What Is AMR?

How it works

AMR was designed for mobile phone voice transmission over GSM and 3G networks. It uses a variable bitrate — the encoder selects the bitrate dynamically based on the complexity of the audio — to achieve very small file sizes while keeping speech intelligible.

There are two variants: AMR-NB (Narrow Band, the original) and AMR-WB (Wide Band, sometimes called AMR+), which offers higher quality speech by encoding a wider frequency range. Both store audio at much lower quality than even the most compressed MP3 settings — they are optimised for voice, not fidelity.

Where AMR files come from

• —Nokia phone voice recorders (2000s–2010s) — the most common source
• —Early Samsung and Sony Ericsson handsets
• —Android voice recorder apps on older or budget phones
• —VOIP call recording applications
• —3GPP video containers (.3gp, .3g2) may embed AMR audio

Strengths

• +Extremely small file size — very efficient for voice recordings
• +Speech remains intelligible even at the lowest bitrate modes
• +Standardised format — widely supported in telecom equipment

Weaknesses

• −Very poor audio quality for anything other than speech
• −Limited playback support on modern devices and software
• −Narrow audio bandwidth — music, background sound, and high frequencies are degraded
• −Converting AMR to a higher bitrate does not improve quality — quality is set at recording time

When to convert AMR

Convert AMR to MP3 or WAV when you need to play the file on modern devices, edit it in audio software, or run it through transcription tools. Most modern audio players and applications do not support AMR natively.

Converting to MP3 is suitable for playback and sharing. Converting to WAV is better for editing — Audacity, transcription services, and speech recognition tools generally accept WAV without issue.