1
0
mirror of https://github.com/mpv-player/mpv synced 2024-12-21 22:30:22 +00:00
mpv/audio/format.c

225 lines
6.7 KiB
C
Raw Normal View History

/*
* Copyright (C) 2005 Alex Beregszaszi
*
* This file is part of MPlayer.
*
* MPlayer is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* MPlayer 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License along
* with MPlayer; if not, write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <inttypes.h>
#include <limits.h>
#include <assert.h>
#include "common/common.h"
#include "audio/filter/af.h"
int af_fmt2bps(int format)
{
2013-11-07 21:12:26 +00:00
switch (format & AF_FORMAT_BITS_MASK) {
case AF_FORMAT_8BIT: return 1;
case AF_FORMAT_16BIT: return 2;
case AF_FORMAT_24BIT: return 3;
case AF_FORMAT_32BIT: return 4;
case AF_FORMAT_64BIT: return 8;
}
return 0;
}
int af_fmt2bits(int format)
{
return af_fmt2bps(format) * 8;
}
static int bits_to_mask(int bits)
{
switch (bits) {
2013-11-07 21:12:26 +00:00
case 8: return AF_FORMAT_8BIT;
case 16: return AF_FORMAT_16BIT;
case 24: return AF_FORMAT_24BIT;
case 32: return AF_FORMAT_32BIT;
case 64: return AF_FORMAT_64BIT;
}
return 0;
}
int af_fmt_change_bits(int format, int bits)
{
audio: cleanup spdif format definitions Before this commit, there was AF_FORMAT_AC3 (the original spdif format, used for AC3 and DTS core), and AF_FORMAT_IEC61937 (used for AC3, DTS and DTS-HD), which was handled as some sort of superset for AF_FORMAT_AC3. There also was AF_FORMAT_MPEG2, which used IEC61937-framing, but still was handled as something "separate". Technically, all of them are pretty similar, but may use different bitrates. Since digital passthrough pretends to be PCM (just with special headers that wrap digital packets), this is easily detectable by the higher samplerate or higher number of channels, so I don't know why you'd need a separate "class" of sample formats (AF_FORMAT_AC3 vs. AF_FORMAT_IEC61937) to distinguish them. Actually, this whole thing is just a mess. Simplify this by handling all these formats the same way. AF_FORMAT_IS_IEC61937() now returns 1 for all spdif formats (even MP3). All AOs just accept all spdif formats now - whether that works or not is not really clear (seems inconsistent due to earlier attempts to make DTS-HD work). But on the other hand, enabling spdif requires manual user interaction, so it doesn't matter much if initialization fails in slightly less graceful ways if it can't work at all. At a later point, we will support passthrough with ao_pulse. It seems the PulseAudio API wants to know the codec type (or maybe not - feeding it DTS while telling it it's AC3 works), add separate formats for each codecs. While this reminds of the earlier chaos, it's stricter, and most code just uses AF_FORMAT_IS_IEC61937(). Also, modify AF_FORMAT_TYPE_MASK (renamed from AF_FORMAT_POINT_MASK) to include special formats, so that it always describes the fundamental sample format type. This also ensures valid AF formats are never 0 (this was probably broken in one of the earlier commits from today).
2014-09-23 20:44:54 +00:00
if (!af_fmt_is_valid(format))
return 0;
int mask = bits_to_mask(bits);
format = (format & ~AF_FORMAT_BITS_MASK) | mask;
return af_fmt_is_valid(format) ? format : 0;
}
static const int planar_formats[][2] = {
{AF_FORMAT_U8P, AF_FORMAT_U8},
{AF_FORMAT_S16P, AF_FORMAT_S16},
{AF_FORMAT_S32P, AF_FORMAT_S32},
{AF_FORMAT_FLOATP, AF_FORMAT_FLOAT},
{AF_FORMAT_DOUBLEP, AF_FORMAT_DOUBLE},
};
// Return the planar format corresponding to the given format.
// If the format is already planar, return it.
// Return 0 if there's no equivalent.
int af_fmt_to_planar(int format)
{
for (int n = 0; n < MP_ARRAY_SIZE(planar_formats); n++) {
if (planar_formats[n][1] == format)
return planar_formats[n][0];
if (planar_formats[n][0] == format)
return format;
}
return 0;
}
// Return the interleaved format corresponding to the given format.
// If the format is already interleaved, return it.
