* Restructuring the division of labor between DV codec and DV format

    [ Based on a patch by Brian Brice (bbrice at newtek dot com) ]

Originally committed as revision 6161 to svn://svn.ffmpeg.org/ffmpeg/trunk

7 files changed, 262 insertions, 285 deletions

diff --git a/libavcodec/dv.c b/libavcodec/dv.c
index 90d101813b..fa2952dc65 100644
--- a/libavcodec/dv.c
+++ b/libavcodec/dv.c
@@ -1069,6 +1069,130 @@ static int dvvideo_decode_frame(AVCodecContext *avctx,
     return s->sys->frame_size;
 }
 
+
+static inline int dv_write_pack(enum dv_pack_type pack_id, DVVideoContext *c, uint8_t* buf)
+{
+    /*
+     * Here's what SMPTE314M says about these two:
+     *    (page 6) APTn, AP1n, AP2n, AP3n: These data shall be identical
+     *             as track application IDs (APTn = 001, AP1n =
+     *             001, AP2n = 001, AP3n = 001), if the source signal
+     *             comes from a digital VCR. If the signal source is
+     *             unknown, all bits for these data shall be set to 1.
+     *    (page 12) STYPE: STYPE defines a signal type of video signal
+     *                     00000b = 4:1:1 compression
+     *                     00100b = 4:2:2 compression
+     *                     XXXXXX = Reserved
+     * Now, I've got two problems with these statements:
+     *   1. it looks like APT == 111b should be a safe bet, but it isn't.
+     *      It seems that for PAL as defined in IEC 61834 we have to set
+     *      APT to 000 and for SMPTE314M to 001.
+     *   2. It is not at all clear what STYPE is used for 4:2:0 PAL
+     *      compression scheme (if any).
+     */
+    int apt = (c->sys->pix_fmt == PIX_FMT_YUV420P ? 0 : 1);
+    int stype = (c->sys->pix_fmt == PIX_FMT_YUV422P ? 4 : 0);
+
+    uint8_t aspect = 0;
+    if((int)(av_q2d(c->avctx->sample_aspect_ratio) * c->avctx->width / c->avctx->height * 10) == 17) /* 16:9 */
+        aspect = 0x02;
+
+    buf[0] = (uint8_t)pack_id;
+    switch (pack_id) {
+    case dv_header525: /* I can't imagine why these two weren't defined as real */
+    case dv_header625: /* packs in SMPTE314M -- they definitely look like ones */
+          buf[1] = 0xf8 |               /* reserved -- always 1 */
+                   (apt & 0x07);        /* APT: Track application ID */
+          buf[2] = (0 << 7)    | /* TF1: audio data is 0 - valid; 1 - invalid */
+                   (0x0f << 3) | /* reserved -- always 1 */
+                   (apt & 0x07); /* AP1: Audio application ID */
+          buf[3] = (0 << 7)    | /* TF2: video data is 0 - valid; 1 - invalid */
+                   (0x0f << 3) | /* reserved -- always 1 */
+                   (apt & 0x07); /* AP2: Video application ID */
+          buf[4] = (0 << 7)    | /* TF3: subcode(SSYB) is 0 - valid; 1 - invalid */
+                   (0x0f << 3) | /* reserved -- always 1 */
+                   (apt & 0x07); /* AP3: Subcode application ID */
+          break;
+    case dv_video_source:
+          buf[1] = 0xff; /* reserved -- always 1 */
+          buf[2] = (1 << 7) | /* B/W: 0 - b/w, 1 - color */
+                   (1 << 6) | /* following CLF is valid - 0, invalid - 1 */
+                   (3 << 4) | /* CLF: color frames id (see ITU-R BT.470-4) */
+                   0xf; /* reserved -- always 1 */
+          buf[3] = (3 << 6) | /* reserved -- always 1 */
+                   (c->sys->dsf << 5) | /*  system: 60fields/50fields */
+                   stype; /* signal type video compression */
+          buf[4] = 0xff; /* VISC: 0xff -- no information */
+          break;
+    case dv_video_control:
+          buf[1] = (0 << 6) | /* Copy generation management (CGMS) 0 -- free */
+                   0x3f; /* reserved -- always 1 */
+          buf[2] = 0xc8 | /* reserved -- always b11001xxx */
+                   aspect;
+          buf[3] = (1 << 7) | /* Frame/field flag 1 -- frame, 0 -- field */
+                   (1 << 6) | /* First/second field flag 0 -- field 2, 1 -- field 1 */
+                   (1 << 5) | /* Frame change flag 0 -- same picture as before, 1 -- different */
+                   (1 << 4) | /* 1 - interlaced, 0 - noninterlaced */
+                   0xc; /* reserved -- always b1100 */
+          buf[4] = 0xff; /* reserved -- always 1 */
+          break;
+    default:
+          buf[1] = buf[2] = buf[3] = buf[4] = 0xff;
+    }
+    return 5;
+}
+
+static void dv_format_frame(DVVideoContext* c, uint8_t* buf)
+{
+    int chan, i, j, k;
+
+    for (chan = 0; chan < c->sys->n_difchan; chan++) {
+        for (i = 0; i < c->sys->difseg_size; i++) {
+            memset(buf, 0xff, 80 * 6); /* First 6 DIF blocks are for control data */
+
+            /* DV header: 1DIF */
+            buf += dv_write_dif_id(dv_sect_header, chan, i, 0, buf);
+            buf += dv_write_pack((c->sys->dsf ? dv_header625 : dv_header525), c, buf);
+            buf += 72; /* unused bytes */
+
+            /* DV subcode: 2DIFs */
+            for (j = 0; j < 2; j++) {
+                buf += dv_write_dif_id(dv_sect_subcode, chan, i, j, buf);
+                for (k = 0; k < 6; k++)
+                     buf += dv_write_ssyb_id(k, (i < c->sys->difseg_size/2), buf) + 5;
+                buf += 29; /* unused bytes */
+            }
+
+            /* DV VAUX: 3DIFS */
+            for (j = 0; j < 3; j++) {
+                buf += dv_write_dif_id(dv_sect_vaux, chan, i, j, buf);
+                buf += dv_write_pack(dv_video_source,  c, buf);
+                buf += dv_write_pack(dv_video_control, c, buf);
+                buf += 7*5;
+                buf += dv_write_pack(dv_video_source,  c, buf);
+                buf += dv_write_pack(dv_video_control, c, buf);
+                buf += 4*5 + 2; /* unused bytes */
+            }
+
+            /* DV Audio/Video: 135 Video DIFs + 9 Audio DIFs */
+            for (j = 0; j < 135; j++) {
+                if (j%15 == 0) {
+                    memset(buf, 0xff, 80);
+                    buf += dv_write_dif_id(dv_sect_audio, chan, i, j/15, buf);
+                    buf += 77; /* audio control & shuffled PCM audio */
+                }
+                buf += dv_write_dif_id(dv_sect_video, chan, i, j, buf);
+                buf += 77; /* 1 video macro block: 1 bytes control
+                              4 * 14 bytes Y 8x8 data
+                              10 bytes Cr 8x8 data
+                              10 bytes Cb 8x8 data */
+            }
+        }
+    }
+}
+
+
+
 static int dvvideo_encode_frame(AVCodecContext *c, uint8_t *buf, int buf_size,
                                 void *data)
 {
@@ -1091,18 +1215,7 @@ static int dvvideo_encode_frame(AVCodecContext *c, uint8_t *buf, int buf_size,
 
