Jump to content

Comparison of audio coding formats

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by MihaiFunny125 (talk | contribs) at 14:37, 12 May 2024 (Added Windows Media Voice. Also: done other fixes.). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Technical details

Lossy formats
Audio compression format Algorithm Sample rate Bit rate Latency CBR VBR Stereo Multichannel
AAC MDCT, Hybrid Subband (AAC-HE) 8–192 kHz[1] 8–529 kbit/s (stereo) 20–405 ms[2] Yes Yes Yes: Dual, Mid/Side, Intensity, Parametric Yes: Up to 48 channels
AC3 MDCT 32, 44.1, 48 kHz 32–640 kbit/s 32–48 ms Yes In theory, it supports VBR, but no high-quality encoder supports that feature. Most high-quality encoders only do CBR encoding Yes Yes: Up to 6 channels
AC4 MDCT 48 kHz 24-1536 kbit/s ? Yes No Yes Yes: Up to 24 channels
AMBE ? 8 kHz 2, 9.6 kbit/s ? ? ? ? ?
AMR ACELP 8 kHz 4.75, 5.15, 5.90, 6.70, 7.40, 7.95, 10.20, 12.20 kbit/s 25 ms Yes Yes No No
AMR-WB
(G.722.2) 		ACELP 16 kHz 6.60, 8.85, 12.65, 14.25, 15.85, 18.25, 19.85, 23.05, 23.85 kbit/s 25 ms Yes Yes Yes: only in MPEG-4 Part 12 container No
AMR-WB+ ACELP 8, 11.025, 16, 22.05, 32, 44.1, 48 kHz 6–36 kbit/s (mono)
7–48 kbit/s (stereo) 		60–90 ms Yes Yes Yes No
aptX Subband ADPCM 24–48 kHz (stereo) 192–384 kbit/s (stereo) 2 ms Yes (4:1) No Yes Yes: Up to 8 channels
ATRAC1 MDCT-Hybrid Subband 44.1 kHz 292 kbit/s >100 ms Yes No Yes: Dual Only No
ATRAC3 MDCT-Hybrid Subband 44.1 kHz 66, 105, 132, 146, 176, 264, 352 kbit/s >100 ms Yes No Yes: Dual (LP2), Mid/Side (LP4) No
ATRAC3plus MDCT-Hybrid Subband 44.1, 48 kHz 32–768 kbit/s >100 ms Yes No Yes Yes: Up to 8 channels
ATRAC9 MDCT-Hybrid Subband 12, 24, 48 kHz 36-672 kbit/s >100 ms Yes No Yes Yes: Up to 8 channels
BroadVoice (BV16, BV32) Two-Stage Noise Feedback Coding (TSNFC) 8, 16 kHz 16, 32 kbit/s[3] 5 ms Yes No No No
Codec2 Speech 8 kHz 0.7, 1.2, 1.3, 1.4, 1.6, 2.4, 3.2 kbit/s 20–40 ms Yes No No No
E-AC3 MDCT 32, 44.1, 48 kHz 32–6144 kbit/s 5.33–48 ms Yes No Yes Yes: Up to 15 channels
E-aptX Subband ADPCM 15–48 kHz 60, 384, 767, 1024 kbit/s, 1.28 Mbit/s ? Yes (4:1) No Yes Yes: Up to 8 channels
Encodec Neural networks 24 kHz (Mono)

48 kHz (Stereo)

1.5 (Mono only), 3, 6, 12, 24 kbit/s ? Yes No Yes No
EVS Speech 8, 16, 32, 48 kHz 5.9, 7.2, 8, 9.6, 13.2, 16.4, 24.4, 32, 48, 64, 96, 128 kbit/s + AMR-WB bitrate modes 20 ms Yes Only 5.9 kbit/s with DTX on Yes No
GSM-HR VSELP 8 kHz 5.6 kbit/s 25 ms Yes No No No
GSM-FR RPE-LTP 8 kHz 13 kbit/s 20–30 ms Yes No No No
GSM-EFR ACELP 8 kHz 12.2 kbit/s 20–30 ms Yes No No No
HE-AAC ? 22 ~ 96 kHz 16 ~ 80 kbit/s ~130 ms[4] Yes Yes Yes: Dual, Mid/Side, Intensity, Parametric Yes: Up to 48 channels
HVXC Speech 8 kHz 2, 4 kbit/s 36 ms Yes Yes No No
iLBC Block Independent LPC 8 kHz 13.33, 15.20 kbit/s 25, 40 ms Yes No No No
iSAC Transform coding 16, 32 kHz 10–52 kbit/s 33, 63 ms Yes Yes No No
LC3 Speech 8, 16, 24, 32, 48 kHz 16-320 kbit/s 7.5-10 ms Yes No Yes No
LDAC Subband ADPCM 44.1, 48, 88.2, 96, 176.4, 192 kHz 303, 606, 909 kbit/s (44.1/88.2/176.4 kHz)

