Compressing sound files

CD-quality, uncompressed stereo audio files require about 150 kilobytes (KB) of hard disk space for every second of sound. A sound file's size depends on the sample rate (frequency), the number of bits per sample (quality or richness of the sound), and whether the sound is mono or stereo. For example, a 22 kilohertz (kHz), 16-bit mono music sample requires a data rate of 2.65 megabytes (MB) per minute to play back as it was originally recorded. Compressing a sound file using an audio codec is the best way to store audio files on your computer's hard disk or play audio files over a network with limited bandwidth.

Audio codecs exist for various audio uses. Some audio codecs are optimized for voice, while others are better suited for low to high bit rate music samples and sounds. Voice-oriented audio codecs (such as DSP Group TrueSpeech or Microsoft Groupe Spécial Mobile [GSM] 6.10) allow efficient compression of voice data. Music-oriented audio codecs (such as Fraunhofer Institut Integrierte Schaltungen IIS [FhG] MPEG Layer-3 or Voxware MetaSound) allow close to CD-quality sound to be compressed to about one-fourth its original size.

Using Sound Recorder, you can select either a voice-oriented or music-oriented audio codec to compress and decompress .wav files. If you compress the .wav file into one of the formats supported by Windows, you can play the sound back directly without decompressing it manually. Windows uses its codecs to decompress the data for playback.

Common audio codecs supported by Windows

Audio CodecDescriptionFrequency/Sampling Rate


The Microsoft International Telecommunications Union (ITU) CCITT A-Law codec provides compatibility with Telephony Application Programming Interface (TAPI) standards for Europe. Also known as G.711, this codec is supported by many hardware configurations and offers a 2:1 compression ratio (from 16 bits to 8 bits per sample).

8.000 kHz, 8-bit, Mono, 7 kbps
8.000 kHz, 8-bit, Stereo, 15 kbps
11.025 kHz, 8-bit, Mono, 10 kbps
11.025 kHz, 8-bit, Stereo, 21 kbps
22.050 kHz, 8-bit, Mono, 21 kbps
22.050 kHz, 8-bit, Stereo, 43 kbps
44.100 kHz, 8-bit, Mono, 43 kbps
44.100 kHz, 8-bit, Stereo, 86 kbps


Same as the CCITT A-Law codec, but provides compatibility with TAPI standards used in North America.

Same as the CCITT A-Law codec.

DSP Group TrueSpeech

The DSP Group TrueSpeech codec provides good compression for voice-oriented sound and is designed for low to mid bit rate voice-oriented sound. TrueSpeech offers a better data rate than GSM 6.10, the other voice-oriented audio codec provided with Windows. TrueSpeech is a good choice when you want to record notes in documents or spreadsheets or store voice mail on your computer, but is not designed for sound other than voice. TrueSpeech does not offer real-time compression rates, but offers real-time decompression rates, making this codec a good alternative for use with modems and networks.

8.000 kHz, 1-bit, Mono, 1 kbps

GSM 6.10

The Microsoft Groupe Spécial Mobile (GSM) codec is designed for the efficient compression of speech and is best for mid to high bit rate voice-oriented sound. GSM provides 2:1, real-time compression (as long as your hardware is fast enough to support it), making this codec a good option for recording voice with Sound Recorder. Using GSM to compress music can result in poor audio quality. GSM conforms to the European Telecommunications Standards Institute recommendation 6.10.

8.000 kHz, Mono, 1 kbps
11.025 kHz, Mono, 2 kbps
22.050 kHz, Mono, 4 kbps
44.100 kHz, Mono, 8 kbps


The Microsoft Interactive Multimedia Association (IMA) ADPCM codec is designed for multiple hardware platforms and offers high quality, real-time compression with high bit rate content. IMA ADPCM is similar to Microsoft ADPCM, but provides faster 4:1 compression.

8.000 kHz, 4-bit, Mono, 3 kbps
8.000 kHz, 4-bit, Stereo, 7 kbps
11.025 kHz, 4-bit, Mono, 5 kbps
11.025 kHz, 4-bit, Stereo, 10 kbps
22.050 kHz, 4-bit, Mono, 10 kbps
22.050 kHz, 4-bit, Stereo, 21 kbps
44.100 kHz, 4-bit, Mono, 21 kbps
44.100 kHz, 4-bit, Stereo, 43 kbps

Lernout & Hauspie CELP
4.8 kbps

The Lernout & Hauspie CELP codec is designed for use with speech and is best for low bit rate voice-oriented sound. Using Lernout & Hauspie CELP to compress music can result in poor audio quality. This codec has only one setting, which compresses everything to 4.8 kbps. Audio compressed this small doesn't have the richness associated with CD-quality audio, however voice audio, such as someone talking, will sound good because Lernout & Hauspie CELP is designed to work best with less dynamic audio sources. Lernout & Hauspie CELP is useful when you want to have a lot of bandwidth left for sending other types of information across the network such as video, images, or script commands.