// Always succeeds if format is actually planar; otherwise return 0.
int af_fmt_from_planar(int format)
{
for (int n = 0; n < MP_ARRAY_SIZE(planar_formats); n++) {
if (planar_formats[n][0] == format)
return planar_formats[n][1];
}
return format;
}
// false for interleaved and AF_FORMAT_UNKNOWN
bool af_fmt_is_planar(int format)
{
return !!(format & AF_FORMAT_PLANAR);
}
const struct af_fmt_entry af_fmtstr_table[] = {
{"u8", AF_FORMAT_U8},
{"s8", AF_FORMAT_S8},
{"u16", AF_FORMAT_U16},
{"s16", AF_FORMAT_S16},
{"u24", AF_FORMAT_U24},
{"s24", AF_FORMAT_S24},
{"u32", AF_FORMAT_U32},
{"s32", AF_FORMAT_S32},
{"float", AF_FORMAT_FLOAT},
{"double", AF_FORMAT_DOUBLE},
{"u8p", AF_FORMAT_U8P},
{"s16p", AF_FORMAT_S16P},
{"s32p", AF_FORMAT_S32P},
{"floatp", AF_FORMAT_FLOATP},
{"doublep", AF_FORMAT_DOUBLEP},
audio: cleanup spdif format definitions Before this commit, there was AF_FORMAT_AC3 (the original spdif format, used for AC3 and DTS core), and AF_FORMAT_IEC61937 (used for AC3, DTS and DTS-HD), which was handled as some sort of superset for AF_FORMAT_AC3. There also was AF_FORMAT_MPEG2, which used IEC61937-framing, but still was handled as something "separate". Technically, all of them are pretty similar, but may use different bitrates. Since digital passthrough pretends to be PCM (just with special headers that wrap digital packets), this is easily detectable by the higher samplerate or higher number of channels, so I don't know why you'd need a separate "class" of sample formats (AF_FORMAT_AC3 vs. AF_FORMAT_IEC61937) to distinguish them. Actually, this whole thing is just a mess. Simplify this by handling all these formats the same way. AF_FORMAT_IS_IEC61937() now returns 1 for all spdif formats (even MP3). All AOs just accept all spdif formats now - whether that works or not is not really clear (seems inconsistent due to earlier attempts to make DTS-HD work). But on the other hand, enabling spdif requires manual user interaction, so it doesn't matter much if initialization fails in slightly less graceful ways if it can't work at all. At a later point, we will support passthrough with ao_pulse. It seems the PulseAudio API wants to know the codec type (or maybe not - feeding it DTS while telling it it's AC3 works), add separate formats for each codecs. While this reminds of the earlier chaos, it's stricter, and most code just uses AF_FORMAT_IS_IEC61937(). Also, modify AF_FORMAT_TYPE_MASK (renamed from AF_FORMAT_POINT_MASK) to include special formats, so that it always describes the fundamental sample format type. This also ensures valid AF formats are never 0 (this was probably broken in one of the earlier commits from today).
2014-09-23 20:44:54 +00:00
{"spdif-aac", AF_FORMAT_S_AAC},
{"spdif-ac3", AF_FORMAT_S_AC3},
{"spdif-dts", AF_FORMAT_S_DTS},
{"spdif-dtshd", AF_FORMAT_S_DTSHD},
{"spdif-eac3", AF_FORMAT_S_EAC3},
{"spdif-mp3", AF_FORMAT_S_MP3},
{"spdif-truehd",AF_FORMAT_S_TRUEHD},
{0}
};
bool af_fmt_is_valid(int format)
{
for (int i = 0; af_fmtstr_table[i].name; i++) {
if (af_fmtstr_table[i].format == format)
return true;
}
return false;
}
const char *af_fmt_to_str(int format)
{
2013-11-07 21:12:26 +00:00
for (int i = 0; af_fmtstr_table[i].name; i++) {
if (af_fmtstr_table[i].format == format)
return af_fmtstr_table[i].name;
}
return "??";
}
int af_fmt_seconds_to_bytes(int format, float seconds, int channels, int samplerate)
{
assert(!af_fmt_is_planar(format));
int bps = af_fmt2bps(format);
int framelen = channels * bps;
int bytes = seconds * bps * samplerate;
if (bytes % framelen)
bytes += framelen - (bytes % framelen);
return bytes;
}
int af_str2fmt_short(bstr str)
{
2013-11-07 21:12:26 +00:00
for (int i = 0; af_fmtstr_table[i].name; i++) {
if (!bstrcasecmp0(str, af_fmtstr_table[i].name))
return af_fmtstr_table[i].format;
2013-11-07 21:12:26 +00:00
}
return 0;
}
void af_fill_silence(void *dst, size_t bytes, int format)
{
bool us = (format & AF_FORMAT_SIGN_MASK) == AF_FORMAT_US;
memset(dst, us ? 0x80 : 0, bytes);
}