     emms_c();
 
-    /* Fill in just enough of the header for dv_frame_profile() to
-       return the correct result, so that the frame can be decoded
-       correctly. The rest of the metadata is filled in by the dvvideo
-       avformat. (this should probably change so that encode_frame()
-       fills in ALL of the metadata - e.g. for Quicktime-wrapped DV
-       streams) */
-
-    /* NTSC/PAL format */
-    buf[3] = s->sys->dsf ? 0x80 : 0x00;
-
-    /* 25Mbps or 50Mbps */
-    buf[80*5 + 48 + 3] = (s->sys->pix_fmt == PIX_FMT_YUV422P) ? 0x4 : 0x0;
+    dv_format_frame(s, buf);
 
     return s->sys->frame_size;
 }
diff --git a/libavcodec/dvdata.h b/libavcodec/dvdata.h
index a3d42d66c1..87d34c0430 100644
--- a/libavcodec/dvdata.h
+++ b/libavcodec/dvdata.h
@@ -2624,6 +2624,29 @@ static const DVprofile dv_profiles[] = {
     }
 };
 
+enum dv_section_type {
+     dv_sect_header  = 0x1f,
+     dv_sect_subcode = 0x3f,
+     dv_sect_vaux    = 0x56,
+     dv_sect_audio   = 0x76,
+     dv_sect_video   = 0x96,
+};
+
+enum dv_pack_type {
+     dv_header525     = 0x3f, /* see dv_write_pack for important details on */
+     dv_header625     = 0xbf, /* these two packs */
+     dv_timecode      = 0x13,
+     dv_audio_source  = 0x50,
+     dv_audio_control = 0x51,
+     dv_audio_recdate = 0x52,
+     dv_audio_rectime = 0x53,
+     dv_video_source  = 0x60,
+     dv_video_control = 0x61,
+     dv_video_recdate = 0x62,
+     dv_video_rectime = 0x63,
+     dv_unknown_pack  = 0xff,
+};
+
 /* minimum number of bytes to read from a DV stream in order to determine the profile */
 #define DV_PROFILE_BYTES (6*80) /* 6 DIF blocks */
 
