Initializing... drag & drop files here
Supports: WAV
WAV stores uncompressed PCM samples — pristine quality, but a 1-minute 44.1 kHz stereo recording is roughly 10 MB. AMR (Adaptive Multi-Rate) is a speech codec built by Ericsson and standardized by 3GPP for GSM and UMTS networks; it strips music-range frequencies and parametrically encodes the human vocal tract, dropping the same minute to 35-90 KB. That's where WAV-to-AMR conversion earns its keep:
See also WAV to MP3 and WAV to M4A when you need a smaller file but still want music-quality output, or AMR to WAV for the reverse direction.
| Property | WAV (source) | AMR (output) |
|---|---|---|
| Compression | Uncompressed PCM | Lossy ACELP, speech-optimized |
| Sample rate | 8 kHz to 192 kHz, freely chosen | 8 kHz (NB) or 16 kHz (WB) only |
| Channels | Mono or stereo | Mono only |
| Bitrate | ~128 kbps (8-bit/8k mono) up to 1,411 kbps (16-bit/44.1k stereo) | 4.75-12.2 kbps (NB), 6.60-23.85 kbps (WB) |
| 1-minute file size | ~1 MB to ~10 MB | ~35 KB to ~180 KB |
| Frequency range | Full audible spectrum (20 Hz - 20 kHz+) | 300 Hz - 3.4 kHz (NB) / 50 Hz - 7 kHz (WB) |
| Music suitability | Excellent | Poor — speech only |
| Best for | Recording, editing, archival | Telephony, voicemail, MMS, IVR prompts |
| Mode | Bitrate | 1-min size | Best for |
|---|---|---|---|
| AMR-NB 4.75 kbps | 4.75 kbps | ~35 KB | Maximum compression — embedded prompts, low-bandwidth telephony |
| AMR-NB 7.40 kbps | 7.40 kbps | ~55 KB | Standard voicemail / IVR — intelligible, very small |
| AMR-NB 12.2 kbps | 12.2 kbps | ~92 KB | GSM full-rate quality — best AMR-NB speech clarity |
| AMR-WB 12.65 kbps | 12.65 kbps | ~95 KB | HD voice baseline — fuller speech with 7 kHz top end |
| AMR-WB 23.85 kbps | 23.85 kbps | ~180 KB | Highest AMR fidelity — VoLTE-grade speech |
No. AMR is a parametric speech codec — it models the human vocal tract and discards everything above 3.4 kHz (NB) or 7 kHz (WB). Music, instruments, and singing are heavily distorted because cymbals, guitar harmonics, and high vocals all live above AMR's cutoff. If you need a small music file, convert to WAV to MP3 at 96-128 kbps or WAV to OPUS at 64 kbps instead.
Match the playback target. AMR-NB (8 kHz sample rate, 4.75-12.2 kbps) is what GSM voicemail, MMS, and pre-VoLTE phones expect — universal compatibility, smallest files. AMR-WB (16 kHz, 6.60-23.85 kbps) is HD Voice / VoLTE quality and sounds noticeably clearer, but only newer handsets and modern PBX systems decode it. When in doubt for telephony deployment, use AMR-NB at 12.2 kbps.
The AMR codec specification only carries a single channel — it has no stereo mode. XConvert downmixes your stereo WAV to mono before encoding (averaging left and right). For speech this is the right call; for music it's another reason AMR isn't appropriate for non-voice content.
Roughly 100x to 300x smaller. A 60-second 44.1 kHz 16-bit stereo WAV is about 10 MB. The same minute as AMR-NB at 12.2 kbps is ~92 KB; at 4.75 kbps it drops to ~35 KB. AMR-WB at 23.85 kbps is the largest AMR option and still only ~180 KB per minute.
Apple dropped native AMR support years ago — iOS Voice Memos and the default player won't open .amr. VLC for iOS and most third-party audio apps decode AMR fine. If you need native iPhone playback, convert to WAV to M4A instead, which is the format iOS uses for voice memos.
Yes. Use the trim section to set a start time and duration in seconds (12.5) or HH:MM:SS.sss (00:00:30.500). Useful for cutting a single voicemail greeting out of a longer studio take or removing silence at the start of an IVR prompt before deploying it.
AMR-NB cuts off at 3.4 kHz — half the frequency range of even AM radio. Sibilants ("s" and "f" sounds), high-pitched voices, and any music content above that cutoff are simply gone. Switch to AMR-WB if your decoder supports it; the 7 kHz top end restores most of what makes speech sound natural. If muffled audio is a deal-breaker, AMR is the wrong target — use MP3, M4A, or OPUS.
Only if storage is the dominant constraint and you've already extracted a transcript. AMR is intelligible but lossy and band-limited; once it's encoded you can't recover the source quality. The standard practice is: keep the WAV master, encode an MP3 or AAC copy for distribution, and only fall back to AMR for legacy telephony deployment or storage-starved embedded devices.