diff --git a/libavcodec/truespeech.c b/libavcodec/truespeech.c
index edaaea686a..fea9ae4195 100644
--- a/libavcodec/truespeech.c
+++ b/libavcodec/truespeech.c
@@ -36,7 +36,6 @@
  * TrueSpeech decoder context
  */
 typedef struct {
-    AVFrame frame;
     DSPContext dsp;
     /* input data */
     DECLARE_ALIGNED(16, uint8_t, buffer)[32];
@@ -73,9 +72,6 @@ static av_cold int truespeech_decode_init(AVCodecContext * avctx)
 
     ff_dsputil_init(&c->dsp, avctx);
 
-    avcodec_get_frame_defaults(&c->frame);
-    avctx->coded_frame = &c->frame;
-
     return 0;
 }
 
@@ -310,6 +306,7 @@ static void truespeech_save_prevvec(TSContext *c)
 static int truespeech_decode_frame(AVCodecContext *avctx, void *data,
                                    int *got_frame_ptr, AVPacket *avpkt)
 {
+    AVFrame *frame     = data;
     const uint8_t *buf = avpkt->data;
     int buf_size = avpkt->size;
     TSContext *c = avctx->priv_data;
@@ -327,12 +324,12 @@ static int truespeech_decode_frame(AVCodecContext *avctx, void *data,
     }
 
     /* get output buffer */
-    c->frame.nb_samples = iterations * 240;
-    if ((ret = ff_get_buffer(avctx, &c->frame)) < 0) {
+    frame->nb_samples = iterations * 240;
+    if ((ret = ff_get_buffer(avctx, frame)) < 0) {
         av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
         return ret;
     }
-    samples = (int16_t *)c->frame.data[0];
+    samples = (int16_t *)frame->data[0];
 
     memset(samples, 0, iterations * 240 * sizeof(*samples));
 
@@ -354,8 +351,7 @@ static int truespeech_decode_frame(AVCodecContext *avctx, void *data,
         truespeech_save_prevvec(c);
     }
 
-    *got_frame_ptr   = 1;
-    *(AVFrame *)data = c->frame;
+    *got_frame_ptr = 1;
 
     return buf_size;
 }
diff --git a/libavcodec/tta.c b/libavcodec/tta.c
index 9b51a05fb4..bcbe037b66 100644
--- a/libavcodec/tta.c
+++ b/libavcodec/tta.c
@@ -61,7 +61,6 @@ typedef struct TTAChannel {
 typedef struct TTAContext {
     AVClass *class;
     AVCodecContext *avctx;
-    AVFrame frame;
     GetBitContext gb;
     const AVCRC *crc_table;
 
@@ -312,15 +311,13 @@ static av_cold int tta_decode_init(AVCodecContext * avctx)
         return AVERROR_INVALIDDATA;
     }
 
-    avcodec_get_frame_defaults(&s->frame);
-    avctx->coded_frame = &s->frame;
-
     return 0;
 }
 
 static int tta_decode_frame(AVCodecContext *avctx, void *data,
                             int *got_frame_ptr, AVPacket *avpkt)
 {
+    AVFrame *frame     = data;
     const uint8_t *buf = avpkt->data;
     int buf_size = avpkt->size;
     TTAContext *s = avctx->priv_data;
@@ -336,15 +333,15 @@ static int tta_decode_frame(AVCodecContext *avctx, void *data,
     init_get_bits(&s->gb, buf, buf_size*8);
 
     /* get output buffer */
-    s->frame.nb_samples = framelen;
-    if ((ret = ff_get_buffer(avctx, &s->frame)) < 0) {
+    frame->nb_samples = framelen;
+    if ((ret = ff_get_buffer(avctx, frame)) < 0) {
         av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
         return ret;
     }
 
     // decode directly to output buffer for 24-bit sample format
     if (s->bps == 3)
-        s->decode_buffer = (int32_t *)s->frame.data[0];
+        s->decode_buffer = (int32_t *)frame->data[0];
 
