil/pixfmt.h
>> index 58f9ad28bd..9c87571f49 100644
>> --- a/libavutil/pixfmt.h
>> +++ b/libavutil/pixfmt.h
>> @@ -83,7 +83,7 @@ enum AVPixelFormat {
>> AV_PIX_FMT_BGR8, ///< packed RGB 3:3:2, 8bpp, (msb)2B 3G 3R(lsb)
>> AV_PIX_FMT_BGR4, ///< packed RGB 1:2:1 bitstream, 4bpp, (msb)1B
>> 2G 1R(lsb), a byte contains two pixels, the first pixel in the byte is the
>> one composed by the 4 msb bits
>> AV_PIX_FMT_BGR4_BYTE, ///< packed RGB 1:2:1, 8bpp, (msb)1B 2G 1R(lsb)
>
>> - AV_PIX_FMT_RGB8, ///< packed RGB 3:3:2, 8bpp, (msb)2R 3G 3B(lsb)
>> + AV_PIX_FMT_RGB8, ///< packed RGB 3:3:2, 8bpp, (msb)3R 3G 2B(lsb)
>
> This is not consistent with pixdesc definition:
> [AV_PIX_FMT_RGB8] = {
> .name = "rgb8",
> .nb_components = 3,
> .log2_chroma_w = 0,
> .log2_chroma_h = 0,
> .comp = {
> { 0, 1, 0, 6, 2 }, /* R */
> { 0, 1, 0, 3, 3 }, /* G */
> { 0, 1, 0, 0, 3 }, /* B */
> },
> .flags = AV_PIX_FMT_FLAG_RGB,
> },
>
> so either the documentation or the pixdesc description is wrong.
>
> Do you confirm "(msb)3R 3G 2B(lsb)" is the correct definition?
Yes, I created three solid-colored RGBA raw samples filled with either
#ff0000ff, #00ff00ff, or #0000ffff, and converted them using `ffmpeg ...
-pix_fmt rgba ... -pix_fmt rgb8 ...` and inspected the output. I also tested
with `ffmpeg ... -pix_fmt rgba ... -pix_fmt yuv420p ...` followed by `ffmpeg
... -pix_fmt yuv420p ... -pix_fmt rgb8 ...`. In these tests the outputs were
consistent with "(msb)3R 3G 2B(lsb)".
> In this case pixdesc should be fixed accordingly.
Got it, will send an updated patch!
Thanks,
Jeff
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
