[FFmpeg-devel] [PATCH] 1 of 5 Add Speex Encoding

[Art Clarke]

Encodes speex audio with the following features:
1) FLT or S16 audio
2) mono or stereo
3) narrow, wide, or ultra band
4) either settable bit-rate or quality settings

Beyond whatever comments you have, I had some requests:
1) Adobe's implementation of speex in FLV asusmes speex to be encoded with
two frames per packet, so that's what this encoder does.  I'd like to make
this configurable, but couldn't find an AVCodecContext field to use for
that.
2) I'd like to support variable bit-rate, but didn't know which
AVCodecContext flag to query for that.

Comments appreciated,

- Art

-- 
http://www.xuggle.com/
xu?ggle (z?' gl) v. To freely encode, decode, and experience audio and
video.

Use Xuggle to get the power of FFmpeg in Java.
-------------- next part --------------
Index: libavcodec/libspeexenc.c
===================================================================
--- libavcodec/libspeexenc.c	(revision 0)
+++ libavcodec/libspeexenc.c	(revision 0)
@@ -0,0 +1,177 @@
+/*
+ * Copyright (c) 2009 by Xuggle Incorporated.  All rights reserved.
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+#include <libavcodec/avcodec.h>
+#include <speex/speex.h>
+#include <speex/speex_header.h>
+
+typedef struct {
+    SpeexBits bits;
+    void *enc_state;
+    SpeexHeader header;
+} LibSpeexEncContext;
+
+
+static av_cold int libspeex_encode_init(AVCodecContext *avctx)
+{
+    LibSpeexEncContext *s = (LibSpeexEncContext*)avctx->priv_data;
+    const SpeexMode *mode;
+
+    if ((avctx->sample_fmt != SAMPLE_FMT_S16 && avctx->sample_fmt != SAMPLE_FMT_FLT) ||
+            avctx->sample_rate <= 0 ||
+            avctx->channels <= 0 ||
+            avctx->channels > 2)
+    {
+        av_log(avctx, AV_LOG_ERROR, "Unsupported sample format, rate, or channels for speex");
+        return -1;
+    }
+
+    if (avctx->sample_rate <= 8000)
+        mode = &speex_nb_mode;
+    else if (avctx->sample_rate <= 16000)
+        mode = &speex_wb_mode;
+    else
+        mode = &speex_uwb_mode;
+
+    speex_bits_init(&s->bits);
+    s->enc_state = speex_encoder_init(mode);
+    if (!s->enc_state)
+    {
+        av_log(avctx, AV_LOG_ERROR, "could not initialize speex encoder");
+        return -1;
+    }
+
+    // initialize the header
+    speex_init_header(&s->header, avctx->sample_rate,
+            avctx->channels, mode);
+
+    // TODO: It'd be nice to support VBR here, but
+    // I'm uncertain what AVCodecContext options to use
+    // to signal whether to turn it on.
+    if (avctx->flags & CODEC_FLAG_QSCALE) {
+        spx_int32_t quality = 0;
+        // Map global_quality's mpeg 1/2/4 scale into Speex's 0-10 scale
+        if (avctx->global_quality > FF_LAMBDA_MAX)
+            quality = 0; // lowest possible quality
+        else
+            quality = (spx_int32_t)((FF_LAMBDA_MAX-avctx->global_quality)*10.0/FF_LAMBDA_MAX);
+        speex_encoder_ctl(s->enc_state, SPEEX_SET_QUALITY, &quality);
+    } else {
+        // default to CBR
+        if (avctx->bit_rate > 0)
+            speex_encoder_ctl(s->enc_state, SPEEX_SET_BITRATE, &avctx->bit_rate);
+        // otherwise just take the default quality setting
+    }
+    // reset the bit-rate to the actual bit rate speex will use
+    speex_encoder_ctl(s->enc_state, SPEEX_GET_BITRATE, &s->header.bitrate);