@@ -2663,3 +2686,37 @@ static inline const DVprofile* dv_codec_profile(AVCodecContext* codec)
 
     return NULL;
 }
+
+static inline int dv_write_dif_id(enum dv_section_type t, uint8_t chan_num, uint8_t seq_num,
+                                  uint8_t dif_num, uint8_t* buf)
+{
+    buf[0] = (uint8_t)t;    /* Section type */
+    buf[1] = (seq_num<<4) | /* DIF seq number 0-9 for 525/60; 0-11 for 625/50 */
+             (chan_num << 3) | /* FSC: for 50Mb/s 0 - first channel; 1 - second */
+             7;             /* reserved -- always 1 */
+    buf[2] = dif_num;       /* DIF block number Video: 0-134, Audio: 0-8 */
+    return 3;
+}
+
+
+static inline int dv_write_ssyb_id(uint8_t syb_num, uint8_t fr, uint8_t* buf)
+{
+    if (syb_num == 0 || syb_num == 6) {
+        buf[0] = (fr<<7) | /* FR ID 1 - first half of each channel; 0 - second */
+                 (0<<4)  | /* AP3 (Subcode application ID) */
+                 0x0f;     /* reserved -- always 1 */
+    }
+    else if (syb_num == 11) {
+        buf[0] = (fr<<7) | /* FR ID 1 - first half of each channel; 0 - second */
+                 0x7f;     /* reserved -- always 1 */
+    }
+    else {
+        buf[0] = (fr<<7) | /* FR ID 1 - first half of each channel; 0 - second */
+                 (0<<4)  | /* APT (Track application ID) */
+                 0x0f;     /* reserved -- always 1 */
+    }
+    buf[1] = 0xf0 |            /* reserved -- always 1 */
+             (syb_num & 0x0f); /* SSYB number 0 - 11   */
+    buf[2] = 0xff;             /* reserved -- always 1 */
+    return 3;
+}
diff --git a/libavformat/avformat.h b/libavformat/avformat.h
index 0429701b21..8f0c9bf43e 100644
--- a/libavformat/avformat.h
+++ b/libavformat/avformat.h
@@ -434,6 +434,16 @@ int fifo_read(FifoBuffer *f, uint8_t *buf, int buf_size, uint8_t **rptr_ptr);
 void fifo_write(FifoBuffer *f, const uint8_t *buf, int size, uint8_t **wptr_ptr);
 int put_fifo(ByteIOContext *pb, FifoBuffer *f, int buf_size, uint8_t **rptr_ptr);
 void fifo_realloc(FifoBuffer *f, unsigned int size);
+static inline uint8_t fifo_peek(FifoBuffer *f, int offs)
+{
+    return f->buffer[(f->rptr - f->buffer + offs) % (f->end - f->buffer)];
+}
+static inline void fifo_drain(FifoBuffer *f, int size)
+{
+    f->rptr += size;
+    if (f->rptr >= f->end)
+        f->rptr = f->buffer + (f->rptr - f->end);
+}
 
 /* media file input */
 AVInputFormat *av_find_input_format(const char *short_name);
diff --git a/libavformat/dv.c b/libavformat/dv.c
index b39d0b16b6..1e139ab3d7 100644
--- a/libavformat/dv.c
+++ b/libavformat/dv.c
@@ -49,93 +49,11 @@ struct DVMuxContext {
     FifoBuffer  audio_data[2]; /* Fifo for storing excessive amounts of PCM */
     int         frames;       /* Number of a current frame */
     time_t      start_time;   /* Start time of recording */
-    uint8_t     aspect;       /* Aspect ID 0 - 4:3, 7 - 16:9 */
     int         has_audio;    /* frame under contruction has audio */
     int         has_video;    /* frame under contruction has video */
     uint8_t     frame_buf[DV_MAX_FRAME_SIZE]; /* frame under contruction */
 };
 