8.000 kHz, 16-bit, Mono, 0 kbps

Lernout & Hauspie SBC
8 kbps

Similar to the Lernout & Hauspie CELP 4.8 kbps codec.

8.000 kHz, 16-bit, Mono, 1 kbps

Lernout & Hauspie SBC
12 kbps

Similar to the Lernout & Hauspie CELP 4.8 kbps codec.

8.000 kHz, 16-bit, Mono, 2 kbps

Lernout & Hauspie SBC
16 kbps

Similar to the Lernout & Hauspie CELP 4.8 kbps codec.

8.000 kHz, 16-bit, Mono, 2 kbps

Microsoft ADPCM

The Microsoft Adaptive Delta Pulse Code Modulation (ADPCM) codec provides high-quality 4:1 for compression, in both real-time and otherwise, and is best for audio streams associated with high bit rate video.

8.000 kHz, 4-bit, Mono, 4 kbps
8.000 kHz, 4-bit, Stereo, 8 kbps
11.025 kHz, 4-bit, Mono, 5 kbps
11.025 kHz, 4-bit, Stereo, 11 kbps
22.050 kHz, 4-bit, Mono, 10 kbps
22.050 kHz, 4 Bit, Stereo, 21 kbps
44.100 kHz, 4-bit, Mono, 21 kbps
44.100 kHz, 4-bit, Stereo, 43 kbps

Microsoft G.723.1

The Microsoft G.723.1 codec is designed for creating active streaming format (.asf) files or ASF audio streams for use over the Internet or with Windows Media Player. The G.723.1 standard specifies the format and algorithm for sending or receiving voice over low bandwidth network connections. Files compressed using Microsoft G.723.1 can contain multiple data types in addition to audio-only, and supports many compression schemes. Microsoft G.723.1 works extremely well on low bit rate network connections, such as 14.4 or 28.8 kbps.

8.000 kHz, Mono, 6400 bit/s 0 kbps
8.000 kHz, Mono, 5333 bit/s 0 kbps

MPEG Layer-3

The Fraunhofer Institut Integrierte Schaltungen IIS (FhG) MPEG Layer-3 codec is designed for creating low-to-mid bit rate CD-quality music files for use on an intranet or the Internet. The MPEG Layer-3 codec offers excellent high-fidelity compression for many different types of audio and is a good choice when you require a high bit rate and low CPU usage.

8 kbps, 8.000 kHz, Mono, 0 kbps
16 kbps, 8.000 kHz, Mono, 1 kbps
8 kbps, 11.025 kHz, Mono, 0 kbps
16 kbps, 11.025 kHz, Mono, 1 kbps
18 kbps, 11.025 kHz, Mono, 2 kbps
20 kbps, 11.025 kHz, Mono, 2 kbps
24 kbps, 11.025 kHz, Mono, 2 kbps
32 kbps, 11.025 kHz, Mono, 3 kbps
8 kbps, 12.000 kHz, Mono, 0 kbps
16 kbps, 12.000 kHz, Mono, 1 kbps
18 kbps, 12.000 kHz, Mono, 2 kbps
20 kbps, 12.000 kHz, Mono, 2 kbps
16 kbps, 16.000 kHz, Mono, 1 kbps
18 kbps, 16.000 kHz, Mono, 2 kbps
20 kbps, 16.000 kHz, Mono, 2 kbps
24 kbps, 16.000 kHz, Mono, 2 kbps
32 kbps, 16.000 kHz, Mono, 3 kbps
24 kbps, 22.050 kHz, Mono, 2 kbps
32 kbps, 22.050 kHz, Mono, 3 kbps
24 kbps, 24.000 kHz, Mono, 2 kbps
32 kbps, 24.000 kHz, Mono, 3 kbps
18 kbps, 8.000 kHz, Stereo, 2 kbps
20 kbps, 8.000 kHz, Stereo, 2 kbps
24 kbps, 8.000 kHz, Stereo, 2 kbps
32 kbps, 8.000 kHz, Stereo, 3 kbps
18 kbps, 11.025 kHz, Stereo, 2 kbps
20 kbps, 11.025 kHz, Stereo, 2 kbps
24 kbps, 11.025 kHz, Stereo, 2 kbps
32 kbps, 11.025 kHz, Stereo, 3 kbps
18 kbps, 12.000 kHz, Stereo, 2 kbps
20 kbps, 12.000 kHz, Stereo, 2 kbps
24 kbps, 12.000 kHz, Stereo, 2 kbps
32 kbps, 12.000 kHz, Stereo, 3 kbps
32 kbps, 16.000 kHz, Stereo, 3 kbps
48 kbps, 16.000 kHz, Stereo, 5 kbps
56 kbps, 16.000 kHz, Stereo, 6 kbps
48 kbps, 22.050 kHz, Stereo, 5 kbps
56 kbps, 22.050 kHz, Stereo, 6 kbps
48 kbps, 24.000 kHz, Stereo, 5 kbps
56 kbps, 24.000 kHz, Stereo, 6 kbps