audio/format: add heuristic to estimate loss on format conversion The added function af_format_conversion_score() can be used to select the best sample format to convert to in order to reduce loss and extra conversion work. It calculates a "loss" score when going from one format to another, and for each conversion that needs to be done a certain score is subtracted. Thus, if you have to convert from one format to a set of other formats, you can calculate the score for each conversion, and pick the one with the highest score. Conversion between int and float is considered the worst case. One odd consequence is that when converting from s32 to u8 or float, u8 will be picked. Test program used to develop this follows: #define MAX_FMT 200 struct entry { const char *name; int score; }; static int compentry(const void *px1, const void *px2) { const struct entry *x1 = px1; const struct entry *x2 = px2; if (x1->score > x2->score) return 1; if (x1->score < x2->score) return -1; return 0; } int main(int argc, char *argv[]) { for (int n = 0; af_fmtstr_table[n].name; n++) { struct entry entry[MAX_FMT]; int entries = 0; for (int i = 0; af_fmtstr_table[i].name; i++) { assert(i < MAX_FMT); entry[entries].name = af_fmtstr_table[i].name; entry[entries].score = af_format_conversion_score(af_fmtstr_table[i].format, af_fmtstr_table[n].format); entries++; } qsort(&entry[0], entries, sizeof(entry[0]), compentry); for (int i = 0; i < entries; i++) { printf("%s -> %s: %d \n", af_fmtstr_table[n].name, entry[i].name, entry[i].score); } } }
2013-11-16 19:18:39 +00:00
#define FMT_DIFF(type, a, b) (((a) & type) - ((b) & type))
// Returns a "score" that serves as heuristic how lossy or hard a conversion is.
// If the formats are equal, 1024 is returned. If they are gravely incompatible
// (like s16<->ac3), INT_MIN is returned. If there is implied loss of precision
// (like s16->s8), a value <0 is returned.
int af_format_conversion_score(int dst_format, int src_format)
{
if (dst_format == AF_FORMAT_UNKNOWN || src_format == AF_FORMAT_UNKNOWN)
return INT_MIN;
if (dst_format == src_format)
return 1024;
// Can't be normally converted
if (AF_FORMAT_IS_SPECIAL(dst_format) || AF_FORMAT_IS_SPECIAL(src_format))
return INT_MIN;
int score = 1024;
if (FMT_DIFF(AF_FORMAT_INTERLEAVING_MASK, dst_format, src_format))
score -= 1; // has to (de-)planarize
if (FMT_DIFF(AF_FORMAT_SIGN_MASK, dst_format, src_format))
score -= 4; // has to swap sign
audio: cleanup spdif format definitions Before this commit, there was AF_FORMAT_AC3 (the original spdif format, used for AC3 and DTS core), and AF_FORMAT_IEC61937 (used for AC3, DTS and DTS-HD), which was handled as some sort of superset for AF_FORMAT_AC3. There also was AF_FORMAT_MPEG2, which used IEC61937-framing, but still was handled as something "separate". Technically, all of them are pretty similar, but may use different bitrates. Since digital passthrough pretends to be PCM (just with special headers that wrap digital packets), this is easily detectable by the higher samplerate or higher number of channels, so I don't know why you'd need a separate "class" of sample formats (AF_FORMAT_AC3 vs. AF_FORMAT_IEC61937) to distinguish them. Actually, this whole thing is just a mess. Simplify this by handling all these formats the same way. AF_FORMAT_IS_IEC61937() now returns 1 for all spdif formats (even MP3). All AOs just accept all spdif formats now - whether that works or not is not really clear (seems inconsistent due to earlier attempts to make DTS-HD work). But on the other hand, enabling spdif requires manual user interaction, so it doesn't matter much if initialization fails in slightly less graceful ways if it can't work at all. At a later point, we will support passthrough with ao_pulse. It seems the PulseAudio API wants to know the codec type (or maybe not - feeding it DTS while telling it it's AC3 works), add separate formats for each codecs. While this reminds of the earlier chaos, it's stricter, and most code just uses AF_FORMAT_IS_IEC61937(). Also, modify AF_FORMAT_TYPE_MASK (renamed from AF_FORMAT_POINT_MASK) to include special formats, so that it always describes the fundamental sample format type. This also ensures valid AF formats are never 0 (this was probably broken in one of the earlier commits from today).