+    avctx->bit_rate = s->header.bitrate;
+
+    // get the actual sample rate
+    speex_encoder_ctl(s->enc_state, SPEEX_GET_SAMPLING_RATE, &s->header.rate);
+    avctx->sample_rate = s->header.rate;
+
+    // get the frame-size.  To align with FLV, we're going to put 2 frames
+    // per packet.  If someone can tell me how to make this configurable
+    // from the avcodec contents, I'll mod this so it's not hard-coded.
+    // but without this, FLV files with speex data won't play correctly
+    // in flash player 10.
+    speex_encoder_ctl(s->enc_state, SPEEX_GET_FRAME_SIZE, &s->header.frame_size);
+    s->header.frames_per_packet = 2; // Need for FLV container support
+    avctx->frame_size = s->header.frame_size*s->header.frames_per_packet;
+
+    // and we'll put a speex header packet into extradata so that muxers
+    // can use it.
+    avctx->extradata = speex_header_to_packet(&s->header, &avctx->extradata_size);
+    return 0;
+}
+
+static av_cold int libspeex_encode_frame(
+        AVCodecContext *avctx, uint8_t *frame,
+        int buf_size, void *data)
+{
+    LibSpeexEncContext *s = (LibSpeexEncContext*)avctx->priv_data;
+    int i = 0;
+
+    if (!data)
+        // nothing to flush
+        return 0;
+
+    speex_bits_reset(&s->bits);
+    for(i = 0; i < s->header.frames_per_packet; i++)
+    {
+        if (avctx->sample_fmt == SAMPLE_FMT_FLT)
+        {
+            if (avctx->channels == 2) {
+                speex_encode_stereo(
+                        (float*)data+i*s->header.frame_size,
+                        s->header.frame_size,
+                        &s->bits);
+            }
+            speex_encode(s->enc_state,
+                    (float*)data+i*s->header.frame_size, &s->bits); 
+        } else {
+            if (avctx->channels == 2) {
+                speex_encode_stereo_int(
+                        (spx_int16_t*)data+i*s->header.frame_size,
+                        s->header.frame_size,
+                        &s->bits);
+            }
+            speex_encode_int(s->enc_state,
+                    (spx_int16_t*)data+i*s->header.frame_size, &s->bits);
+        }
+    }
+    // put in a terminator so this will fit in a OGG or FLV packet
+    speex_bits_insert_terminator(&s->bits);
+
+    if (buf_size >= speex_bits_nbytes(&s->bits)) {
+        return speex_bits_write(&s->bits, frame, buf_size);
+    } else {
+        av_log(avctx, AV_LOG_ERROR, "output buffer too small");
+        return -1;
+    }
+}
+
+static av_cold int libspeex_encode_close(AVCodecContext *avctx)
+{
+    LibSpeexEncContext *s = (LibSpeexEncContext*)avctx->priv_data;
+
+    speex_bits_destroy(&s->bits);
+    speex_encoder_destroy(s->enc_state);
+    s->enc_state = 0;
+    if (avctx->extradata)
+        speex_header_free(avctx->extradata);
+    avctx->extradata = 0;
+    avctx->extradata_size = 0;
+
+    return 0;
+}
+
+AVCodec libspeex_encoder = {
+    "libspeex",
+    CODEC_TYPE_AUDIO,
+    CODEC_ID_SPEEX,
+    sizeof(LibSpeexEncContext),
+    libspeex_encode_init,
+    libspeex_encode_frame,
+    libspeex_encode_close,
+    0,
+    .capabilities = CODEC_CAP_DELAY,
+    .supported_samplerates = (const int[]){8000, 16000, 32000, 0},
+    .sample_fmts = (enum SampleFormat[]){SAMPLE_FMT_S16,SAMPLE_FMT_FLT,SAMPLE_FMT_NONE},
+    .long_name = NULL_IF_CONFIG_SMALL("libspeex Speex Encoder"),
+};