Fix standalone compilation of the WebVTT decoder.

[ffmpeg] / libavcodec / h264_mp4toannexb_bsf.c

diff --git a/libavcodec/h264_mp4toannexb_bsf.c b/libavcodec/h264_mp4toannexb_bsf.c
index a7b4f4158ea475ed8efaa08c7d45839d9ae6738a..e0c1385f0794918eff8762e61b1a522c503431ce 100644 (file)
--- a/libavcodec/h264_mp4toannexb_bsf.c
+++ b/libavcodec/h264_mp4toannexb_bsf.c
@@ -176,9 +176,14 @@ static int h264_mp4toannexb_filter(AVBitStreamFilterContext *bsfc,
         if (buf + nal_size > buf_end || nal_size < 0)
             goto fail;
 
-        if (ctx->first_idr && (unit_type == 7 || unit_type == 8))
+        if (unit_type == 7 || unit_type == 8)
             ctx->idr_sps_pps_seen = 1;
 
+        /* if this is a new IDR picture following an IDR picture, reset the idr flag.
+         * Just check first_mb_in_slice to be 0 as this is the simplest solution.
+         * This could be checking idr_pic_id instead, but would complexify the parsing. */
+        if (!ctx->first_idr && unit_type == 5 && (buf[1] & 0x80))
+            ctx->first_idr = 1;
 
         /* prepend only to the first type 5 NAL unit of an IDR picture, if no sps/pps are already present */
         if (ctx->first_idr && unit_type == 5 && !ctx->idr_sps_pps_seen) {