2014-09-23 20:44:54 +00:00
if (FMT_DIFF(AF_FORMAT_TYPE_MASK, dst_format, src_format)) {
audio/format: add heuristic to estimate loss on format conversion The added function af_format_conversion_score() can be used to select the best sample format to convert to in order to reduce loss and extra conversion work. It calculates a "loss" score when going from one format to another, and for each conversion that needs to be done a certain score is subtracted. Thus, if you have to convert from one format to a set of other formats, you can calculate the score for each conversion, and pick the one with the highest score. Conversion between int and float is considered the worst case. One odd consequence is that when converting from s32 to u8 or float, u8 will be picked. Test program used to develop this follows: #define MAX_FMT 200 struct entry { const char *name; int score; }; static int compentry(const void *px1, const void *px2) { const struct entry *x1 = px1; const struct entry *x2 = px2; if (x1->score > x2->score) return 1; if (x1->score < x2->score) return -1; return 0; } int main(int argc, char *argv[]) { for (int n = 0; af_fmtstr_table[n].name; n++) { struct entry entry[MAX_FMT]; int entries = 0; for (int i = 0; af_fmtstr_table[i].name; i++) { assert(i < MAX_FMT); entry[entries].name = af_fmtstr_table[i].name; entry[entries].score = af_format_conversion_score(af_fmtstr_table[i].format, af_fmtstr_table[n].format); entries++; } qsort(&entry[0], entries, sizeof(entry[0]), compentry); for (int i = 0; i < entries; i++) { printf("%s -> %s: %d \n", af_fmtstr_table[n].name, entry[i].name, entry[i].score); } } }
2013-11-16 19:18:39 +00:00
int dst_bits = dst_format & AF_FORMAT_BITS_MASK;
audio: cleanup spdif format definitions Before this commit, there was AF_FORMAT_AC3 (the original spdif format, used for AC3 and DTS core), and AF_FORMAT_IEC61937 (used for AC3, DTS and DTS-HD), which was handled as some sort of superset for AF_FORMAT_AC3. There also was AF_FORMAT_MPEG2, which used IEC61937-framing, but still was handled as something "separate". Technically, all of them are pretty similar, but may use different bitrates. Since digital passthrough pretends to be PCM (just with special headers that wrap digital packets), this is easily detectable by the higher samplerate or higher number of channels, so I don't know why you'd need a separate "class" of sample formats (AF_FORMAT_AC3 vs. AF_FORMAT_IEC61937) to distinguish them. Actually, this whole thing is just a mess. Simplify this by handling all these formats the same way. AF_FORMAT_IS_IEC61937() now returns 1 for all spdif formats (even MP3). All AOs just accept all spdif formats now - whether that works or not is not really clear (seems inconsistent due to earlier attempts to make DTS-HD work). But on the other hand, enabling spdif requires manual user interaction, so it doesn't matter much if initialization fails in slightly less graceful ways if it can't work at all. At a later point, we will support passthrough with ao_pulse. It seems the PulseAudio API wants to know the codec type (or maybe not - feeding it DTS while telling it it's AC3 works), add separate formats for each codecs. While this reminds of the earlier chaos, it's stricter, and most code just uses AF_FORMAT_IS_IEC61937(). Also, modify AF_FORMAT_TYPE_MASK (renamed from AF_FORMAT_POINT_MASK) to include special formats, so that it always describes the fundamental sample format type. This also ensures valid AF formats are never 0 (this was probably broken in one of the earlier commits from today).
2014-09-23 20:44:54 +00:00
if ((dst_format & AF_FORMAT_TYPE_MASK) == AF_FORMAT_F) {
audio/format: add heuristic to estimate loss on format conversion The added function af_format_conversion_score() can be used to select the best sample format to convert to in order to reduce loss and extra conversion work. It calculates a "loss" score when going from one format to another, and for each conversion that needs to be done a certain score is subtracted. Thus, if you have to convert from one format to a set of other formats, you can calculate the score for each conversion, and pick the one with the highest score. Conversion between int and float is considered the worst case. One odd consequence is that when converting from s32 to u8 or float, u8 will be picked. Test program used to develop this follows: #define MAX_FMT 200 struct entry { const char *name; int score; }; static int compentry(const void *px1, const void *px2) { const struct entry *x1 = px1; const struct entry *x2 = px2; if (x1->score > x2->score) return 1; if (x1->score < x2->score) return -1; return 0; } int main(int argc, char *argv[]) { for (int n = 0; af_fmtstr_table[n].name; n++) { struct entry entry[MAX_FMT]; int entries = 0; for (int i = 0; af_fmtstr_table[i].name; i++) { assert(i < MAX_FMT); entry[entries].name = af_fmtstr_table[i].name; entry[entries].score = af_format_conversion_score(af_fmtstr_table[i].format, af_fmtstr_table[n].format); entries++; } qsort(&entry[0], entries, sizeof(entry[0]), compentry); for (int i = 0; i < entries; i++) { printf("%s -> %s: %d \n", af_fmtstr_table[n].name, entry[i].name, entry[i].score); } } }
2013-11-16 19:18:39 +00:00
// For int->float, always prefer 32 bit float.