330, 660, 990 kbit/s (48/96/192 kHz)

? Yes No Yes No
LHDC Subband ADPCM 44.1, 48, 96 kHz 400, 560, 900 kbit/s ? Yes No Yes No
L2HC Subband ADPCM 44.1, 48, 96, 192 kHz 400, 560, 960, 1920 kbit/s ? Yes No Yes ?
Lyra Speech 8, 16, 32, 48 kHz 3.2, 6, 9.2 kbit/s 20 ms Yes No No No
MP3 (MPEG-1, 2, 2.5 Audio Layer III) MDCT, Hybrid Subband 8, 11.025, 12, 16, 22.05, 24, 32, 44.1, 48 kHz 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144, 160, 192, 224, 256, 288, 320 kbit/s >100 ms Yes Yes Yes: Dual, Mid/Side, Intensity No
MPEG-1 Audio Layer II (MP2) Subband 32, 44.1, 48 kHz 32, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256, 320, 384 kbit/s Yes Yes: but decoders are not required to support it Yes: Dual, Intensity No
MPEG-2 Audio Layer II (MP2) Subband 16, 22.05, 24 kHz[5][6] 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144, 160 kbit/s[5] Yes Yes: but decoders are not required to support it Yes Yes: up to 5 full range audio channels and an LFE-channel with MPEG Multichannel
Musepack Subband 32, 37.8, 44.1, 48 kHz 20–350 kbit/s ? No Yes Yes Yes: Up to 8 channels
Opus MDCT, LPC, LTP 8–48 kHz 6–510 kbit/s 5–66.5 ms Yes Yes Yes Yes: Up to 255 channels[7]
RealAudio MDCT Varies (see article) Varies (see article) Varies Yes Yes Yes Yes: Up to 6 channels
Satin Speech 8, 16, 32, 48 kHz 6-36 kbit/s ? Yes No Yes No
SILK LPC, LTP 8, 12, 16, 24 kHz 6–40 kbit/s 25 ms Yes ? ? ?
Siren 7 Derived from PT716plus, MLT 16 kHz 16, 24, 32 kbit/s 40 ms Yes No No No
Siren 14 MLT 32 kHz 24, 32, 48 kbit/s (mono)
48, 64, 96 kbit/s (stereo) 		40 ms Yes No Yes No
Siren 22 MLT 48 kHz 32, 48, 64 kbit/s (mono)
64, 96, 128 kbit/s (stereo) 		40 ms Yes No Yes No
SNAC Neural networks 24, 32, 44.1 kHz 0.98 (24 kHz), 1.9 (32 kHz), 2.6 (44.1 kHz) kbit/s ? Yes No No No
Speex CELP 8, 16, 32, (48) kHz 2.15–24.6 kbit/s (NB)
4–44.2 kbit/s (WB) 		30 ms (NB)
34 ms (WB) 		Yes Yes Yes: Intensity Yes
TSAC Neural networks, modified version of Descript Audio Codec, extended for stereo with a transformer model to shrink even more while keeping the quality high. 44.1 kHz (Mono and Stereo). Other sample rates: Uncertain. ~0.45-5.5 kbps (mono)

~0.6-7.5 kbps (stereo)

? Yes No Yes No
USAC ? 7.35, 8, 11.025, 12, 16, 19.2, 22.05, 24, 32, 38.4, 44.1, 48, 57.6, 64, 88.2, 96 kHz 6-128 kbit/s (mono)

12-320 kbit/s (stereo)

