Initializing... drag & drop files here
Supports: OGA
OGA is the audio-only flavor of the Ogg container — usually full-band Vorbis or Opus that can carry music. AMR is a narrowband speech codec built for mobile voice, not music. This converter re-encodes an OGA file into a small .amr clip for the specific cases where you need a tiny voice file: an old feature phone, an MMS message, or a voice-memo workflow that only accepts AMR.
This is a downward, lossy-to-lossy transcode. AMR-NB downsamples to 8 kHz mono and keeps only the speech band, so any music or high frequencies in the source will sound muffled and telephone-quality afterward. If your goal is just a smaller, shareable file with good fidelity, convert to MP3 instead — AMR is the right target only when something specifically requires it.
| Property | Value |
|---|---|
| Container | Ogg (audio-only profile) |
| Extension introduced | 2007 by the Xiph.Org Foundation |
| Typical codec | Vorbis (lossy, v1.0 May 2000); also Opus, FLAC, Speex |
| Compression | Lossy (Vorbis/Opus) or lossless (FLAC) depending on payload |
| Sample rate / channels | Full-band, mono or stereo |
| Typical Vorbis bitrate | ~45–500 kbit/s at 44.1 kHz |
| Best for | General audio and music on open platforms |
| Property | Value |
|---|---|
| Full name | Adaptive Multi-Rate (AMR-NB) |
| Standard | 3GPP TS 26.071 / TS 26.090, adopted October 1999 |
| Sample rate / channels | 8 kHz, mono only |
| Audio bandwidth | ~200–3400 Hz (telephone speech band) |
| Compression | Lossy, speech-optimized |
| Bitrate modes | 4.75, 5.15, 5.90, 6.70, 7.40, 7.95, 10.2, 12.2 kbit/s |
| Extension | .amr |
| Best for | Mobile voice, MMS, voice memos on legacy devices |
.amr file. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. No sign-up, no watermark.AMR-NB is a narrowband speech codec: it downsamples to 8 kHz mono and keeps only the ~200–3400 Hz telephone band, discarding everything above it. That is by design for voice. Any music, cymbals, or vocal sibilance in the source is simply gone after the conversion — there is no setting that restores it. If you need fidelity, convert to MP3 or keep the audio in WAV instead.
When something specifically requires AMR: sending a voice clip over MMS, loading a ringtone or voice memo onto an older feature phone, or feeding a speech-only pipeline that accepts only .amr. For voice-only content the quality drop is acceptable and the files are tiny — a minute of speech is often well under 100 KB.
Almost always, yes. AMR-NB tops out at 12.2 kbit/s, far below a typical Vorbis stream. In our testing, a one-minute OGA voice clip re-encoded to AMR-NB at the highest mode lands around 90 KB. The trade-off is fidelity: the savings come entirely from throwing away bandwidth and channels, not from clever compression.
AMR-NB runs at 8 kHz with a ~200–3400 Hz band and 8 bitrate modes up to 12.2 kbit/s — this is classic mobile-phone voice. AMR Wide Band (AMR-WB, ITU-T G.722.2) samples at 16 kHz with a 50–7000 Hz band and modes up to 23.85 kbit/s, so speech sounds noticeably clearer. Both are mono and speech-focused; pick Wide Band only if the receiving device or service supports it.
No. AMR-NB permanently discards the high frequencies and the second channel, so converting back only re-wraps what is left — it cannot rebuild what was removed. Keep your original OGA if you might need full quality later, and treat the AMR file as a one-way, voice-only export.
Yes. Your file is uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. There is no sign-up, no watermark, and files are never shared or made public.