score -= dst_bits == AF_FORMAT_32BIT ? 8 : 0;
} else {
// For float->int, always prefer highest bit depth int
score -= 8 * (AF_FORMAT_64BIT - dst_bits);
}
} else {
int bits = FMT_DIFF(AF_FORMAT_BITS_MASK, dst_format, src_format);
if (bits > 0) {
score -= 8 * bits; // has to add padding
} else if (bits < 0) {
score -= 1024 - 8 * bits; // has to reduce bit depth
}
}
// Consider this the worst case.
audio: cleanup spdif format definitions Before this commit, there was AF_FORMAT_AC3 (the original spdif format, used for AC3 and DTS core), and AF_FORMAT_IEC61937 (used for AC3, DTS and DTS-HD), which was handled as some sort of superset for AF_FORMAT_AC3. There also was AF_FORMAT_MPEG2, which used IEC61937-framing, but still was handled as something "separate". Technically, all of them are pretty similar, but may use different bitrates. Since digital passthrough pretends to be PCM (just with special headers that wrap digital packets), this is easily detectable by the higher samplerate or higher number of channels, so I don't know why you'd need a separate "class" of sample formats (AF_FORMAT_AC3 vs. AF_FORMAT_IEC61937) to distinguish them. Actually, this whole thing is just a mess. Simplify this by handling all these formats the same way. AF_FORMAT_IS_IEC61937() now returns 1 for all spdif formats (even MP3). All AOs just accept all spdif formats now - whether that works or not is not really clear (seems inconsistent due to earlier attempts to make DTS-HD work). But on the other hand, enabling spdif requires manual user interaction, so it doesn't matter much if initialization fails in slightly less graceful ways if it can't work at all. At a later point, we will support passthrough with ao_pulse. It seems the PulseAudio API wants to know the codec type (or maybe not - feeding it DTS while telling it it's AC3 works), add separate formats for each codecs. While this reminds of the earlier chaos, it's stricter, and most code just uses AF_FORMAT_IS_IEC61937(). Also, modify AF_FORMAT_TYPE_MASK (renamed from AF_FORMAT_POINT_MASK) to include special formats, so that it always describes the fundamental sample format type. This also ensures valid AF formats are never 0 (this was probably broken in one of the earlier commits from today).
2014-09-23 20:44:54 +00:00
if (FMT_DIFF(AF_FORMAT_TYPE_MASK, dst_format, src_format))
audio/format: add heuristic to estimate loss on format conversion The added function af_format_conversion_score() can be used to select the best sample format to convert to in order to reduce loss and extra conversion work. It calculates a "loss" score when going from one format to another, and for each conversion that needs to be done a certain score is subtracted. Thus, if you have to convert from one format to a set of other formats, you can calculate the score for each conversion, and pick the one with the highest score. Conversion between int and float is considered the worst case. One odd consequence is that when converting from s32 to u8 or float, u8 will be picked. Test program used to develop this follows: #define MAX_FMT 200 struct entry { const char *name; int score; }; static int compentry(const void *px1, const void *px2) { const struct entry *x1 = px1; const struct entry *x2 = px2; if (x1->score > x2->score) return 1; if (x1->score < x2->score) return -1; return 0; } int main(int argc, char *argv[]) { for (int n = 0; af_fmtstr_table[n].name; n++) { struct entry entry[MAX_FMT]; int entries = 0; for (int i = 0; af_fmtstr_table[i].name; i++) { assert(i < MAX_FMT); entry[entries].name = af_fmtstr_table[i].name; entry[entries].score = af_format_conversion_score(af_fmtstr_table[i].format, af_fmtstr_table[n].format); entries++; } qsort(&entry[0], entries, sizeof(entry[0]), compentry); for (int i = 0; i < entries; i++) { printf("%s -> %s: %d \n", af_fmtstr_table[n].name, entry[i].name, entry[i].score); } } }
2013-11-16 19:18:39 +00:00
score -= 2048; // has to convert float<->int
return score;
}