-enum dv_section_type {
-     dv_sect_header  = 0x1f,
-     dv_sect_subcode = 0x3f,
-     dv_sect_vaux    = 0x56,
-     dv_sect_audio   = 0x76,
-     dv_sect_video   = 0x96,
-};
-
-enum dv_pack_type {
-     dv_header525     = 0x3f, /* see dv_write_pack for important details on */
-     dv_header625     = 0xbf, /* these two packs */
-     dv_timecode      = 0x13,
-     dv_audio_source  = 0x50,
-     dv_audio_control = 0x51,
-     dv_audio_recdate = 0x52,
-     dv_audio_rectime = 0x53,
-     dv_video_source  = 0x60,
-     dv_video_control = 0x61,
-     dv_video_recdate = 0x62,
-     dv_video_rectime = 0x63,
-     dv_unknown_pack  = 0xff,
-};
-
-
-
-/*
- * The reason why the following three big ugly looking tables are
- * here is my lack of DV spec IEC 61834. The tables were basically
- * constructed to make code that places packs in SSYB, VAUX and
- * AAUX blocks very simple and table-driven. They conform to the
- * SMPTE 314M and the output of my personal DV camcorder, neither
- * of which is sufficient for a reliable DV stream producing. Thus
- * while code is still in development I'll be gathering input from
- * people with different DV equipment and modifying the tables to
- * accommodate all the quirks. Later on, if possible, some of them
- * will be folded into smaller tables and/or switch-if logic. For
- * now, my only excuse is -- they don't eat up that much of a space.
- */
-
-static const int dv_ssyb_packs_dist[12][6] = {
-    { 0x13, 0x13, 0x13, 0x13, 0x13, 0x13 },
-    { 0x13, 0x13, 0x13, 0x13, 0x13, 0x13 },
-    { 0x13, 0x13, 0x13, 0x13, 0x13, 0x13 },
-    { 0x13, 0x13, 0x13, 0x13, 0x13, 0x13 },
-    { 0x13, 0x13, 0x13, 0x13, 0x13, 0x13 },
-    { 0x13, 0x13, 0x13, 0x13, 0x13, 0x13 },
-    { 0x13, 0x62, 0x63, 0x13, 0x62, 0x63 },
-    { 0x13, 0x62, 0x63, 0x13, 0x62, 0x63 },
-    { 0x13, 0x62, 0x63, 0x13, 0x62, 0x63 },
-    { 0x13, 0x62, 0x63, 0x13, 0x62, 0x63 },
-    { 0x13, 0x62, 0x63, 0x13, 0x62, 0x63 },
-    { 0x13, 0x62, 0x63, 0x13, 0x62, 0x63 },
-};
-
-static const int dv_vaux_packs_dist[12][15] = {
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-    { 0x60, 0x61, 0x62, 0x63, 0xff, 0xff, 0xff, 0xff, 0xff,
-      0x60, 0x61, 0x62, 0x63, 0xff, 0xff },
-};
-
 static const int dv_aaux_packs_dist[12][9] = {
     { 0xff, 0xff, 0xff, 0x50, 0x51, 0x52, 0x53, 0xff, 0xff },
     { 0x50, 0x51, 0x52, 0x53, 0xff, 0xff, 0xff, 0xff, 0xff },
@@ -189,20 +107,6 @@ static int dv_write_pack(enum dv_pack_type pack_id, DVMuxContext *c, uint8_t* bu
 