     // init per channel states
     for (i = 0; i < s->channels; i++) {
@@ -433,7 +430,7 @@ static int tta_decode_frame(AVCodecContext *avctx, void *data,
             i++;
             // check for last frame
             if (i == s->last_frame_length && get_bits_left(&s->gb) / 8 == 4) {
-                s->frame.nb_samples = framelen = s->last_frame_length;
+                frame->nb_samples = framelen = s->last_frame_length;
                 break;
             }
         }
@@ -449,20 +446,20 @@ static int tta_decode_frame(AVCodecContext *avctx, void *data,
     // convert to output buffer
     switch (s->bps) {
     case 1: {
-        uint8_t *samples = (uint8_t *)s->frame.data[0];
+        uint8_t *samples = (uint8_t *)frame->data[0];
         for (p = s->decode_buffer; p < s->decode_buffer + (framelen * s->channels); p++)
             *samples++ = *p + 0x80;
         break;
         }
     case 2: {
-        int16_t *samples = (int16_t *)s->frame.data[0];
+        int16_t *samples = (int16_t *)frame->data[0];
         for (p = s->decode_buffer; p < s->decode_buffer + (framelen * s->channels); p++)
             *samples++ = *p;
         break;
         }
     case 3: {
         // shift samples for 24-bit sample format
-        int32_t *samples = (int32_t *)s->frame.data[0];
+        int32_t *samples = (int32_t *)frame->data[0];
         for (p = s->decode_buffer; p < s->decode_buffer + (framelen * s->channels); p++)
             *samples++ <<= 8;
         // reset decode buffer
@@ -471,8 +468,7 @@ static int tta_decode_frame(AVCodecContext *avctx, void *data,
         }
     }
 
-    *got_frame_ptr   = 1;
-    *(AVFrame *)data = s->frame;
+    *got_frame_ptr = 1;
 
     return buf_size;
 error:
diff --git a/libavcodec/twinvq.c b/libavcodec/twinvq.c
index 8bb44cce29..51a4478825 100644
--- a/libavcodec/twinvq.c
+++ b/libavcodec/twinvq.c
@@ -177,7 +177,6 @@ static const ModeTab mode_44_48 = {
 
 typedef struct TwinContext {
     AVCodecContext *avctx;
-    AVFrame frame;
     AVFloatDSPContext fdsp;
     FFTContext mdct_ctx[3];
 
@@ -811,6 +810,7 @@ static void read_and_decode_spectrum(TwinContext *tctx, GetBitContext *gb,
 static int twin_decode_frame(AVCodecContext * avctx, void *data,
                              int *got_frame_ptr, AVPacket *avpkt)
 {
+    AVFrame *frame     = data;
     const uint8_t *buf = avpkt->data;
     int buf_size = avpkt->size;
     TwinContext *tctx = avctx->priv_data;
@@ -832,12 +832,12 @@ static int twin_decode_frame(AVCodecContext * avctx, void *data,
 
     /* get output buffer */
     if (tctx->discarded_packets >= 2) {
-        tctx->frame.nb_samples = mtab->size;
-        if ((ret = ff_get_buffer(avctx, &tctx->frame)) < 0) {
+        frame->nb_samples = mtab->size;
+        if ((ret = ff_get_buffer(avctx, frame)) < 0) {
             av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
             return ret;
         }
-        out = (float **)tctx->frame.extended_data;
+        out = (float **)frame->extended_data;
     }
 
     init_get_bits(&gb, buf, buf_size * 8);
@@ -863,8 +863,7 @@ static int twin_decode_frame(AVCodecContext * avctx, void *data,
         return buf_size;
     }
 
-    *got_frame_ptr   = 1;
-    *(AVFrame *)data = tctx->frame;
+    *got_frame_ptr = 1;
 
     return buf_size;
 }
@@ -1171,9 +1170,6 @@ static av_cold int twin_decode_init(AVCodecContext *avctx)
 
     memset_float(tctx->bark_hist[0][0], 0.1, FF_ARRAY_ELEMS(tctx->bark_hist));
 
-    avcodec_get_frame_defaults(&tctx->frame);
-    avctx->coded_frame = &tctx->frame;
-
     return 0;
 }
 
diff --git a/libavcodec/vmdav.c b/libavcodec/vmdav.c
index 6da60c14a1..d7f136c8a1 100644
--- a/libavcodec/vmdav.c
+++ b/libavcodec/vmdav.c
@@ -469,7 +469,6 @@ static av_cold int vmdvideo_decode_end(AVCodecContext *avctx)
 #define BLOCK_TYPE_SILENCE  3
 
 typedef struct VmdAudioContext {
-    AVFrame frame;
     int out_bps;
     int chunk_size;
 } VmdAudioContext;
@@ -514,9 +513,6 @@ static av_cold int vmdaudio_decode_init(AVCodecContext *avctx)
 
     s->chunk_size = avctx->block_align + avctx->channels * (s->out_bps == 2);
 