The Microsoft Pulse-Code Modulation (PCM) Converter provides uncompressed audio for higher bit rate content. With PCM, you can play an audio sample at one kilohertz rate on a sound card that supports another rate. For example, 8-bit cards can play 16-bit samples by reducing the quality of the audio to match the sound card's capabilities.

8.000 kHz, 8-bit, Mono, 7 kbps
8.000 kHz, 8-bit, Stereo, 15 kbps
8.000 kHz, 16-bit, Mono, 15 kbps
8.000 kHz, 16-bit, Stereo, 31 kbps
11.025 kHz, 8-bit, Mono, 10 kbps
11.025 kHz, 8-bit, Stereo, 21 kbps
11.025 kHz, 16-bit, Mono, 21 kbps
11.025 kHz, 16-bit, Stereo, 43 kbps
22.050 kHz, 8-bit, Mono, 21 kbps
22.050 kHz, 8-bit, Stereo, 43 kbps
22.050 kHz, 16-bit, Mono, 43 kbps
22.050 kHz, 16-bit, Stereo, 86 kbps
44.100 kHz, 8-bit, Mono, 43 kbps
44.100 kHz, 8-bit, Stereo, 86 kbps
44.100 kHz, 16-bit, Mono, 86 kbps
44.100 kHz, 16-bit, Stereo, 172 kbps
48.000 kHz, 8-bit, Mono, 46 kbps
48.000 kHz, 8-bit, Stereo, 93 kbps
48.000 kHz, 16-bit, Mono, 93 kbps
48.000 kHz, 16-bit, Stereo, 187 kbps

VivoActive G.723.1

The VivoActive G.723.1 codec is good for creating low bit rate general purpose audio files, active streaming format (.asf) files, or ASF audio streams for use over an intranet. The G.723.1 standard specifies the format and algorithm for sending or receiving voice over low bandwidth network connections.

8.000 kHz, Mono, 0 kbps

VivoActive Siren

The VivoActive Siren codec is good for creating low bit rate general purpose audio files.

16.000 kHz, Mono, 1 kbps

Voxware MetaSound

The Voxware MetaSound codec is designed for creating low-to-mid bit rate music files, ranging from 6 kbps Mono to 96 kbps high-fidelity stereo. This codec offers excellent high-fidelity compression for many different types of audio. However, MPEG Layer-3 is a good alternative if you require a high bit rate and low CPU usage.

AC8 File, 8 kHz, Mono, 0 kbps
AC10 File, 11 kHz, Mono, 1 kbps
AC16 File, 16 kHz, Mono, 1 kbps
AC24 File, 22 kHz, Mono, 2 kbps
AC8 Bitstream, 8 kHz, Mono, 0 kbps
AC10 Bitstream, 11 kHz, Mono, 1 kbps
AC16 Bitstream, 16 kHz, Mono, 1 kbps
AC24 Bitstream, 22 kHz, Mono, 2 kbps

Voxware MetaVoice

The Voxware MetaVoice codec provides good performance for extremely low bit rate voice-only audio tracks. Using the RT24 setting at 2.4 kbps, the MetaVoice codec compresses audio smaller than any of the other codecs installed with Windows Media Player.

RT24 File, 8 kHz, Mono, 0 kbps
RT29HQ File, 8 kHz, Mono, 0 kbps
RT24 Bitstream, 8 kHz, Mono, 0 kbps
RT29HQ Bitstream, 8 kHz, Mono, 0 kbps


Compressing a .wav file or increasing the compression ratio does not usually improve sound quality.

Changing a .wav file's attributes may increase or decrease file size, but does not usually improve sound quality.

You may have other audio codecs on your computer as a result of installing software provided by other software companies.

Related Topics

Sounds and Audio Devices overview

Using Sound Recorder

© 2017 Microsoft Corporation. All rights reserved. Contact Us |Terms of Use |Trademarks |Privacy & Cookies