     buf[0] = (uint8_t)pack_id;
     switch (pack_id) {
-    case dv_header525: /* I can't imagine why these two weren't defined as real */
-    case dv_header625: /* packs in SMPTE314M -- they definitely look like ones */
-          buf[1] = 0xf8 |               /* reserved -- always 1 */
-                   (apt & 0x07);        /* APT: Track application ID */
-          buf[2] = (0 << 7)    | /* TF1: audio data is 0 - valid; 1 - invalid */
-                   (0x0f << 3) | /* reserved -- always 1 */
-                   (apt & 0x07); /* AP1: Audio application ID */
-          buf[3] = (0 << 7)    | /* TF2: video data is 0 - valid; 1 - invalid */
-                   (0x0f << 3) | /* reserved -- always 1 */
-                   (apt & 0x07); /* AP2: Video application ID */
-          buf[4] = (0 << 7)    | /* TF3: subcode(SSYB) is 0 - valid; 1 - invalid */
-                   (0x0f << 3) | /* reserved -- always 1 */
-                   (apt & 0x07); /* AP3: Subcode application ID */
-          break;
     case dv_timecode:
           ct = (time_t)(c->frames / ((float)c->sys->frame_rate /
                                      (float)c->sys->frame_rate_base));
@@ -293,117 +197,13 @@ static int dv_write_pack(enum dv_pack_type pack_id, DVMuxContext *c, uint8_t* bu
                    ((tc.tm_hour / 10) << 4) | /* Tens of hours */
                    (tc.tm_hour % 10);         /* Units of hours */
           break;
-    case dv_video_source:
-          buf[1] = 0xff; /* reserved -- always 1 */
-          buf[2] = (1 << 7) | /* B/W: 0 - b/w, 1 - color */
-                   (1 << 6) | /* following CLF is valid - 0, invalid - 1 */
-                   (3 << 4) | /* CLF: color frames id (see ITU-R BT.470-4) */
-                   0xf; /* reserved -- always 1 */
-          buf[3] = (3 << 6) | /* reserved -- always 1 */
-                   (c->sys->dsf << 5) | /*  system: 60fields/50fields */
-                   (apt << 2); /* signal type video compression */
-          buf[4] = 0xff; /* VISC: 0xff -- no information */
-          break;
-    case dv_video_control:
-          buf[1] = (0 << 6) | /* Copy generation management (CGMS) 0 -- free */
-                   0x3f; /* reserved -- always 1 */
-          buf[2] = 0xc8 | /* reserved -- always b11001xxx */
-                   c->aspect;
-          buf[3] = (1 << 7) | /* Frame/field flag 1 -- frame, 0 -- field */
-                   (1 << 6) | /* First/second field flag 0 -- field 2, 1 -- field 1 */
-                   (1 << 5) | /* Frame change flag 0 -- same picture as before, 1 -- different */
-                   (1 << 4) | /* 1 - interlaced, 0 - noninterlaced */
-                   0xc; /* reserved -- always b1100 */
-          buf[4] = 0xff; /* reserved -- always 1 */
-          break;
     default:
           buf[1] = buf[2] = buf[3] = buf[4] = 0xff;
     }
     return 5;
 }
 
-static inline int dv_write_dif_id(enum dv_section_type t, uint8_t chan_num, uint8_t seq_num,
-                                  uint8_t dif_num, uint8_t* buf)
-{
-    buf[0] = (uint8_t)t;    /* Section type */
-    buf[1] = (seq_num<<4) | /* DIF seq number 0-9 for 525/60; 0-11 for 625/50 */
-             (chan_num << 3) | /* FSC: for 50Mb/s 0 - first channel; 1 - second */
-             7;             /* reserved -- always 1 */
-    buf[2] = dif_num;       /* DIF block number Video: 0-134, Audio: 0-8 */
-    return 3;
-}
-
-static inline int dv_write_ssyb_id(uint8_t syb_num, uint8_t fr, uint8_t* buf)
-{
-    if (syb_num == 0 || syb_num == 6) {
-        buf[0] = (fr<<7) | /* FR ID 1 - first half of each channel; 0 - second */
-                 (0<<4)  | /* AP3 (Subcode application ID) */
-                 0x0f;     /* reserved -- always 1 */
-    }
-    else if (syb_num == 11) {
-        buf[0] = (fr<<7) | /* FR ID 1 - first half of each channel; 0 - second */
-                 0x7f;     /* reserved -- always 1 */
-    }
-    else {
-        buf[0] = (fr<<7) | /* FR ID 1 - first half of each channel; 0 - second */
-                 (0<<4)  | /* APT (Track application ID) */
-                 0x0f;     /* reserved -- always 1 */
-    }
-    buf[1] = 0xf0 |            /* reserved -- always 1 */
-             (syb_num & 0x0f); /* SSYB number 0 - 11   */
-    buf[2] = 0xff;             /* reserved -- always 1 */
-    return 3;
-}
-
-static void dv_format_frame(DVMuxContext *c, uint8_t* buf)
-{
-    int chan, i, j, k;
-
-    for (chan = 0; chan < c->sys->n_difchan; chan++) {
-        for (i = 0; i < c->sys->difseg_size; i++) {
-            memset(buf, 0xff, 80 * 6); /* First 6 DIF blocks are for control data */
-
-            /* DV header: 1DIF */
-            buf += dv_write_dif_id(dv_sect_header, chan, i, 0, buf);
-            buf += dv_write_pack((c->sys->dsf ? dv_header625 : dv_header525), c, buf);
-            buf += 72; /* unused bytes */
-
-            /* DV subcode: 2DIFs */
-            for (j = 0; j < 2; j++) {
-                buf += dv_write_dif_id(dv_sect_subcode, chan, i, j, buf);
-                for (k = 0; k < 6; k++) {
-                    buf += dv_write_ssyb_id(k, (i < c->sys->difseg_size/2), buf);
-                    buf += dv_write_pack(dv_ssyb_packs_dist[i][k], c, buf);
-                }
-                buf += 29; /* unused bytes */
-            }
-
-            /* DV VAUX: 3DIFs */
-            for (j = 0; j < 3; j++) {
-                buf += dv_write_dif_id(dv_sect_vaux, chan, i, j, buf);
-                for (k = 0; k < 15 ; k++)
-                    buf += dv_write_pack(dv_vaux_packs_dist[i][k], c, buf);
-                buf += 2; /* unused bytes */
-            }
-
-            /* DV Audio/Video: 135 Video DIFs + 9 Audio DIFs */
-            for (j = 0; j < 135; j++) {
-                if (j%15 == 0) {
-                    memset(buf, 0xff, 80);
-                    buf += dv_write_dif_id(dv_sect_audio, chan, i, j/15, buf);
-                    buf += 77; /* audio control & shuffled PCM audio */
-                }
-                buf += dv_write_dif_id(dv_sect_video, chan, i, j, buf);
-                buf += 77; /* 1 video macro block: 1 bytes control
-                              4 * 14 bytes Y 8x8 data
-                              10 bytes Cr 8x8 data
-                              10 bytes Cb 8x8 data */
-            }
-        }
-    }
-}
-
-static void dv_inject_audio(DVMuxContext *c, const uint8_t* pcm, int channel, uint8_t* frame_ptr)
+static void dv_inject_audio(DVMuxContext *c, int channel, uint8_t* frame_ptr)
 {
     int i, j, d, of, size;
     size = 4 * dv_audio_frame_size(c->sys, c->frames);
@@ -417,30 +217,40 @@ static void dv_inject_audio(DVMuxContext *c, const uint8_t* pcm, int channel, ui
              if (of*2 >= size)
                  continue;
 
