[FFmpeg-devel] [PATCH v2] lavf/rtpenc_jpeg: Retrieve sampling factor from SOF0
Xiang, Haihao
haihao.xiang at intel.com
Mon Jun 12 05:46:21 EEST 2023
On Wo, 2023-01-11 at 01:35 +0000, Yeoh, Hoong Tee wrote:
> In ff_rtp_send_jpeg, the type is defined based on PIX_FMT and
> color-range parsed in. There is limitation on current design
> where need to include support newly introduced PIX_FMT such as
> AV_PIX_FMT_QSV and there might be more and more in future. Hence,
> retrive the sampling factor from SOF0 in JPEG compressed header
> directly. This introduces flexibility to handle different type of
> new codec introduced in future.
>
> Signed-off-by: Yeoh, Hoong Tee <hoong.tee.yeoh at intel.com>
> ---
> libavformat/rtpenc_jpeg.c | 41 +++++++++++++++++++++++++--------------
> 1 file changed, 26 insertions(+), 15 deletions(-)
>
> diff --git a/libavformat/rtpenc_jpeg.c b/libavformat/rtpenc_jpeg.c
> index 9111683..53be188 100644
> --- a/libavformat/rtpenc_jpeg.c
> +++ b/libavformat/rtpenc_jpeg.c
> @@ -30,7 +30,7 @@ void ff_rtp_send_jpeg(AVFormatContext *s1, const uint8_t
> *buf, int size)
> RTPMuxContext *s = s1->priv_data;
> const uint8_t *qtables[4] = { NULL };
> int nb_qtables = 0;
> - uint8_t type;
> + uint8_t type = 2; /* initialized non-0/1 value for RTP/JPEG type check*/
> uint8_t w, h;
> uint8_t *p;
> int off = 0; /* fragment offset of the current JPEG frame */
> @@ -45,20 +45,6 @@ void ff_rtp_send_jpeg(AVFormatContext *s1, const uint8_t
> *buf, int size)
> w = AV_CEIL_RSHIFT(s1->streams[0]->codecpar->width, 3);
> h = AV_CEIL_RSHIFT(s1->streams[0]->codecpar->height, 3);
>
> - /* get the pixel format type or fail */
> - if (s1->streams[0]->codecpar->format == AV_PIX_FMT_YUVJ422P ||
> - (s1->streams[0]->codecpar->color_range == AVCOL_RANGE_JPEG &&
> - s1->streams[0]->codecpar->format == AV_PIX_FMT_YUV422P)) {
> - type = 0;
> - } else if (s1->streams[0]->codecpar->format == AV_PIX_FMT_YUVJ420P ||
> - (s1->streams[0]->codecpar->color_range == AVCOL_RANGE_JPEG &&
> - s1->streams[0]->codecpar->format == AV_PIX_FMT_YUV420P)) {
> - type = 1;
> - } else {
> - av_log(s1, AV_LOG_ERROR, "Unsupported pixel format\n");
> - return;
> - }
> -
> /* preparse the header for getting some info */
> for (i = 0; i < size; i++) {
> if (buf[i] != 0xff)
> @@ -90,6 +76,23 @@ void ff_rtp_send_jpeg(AVFormatContext *s1, const uint8_t
> *buf, int size)
> "Only 1x1 chroma blocks are supported. Aborted!\n");
> return;
> }
> +
> + /*
> + * Find out the sampling factor in SOF0.
> + * In SOF0, hsample/vsample is inserted in form of (2<<4) | (type
> ? 2 : 1).
> + * First 4-bit is hsample while Last 4-bit is vsample.
> + */
> +
> + /* Luma channel sampling factor in 4:2:2 chroma subsampling are
> 2x1 */
> + if (buf[i + 11] == 33) {
> + type = 0;
> + /* Luma channel sampling factor in 4:2:0 chroma subsampling are
> 2x2 */
> + } else if (buf[i + 11] == 34) {
> + type = 1;
> + } else {
> + av_log(s1, AV_LOG_ERROR, "Unsupported pixel format\n");
> + return;
> + }
> } else if (buf[i + 1] == DHT) {
> int dht_size = AV_RB16(&buf[i + 2]);
> default_huffman_tables |= 1 << 4;
> @@ -163,6 +166,14 @@ void ff_rtp_send_jpeg(AVFormatContext *s1, const uint8_t
> *buf, int size)
> break;
> }
> }
> +
> + /* Check validity of RTP/JPEG type */
> + if (type != 0 && type != 1) {
> + av_log(s1, AV_LOG_ERROR,
> + "Invalid RTP/JPEG type\n");
> + return;
> + }
> +
> if (default_huffman_tables && default_huffman_tables != 31) {
> av_log(s1, AV_LOG_ERROR,
> "RFC 2435 requires standard Huffman tables for jpeg\n");
Sorry to pick up this old thread. This patch looks good and works well for me. I
will merge it if no one objects in a few days.
Thanks
Haihao
More information about the ffmpeg-devel
mailing list