? Yes Yes Yes No
VMR-WB ACELP 16 kHz 8.55, 4.0, 0.8, 13.3, 6.2, 2.7, 1.0 kbit/s 33.75 ms Yes Yes ? ?
Vorbis (Ogg) MDCT 8–192 kHz 45-500 kbit/s (32-500 kbit/s for aoTuV tunings) >100 ms Yes/ABR Yes Yes: Dual, Lossless, Phase, Point (Intensity) Yes: Up to 255 channels
WavPack Lossy Prediction, Quantization 1 Hz to 16.777216 MHz 196 kbit/s and up in lossy mode (for CD audio) 3523.8 ms Yes Yes Yes Yes: Up to 256 channels
Windows Media Audio Standard MDCT 8, 11.025, 16, 22.05, 32, 44.1, 48 kHz 8–768 kbit/s >100 ms Yes Yes Yes It can be hacked to support 5.1/7.1 support, but officially it doesn't support multi-channel
Windows Media Audio Pro MDCT 8, 11.025, 16, 22.05, 32, 44.1, 48, 88.2, 96 kHz (8-22.05kHz not avabile in encoders that uses DirectShow like Switch by NCH Software or Windows Media Encoder) 4–768 kbit/s >100 ms Yes Yes Yes Yes: At least 8 channels, expandable
Windows Media Audio Voice ACELP? 8, 11.025, 16, 22.05 kHz officially (can be hacked to support higher sample rates) 4-20 kbit/s officially (can be hacked to support higher bitrates) ? Yes No It can be hacked to support stereo support, but officially it doesn't support stereo It can be hacked to support 5.1/7.1 support, but officially it doesn't support multi-channel
Lossless formats
Audio compression format Algorithm Sample rate Bits per sample Latency Stereo Multichannel
ALAC Lossless 1–384 kHz 16, 20, 24, 32[8] ? Yes Yes: Up to 8 channels
ALS Lossless 0–4 GHz 8, 16, 24, 32 (int or float) ? Yes Yes: Up to 65536 channels
ATRAC Advanced Lossless Lossless 44.1 kHz 16 ? Yes No
Dolby TrueHD Lossless 48, 96, 192 kHz 16, 20, 24 ? Yes Yes: 5.1, 7.1, and Dolby Atmos object-based audio[9]
DTS-HD Master Audio Lossless 48, 96, 192 kHz 16, 24 ? Yes Yes: Up to 7.1[10]
FLAC Lossless 1–1.048575 MHz (>655.350 kHz need not be streamable)[11] 4–32 4.3–92 ms (46.4 ms typical) Yes Yes: Up to 8 channels
HD-AAC ? ? ? ? ? ?
LA Lossless 0–4 GiHz 16 ? Yes No
L2HC Lossless 44.1kHz 16 ? Yes No
Monkey's Audio Lossless 1–655.350 kHz[citation needed] 8, 16, 24, 32 and 32-bit float 1670 to 26,749 ms (varies with compression) Yes Yes: Up to 32 channels since version 5.50[12]
OptimFROG Lossless 0–4 GiHz[13] 8, 16, 24, 32 (int or float) ? Yes No
OSQ Lossless 6, 8, 11.025, 16, 22.05, 32, 44.1, 48, 64, 88.2, 96 kHz 8, 16, 20, 24 ? Yes No
RealAudio Lossless Lossless 8, 11.025, 16, 22.05, 32, 44.1 kHz 16 Varies Yes No
Sac Lossless 1–48 kHz 16 ? Yes No
TAK Lossless 8–192 kHz 8, 16, 24 ? Yes Yes: Up to 6 channels
True Audio (TTA) Lossless 0–4 GiHz 8, 16, 24 approx. 1045 ms[14] Yes Yes: Up to 16 (with ffmpeg), but no channel allocation
WavPack Lossless Lossless, Hybrid 1 Hz to 1 GiHz Up to 32 (and 32-bit float), and 1-bit DSD. ? Yes Yes: Up to 256 channels
Windows Media Audio Lossless Lossless 8, 11.025, 16, 22.05, 32, 44.1, 48, 88.2, 96 kHz 16, 24 >100 ms Yes Yes: Up to 6 channels
ITU-T formats
Audio compression format Algorithm Sample rate Bit rate Bits per sample Latency CBR VBR Stereo Multichannel
G.711 companding A-law or μ-law, PCM 8 kHz 64 kbit/s 8 bit (log) 125 μs (typical) Yes No No No
G.711.0 Lossless compression of G.711 8 kHz 0.2–65.6 kbit/s 8 bit (log) 5–40 ms No Yes No No
G.711.1 MDCT, A-law, μ-law 8, 16 kHz 64, 80, 96 kbit/s 16 bit 11.875 ms Yes Yes No No
G.718 CELP, MDCT, Lossy 8, 16 kHz 8, 12, 12.65, 16, 24, 32 kbit/s 16 bit 42.875–43.875 ms Yes Yes No No
G.718B CELP, MDCT, Lossy, Sinusoidal Coding 32 kHz 36, 40, 48 kbit/s 16 bit 49.625 ms Yes Yes No No
G.719 MDCT,[15] Lossy (incorporates elements of Siren Codec and Ericsson technology) 48 kHz 32–88 kbit/s in 4 kbit/s steps, 88–128 kbit/s in 8 kbit/s steps 16 bit 40 ms Yes No Yes: only in MPEG-4 Part 12 container Yes: only in MPEG-4 Part 12 container
G.721 ADPCM, Lossy 8 kHz 32 kbit/s 13 bit Yes No No No
G.722 sub-band ADPCM, Lossy 16 kHz 64 kbit/s
(comprises 48, 56 or 64 kbit/s audio and 16, 8 or 0 kbit/s auxiliary data) 		14 bit 4 ms Yes No No No
G.722.1 Modulated Lapped Transform (MDCT), Lossy (based on Siren Codec) 16 kHz 24, 32 kbit/s 16 bit 40 ms Yes No No No
G.722.1C Modulated Lapped Transform (MDCT), Lossy (based on Siren Codec) 32 kHz 24, 32, 48 kbit/s 16 bit 40 ms Yes No No No
G.722.2 (AMR-WB) multi-rate wideband ACELP, Lossy 16 kHz 6.60, 8.85, 12.65, 14.25, 15.85, 18.25, 19.85, 23.05, 23.85 kbit/s 14 bit 25 ms Yes Yes Yes: only in MPEG-4 Part 12 container No
G.723 ADPCM, Lossy 8 kHz 24, 40 kbit/s 13 bit Yes No No No
G.723.1 MP-MLQ, ACELP, Lossy 8 kHz 5.3, 6.3 kbit/s 13 bit 37.5 ms Yes No No No
G.726 ADPCM, Lossy 8 kHz 16, 24, 32, 40 kbit/s 13 bit 125 μs Yes No No No
G.727 ADPCM, Lossy 8 kHz 16, 24, 32, 40 kbit/s 13 bit Yes Yes No No
G.728 low-delay CELP, Lossy 8 kHz 16 kbit/s 13 bit 0.625 ms Yes No No No
G.729 CS-ACELP, Lossy 8 kHz 8 kbit/s 13 bit 15 ms Yes No No No
G.729D CS-ACELP, Lossy 8 kHz 6.4 kbit/s 13 bit Yes No No No
G.729E CS-ACELP, Lossy 8 kHz 11.8 kbit/s 13 bit 15 ms Yes No No No
G.729.1 CELP, TDBWE, TDAC (MDCT), Lossy 8, 16 kHz 8 kbit/s, 12–32 kbit/s in 2 kbit/s steps 16 bit 48.9375 ms Yes Yes No No