-             frame_ptr[d] = pcm[of*2+1]; // FIXME: may be we have to admit
-             frame_ptr[d+1] = pcm[of*2]; //        that DV is a big endian PCM
+             frame_ptr[d] = fifo_peek(&c->audio_data[channel], of*2+1); // FIXME: may be we have to admit
+             frame_ptr[d+1] = fifo_peek(&c->audio_data[channel], of*2); //        that DV is a big endian PCM
           }
           frame_ptr += 16 * 80; /* 15 Video DIFs + 1 Audio DIF */
        }
     }
 }
 
-static void dv_inject_video(DVMuxContext *c, const uint8_t* video_data, uint8_t* frame_ptr)
+static void dv_inject_metadata(DVMuxContext *c, uint8_t* frame)
 {
-    int chan, i, j;
-    int ptr = 0;
-
-    for (chan = 0; chan < c->sys->n_difchan; chan++) {
-        for (i = 0; i < c->sys->difseg_size; i++) {
-            ptr += 6 * 80; /* skip DIF segment header */
-            for (j = 0; j < 135; j++) {
-                if (j%15 == 0)
-                    ptr += 80; /* skip Audio DIF */
-                ptr += 3;
-                memcpy(frame_ptr + ptr, video_data + ptr, 77);
-                ptr += 77;
+    int j, k;
+    uint8_t* buf;
+
+    for (buf = frame; buf < frame + c->sys->frame_size; buf += 150 * 80) {
+        /* DV subcode: 2nd and 3d DIFs */
+        for (j = 80; j < 80 * 3; j += 80) {
+            for (k = 6; k < 6 * 8; k += 8)
+                dv_write_pack(dv_timecode, c, &buf[j+k]);
+
+            if (((long)(buf-frame)/(c->sys->frame_size/(c->sys->difseg_size*c->sys->n_difchan))%c->sys->difseg_size) > 5) { /* FIXME: is this really needed ? */
+                dv_write_pack(dv_video_recdate, c, &buf[j+14]);
+                dv_write_pack(dv_video_rectime, c, &buf[j+22]);
+                dv_write_pack(dv_video_recdate, c, &buf[j+38]);
+                dv_write_pack(dv_video_rectime, c, &buf[j+46]);
             }
         }
+
+        /* DV VAUX: 4th, 5th and 6th 3DIFs */
+        for (j = 80*3 + 3; j < 80*6; j += 80) {
+            dv_write_pack(dv_video_recdate, c, &buf[j+5*2]);
+            dv_write_pack(dv_video_rectime, c, &buf[j+5*3]);
+            dv_write_pack(dv_video_recdate, c, &buf[j+5*11]);
+            dv_write_pack(dv_video_rectime, c, &buf[j+5*12]);
+        }
     }
 }
 
@@ -643,60 +453,52 @@ static int dv_extract_video_info(DVDemuxContext *c, uint8_t* frame)
 int dv_assemble_frame(DVMuxContext *c, AVStream* st,
                       const uint8_t* data, int data_size, uint8_t** frame)
 {
-    uint8_t pcm[8192];
-    int i;
+    int i, reqasize;
 