-    avcodec_get_frame_defaults(&s->frame);
-    avctx->coded_frame = &s->frame;
-
     av_log(avctx, AV_LOG_DEBUG, "%d channels, %d bits/sample, "
            "block align = %d, sample rate = %d\n",
            avctx->channels, avctx->bits_per_coded_sample, avctx->block_align,
@@ -557,6 +553,7 @@ static void decode_audio_s16(int16_t *out, const uint8_t *buf, int buf_size,
 static int vmdaudio_decode_frame(AVCodecContext *avctx, void *data,
                                  int *got_frame_ptr, AVPacket *avpkt)
 {
+    AVFrame *frame     = data;
     const uint8_t *buf = avpkt->data;
     const uint8_t *buf_end;
     int buf_size = avpkt->size;
@@ -601,13 +598,14 @@ static int vmdaudio_decode_frame(AVCodecContext *avctx, void *data,
     audio_chunks = buf_size / s->chunk_size;
 
     /* get output buffer */
-    s->frame.nb_samples = ((silent_chunks + audio_chunks) * avctx->block_align) / avctx->channels;
-    if ((ret = ff_get_buffer(avctx, &s->frame)) < 0) {
+    frame->nb_samples = ((silent_chunks + audio_chunks) * avctx->block_align) /
+                        avctx->channels;
+    if ((ret = ff_get_buffer(avctx, frame)) < 0) {
         av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
         return ret;
     }
-    output_samples_u8  = s->frame.data[0];
-    output_samples_s16 = (int16_t *)s->frame.data[0];
+    output_samples_u8  =            frame->data[0];
+    output_samples_s16 = (int16_t *)frame->data[0];
 
     /* decode silent chunks */
     if (silent_chunks > 0) {
@@ -637,8 +635,7 @@ static int vmdaudio_decode_frame(AVCodecContext *avctx, void *data,
         }
     }
 
-    *got_frame_ptr   = 1;
-    *(AVFrame *)data = s->frame;
+    *got_frame_ptr = 1;
 
     return avpkt->size;
 }
diff --git a/libavcodec/vorbisdec.c b/libavcodec/vorbisdec.c
index deaf4fd7dc..645673c0c6 100644
--- a/libavcodec/vorbisdec.c
+++ b/libavcodec/vorbisdec.c
@@ -125,7 +125,6 @@ typedef struct {
 
 typedef struct vorbis_context_s {
     AVCodecContext *avccontext;
-    AVFrame frame;
     GetBitContext gb;
     VorbisDSPContext dsp;
     AVFloatDSPContext fdsp;
@@ -1040,9 +1039,6 @@ static av_cold int vorbis_decode_init(AVCodecContext *avccontext)
     avccontext->channels    = vc->audio_channels;
     avccontext->sample_rate = vc->audio_samplerate;
 
-    avcodec_get_frame_defaults(&vc->frame);
-    avccontext->coded_frame = &vc->frame;
-
     return 0;
 }
 
@@ -1653,6 +1649,7 @@ static int vorbis_decode_frame(AVCodecContext *avccontext, void *data,
     const uint8_t *buf = avpkt->data;
     int buf_size       = avpkt->size;
     vorbis_context *vc = avccontext->priv_data;
+    AVFrame *frame     = data;
     GetBitContext *gb = &vc->gb;
     float *channel_ptrs[255];
     int i, len, ret;
@@ -1699,19 +1696,19 @@ static int vorbis_decode_frame(AVCodecContext *avccontext, void *data,
     }
 
     /* get output buffer */
-    vc->frame.nb_samples = vc->blocksize[1] / 2;
-    if ((ret = ff_get_buffer(avccontext, &vc->frame)) < 0) {
+    frame->nb_samples = vc->blocksize[1] / 2;
+    if ((ret = ff_get_buffer(avccontext, frame)) < 0) {
         av_log(avccontext, AV_LOG_ERROR, "get_buffer() failed\n");
         return ret;
     }
 
     if (vc->audio_channels > 8) {
         for (i = 0; i < vc->audio_channels; i++)
-            channel_ptrs[i] = (float *)vc->frame.extended_data[i];
+            channel_ptrs[i] = (float *)frame->extended_data[i];
     } else {
         for (i = 0; i < vc->audio_channels; i++) {
             int ch = ff_vorbis_channel_layout_offsets[vc->audio_channels - 1][i];
-            channel_ptrs[ch] = (float *)vc->frame.extended_data[i];
+            channel_ptrs[ch] = (float *)frame->extended_data[i];
         }
     }
 
@@ -1729,9 +1726,8 @@ static int vorbis_decode_frame(AVCodecContext *avccontext, void *data,
     av_dlog(NULL, "parsed %d bytes %d bits, returned %d samples (*ch*bits) \n",
             get_bits_count(gb) / 8, get_bits_count(gb) % 8, len);
 
-    vc->frame.nb_samples = len;
-    *got_frame_ptr   = 1;
-    *(AVFrame *)data = vc->frame;
+    frame->nb_samples = len;
+    *got_frame_ptr    = 1;
 
     return buf_size;
 }