Notes

  • The latency listed here is the total delay (frame size, plus all lookahead) at the normal operating sample rate (typically 44.1 kHz).
  • Lossless compression will have a variable bit rate.
  1. ^ MPEG 2 AAC was limited to a 96 kHz sampling rate, however, with MPEG 4 AAC, a later version part of the MPEG 4 specification, the maximum sampling rate has been increased to 192 kHz.
  2. ^ "Algorithmic delay and synchronization in MPEG audio codecs - EE Times". EETimes. Retrieved 6 May 2017.
  3. ^ "Support Documents and Downloads". www.broadcom.com. Retrieved 6 May 2017.
  4. ^ Schnell, Markus; Schmidt, Markus; Jander, Manuel; Albert, Tobias; Geiger, Ralf; Ruoppila, Vesa; Ekstrand, Per; Bernhard, Grill (2008-10-01). "MPEG-4 Enhanced Low Delay AAC - A New Standard for High Quality Communication". Audio Engineering Society. Available from fraunhofer.de
  5. ^ a b Predrag Supurovic (September 1998). "MPEG Audio Frame Header". Retrieved 2009-07-11.
  6. ^ ISO/IEC JTC1/SC29/WG11 - Audio Subgroup. "MPEG Audio FAQ Version 9, MPEG-1 and MPEG-2 BC". Retrieved 2010-02-27.{{cite web}}: CS1 maint: numeric names: authors list (link)
  7. ^ "Opus Codec". Retrieved 2014-07-02.
  8. ^ "macosforge/alac". GitHub. Retrieved 6 May 2017.
  9. ^ "Dolby Atmos for Home Theaters". Dolby.com. Archived from the original on 20 May 2018. Retrieved 4 April 2018.
  10. ^ "DTS-HD and DTS-HD Master Audio: Enabling HD Audio Across an Evolving Media Delivery Landscape", DTS, Inc., October 2012
  11. ^ sample rates from 1 up to 1048575 hertz and bit depths from 4 up to 32 bits [1]
  12. ^ "Monkey's Audio - a fast and powerful lossless audio compressor".
  13. ^ 4 bytes - sample rate
  14. ^ "Tau Projects". en.true-audio.com. Archived from the original on 27 October 2015. Retrieved 6 May 2017.
  15. ^ Britanak, Vladimir; Rao, K. R. (2017). Cosine-/Sine-Modulated Filter Banks: General Properties, Fast Algorithms and Integer Approximations. Springer. pp. 31, 478. ISBN 9783319610801.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Comparison_of_audio_coding_formats&oldid=1223495629"