     *frame = &c->frame_buf[0];
-    if (c->has_audio && c->has_video &&
-        (c->has_audio == -1 || c->has_audio == c->n_ast)) {
-        /* must be a stale frame */
-        dv_format_frame(c, *frame);
-        c->frames++;
-        if (c->has_audio > 0)
-            c->has_audio = 0;
-        c->has_video = 0;
-    }
+    reqasize = 4 * dv_audio_frame_size(c->sys, c->frames);
+
+    switch (st->codec->codec_type) {
+    case CODEC_TYPE_VIDEO:
+          /* FIXME: we have to have more sensible approach than this one */
+          if (c->has_video)
+              av_log(st->codec, AV_LOG_ERROR, "Can't process DV frame #%d. Insufficient audio data or severe sync problem.\n", c->frames);
 
-    if (st->codec->codec_type == CODEC_TYPE_VIDEO) {
-        /* FIXME: we have to have more sensible approach than this one */
-        if (c->has_video)
-            av_log(st->codec, AV_LOG_ERROR, "Can't process DV frame #%d. Insufficient audio data or severe sync problem.\n", c->frames);
+          memcpy(*frame, data, c->sys->frame_size);
+          c->has_video = 1;
+          break;
+    case CODEC_TYPE_AUDIO:
+          for (i = 0; i < c->n_ast && st != c->ast[i]; i++);
+
+          /* FIXME: we have to have more sensible approach than this one */
+          if (fifo_size(&c->audio_data[i], c->audio_data[i].rptr) + data_size >= 100*AVCODEC_MAX_AUDIO_FRAME_SIZE)
+              av_log(st->codec, AV_LOG_ERROR, "Can't process DV frame #%d. Insufficient video data or severe sync problem.\n", c->frames);
+          fifo_write(&c->audio_data[i], data, data_size, &c->audio_data[i].wptr);
 
-        dv_inject_video(c, data, *frame);
-        c->has_video = 1;
-        data_size = 0;
-        if (c->has_audio < 0)
-            goto out;
+          /* Lets see if we've got enough audio for one DV frame */
+          c->has_audio |= ((reqasize <= fifo_size(&c->audio_data[i], c->audio_data[i].rptr)) << i);
+
+          break;
+    default:
+          break;
     }
 
-    for (i = 0; i < c->n_ast; i++) {
-        int reqasize, fsize;
-        if (st != c->ast[i])
-            continue;
-        reqasize = 4 * dv_audio_frame_size(c->sys, c->frames);
-        fsize = fifo_size(&c->audio_data[i], c->audio_data[i].rptr);
-        if (st->codec->codec_type == CODEC_TYPE_AUDIO ||
-            (c->has_video && fsize >= reqasize)) {
-            if (fsize + data_size >= reqasize && (c->has_audio < c->n_ast)) {
-                if (fsize >= reqasize) {
-                    fifo_read(&c->audio_data[i], &pcm[0], reqasize, &c->audio_data[i].rptr);
-                } else {
-                    fifo_read(&c->audio_data[i], &pcm[0], fsize, &c->audio_data[i].rptr);
-                    memcpy(&pcm[fsize], &data[0], reqasize - fsize);
-                    data += reqasize - fsize;
-                    data_size -= reqasize - fsize;
-                }
-                dv_inject_audio(c, &pcm[0], i, *frame);
-                c->has_audio += 1;
-            }
-            /* FIXME: we have to have more sensible approach than this one */
-            if (fifo_size(&c->audio_data[i], c->audio_data[i].rptr) + data_size >= 100*AVCODEC_MAX_AUDIO_FRAME_SIZE)
-                av_log(st->codec, AV_LOG_ERROR, "Can't process DV frame #%d. Insufficient video data or severe sync problem.\n", c->frames);
-            fifo_write(&c->audio_data[i], data, data_size, &c->audio_data[i].wptr);
+    /* Lets see if we have enough data to construct one DV frame */
+    if (c->has_video == 1 && c->has_audio + 1 == 1<<c->n_ast) {
+        dv_inject_metadata(c, *frame);
+        for (i=0; i<c->n_ast; i++) {
+             dv_inject_audio(c, i, *frame);
+             fifo_drain(&c->audio_data[i], reqasize);
         }
+
+        c->has_video = 0;
+        c->has_audio = 0;
+        c->frames++;
+
+        return c->sys->frame_size;
     }
-out:
-    return ((c->has_audio == -1 || c->has_audio == c->n_ast) && c->has_video) ? c->sys->frame_size : 0;
+
+    return 0;
 }
 
 DVMuxContext* dv_init_mux(AVFormatContext* s)
@@ -750,12 +552,9 @@ DVMuxContext* dv_init_mux(AVFormatContext* s)
 
     /* Ok, everything seems to be in working order */
     c->frames = 0;
-    c->has_audio = c->n_ast ? 0 : -1;
+    c->has_audio = 0;
     c->has_video = 0;
     c->start_time = (time_t)s->timestamp;
-    c->aspect = 0; /* 4:3 is the default */
-    if ((int)(av_q2d(vst->codec->sample_aspect_ratio) * vst->codec->width / vst->codec->height * 10) == 17) /* 16:9 */
-        c->aspect = 0x07;
 
     for (i=0; i<c->n_ast; i++) {
         if (c->ast[i] && fifo_init(&c->audio_data[i], 100*AVCODEC_MAX_AUDIO_FRAME_SIZE) < 0) {
@@ -767,8 +566,6 @@ DVMuxContext* dv_init_mux(AVFormatContext* s)
         }
     }
 
-    dv_format_frame(c, &c->frame_buf[0]);
-
     return c;
 
 bail_out:
diff --git a/tests/ffmpeg.regression.ref b/tests/ffmpeg.regression.ref
index 5958ae26ba..1aec954d3b 100644
--- a/tests/ffmpeg.regression.ref
+++ b/tests/ffmpeg.regression.ref
@@ -141,11 +141,11 @@ f998347483cef957af497aa0055dca0c *./data/a-snow53.avi
 3534748 ./data/a-snow53.avi
 799d3db687f6cdd7a837ec156efc171f *./data/out.yuv
 stddev:  0.00 PSNR:99.99 bytes:7602176
-85529d9838c5a8aef041445fa8ef3c85 *./data/a-dv.dv
+2fcbcdc63816e1321bf4b6b5380338d2 *./data/a-dv.dv
 7200000 ./data/a-dv.dv
 c2082cd8adf417c4ebc32654e446cba1 *./data/out.yuv
 stddev:  8.86 PSNR:29.17 bytes:7602176
-ff5eb084624f9a2846840a215b95a112 *./data/a-dv.dv
+4f71942eb699bf3b12508a9e777a319f *./data/a-dv.dv
 14400000 ./data/a-dv.dv
 fc866b8879a34f5b440647e5135e4bfb *./data/out.yuv
 stddev:  8.45 PSNR:29.58 bytes:7602176
diff --git a/tests/libav.regression.ref b/tests/libav.regression.ref
index 458ae427f4..f337e05ee1 100644
--- a/tests/libav.regression.ref
+++ b/tests/libav.regression.ref
@@ -28,7 +28,7 @@ ffmpeg regression test
 8bf16d40a2ec19fa36b124a928e47e23 *./data/b-libav.nut
  332358 ./data/b-libav.nut
 ./data/b-libav.nut CRC=0x1fdd352c
-2cc647904d3cca01703f36056b9d3f42 *./data/b-libav.dv
+26c41db318d9aacfd6b9e734c0ea4d94 *./data/b-libav.dv
 3600000 ./data/b-libav.dv
 ./data/b-libav.dv CRC=0xa6b8b635
 304ee17506e4526ba04285f6cd14630f *./data/b-libav.gxf
diff --git a/tests/rotozoom.regression.ref b/tests/rotozoom.regression.ref
index 3af0b93a5d..5652f2d62b 100644
--- a/tests/rotozoom.regression.ref
+++ b/tests/rotozoom.regression.ref
@@ -141,11 +141,11 @@ stddev: 10.93 PSNR:27.34 bytes:7602176
 2726606 ./data/a-snow53.avi
 dde5895817ad9d219f79a52d0bdfb001 *./data/out.yuv
 stddev:  0.00 PSNR:99.99 bytes:7602176
-6568326ee439099e8a0e1680336ed351 *./data/a-dv.dv
+af9f474238c9c68cb32e389659ee25ab *./data/a-dv.dv
 7200000 ./data/a-dv.dv
 bb69dda7a84a5b166434e28e1243d3d1 *./data/out.yuv
 stddev:  2.99 PSNR:38.59 bytes:7602176
-177cbbbe50cbb67ae3bd6e7f10ff968b *./data/a-dv.dv
+91dc1c1dc4a8cca72f27d85db6d7636f *./data/a-dv.dv
 14400000 ./data/a-dv.dv
 74b01209bb5e096d570dd4df112bb82d *./data/out.yuv
 stddev:  2.98 PSNR:38.61 bytes:7602176