A ringtone, a podcast intro, a sample, a voice memo with three minutes of silence at the start: trimming is the most common audio edit there is. The only real question is whether the cut should re-encode the audio or preserve it bit-for-bit. The answer changes the tool you should use.
What people actually trim
- Ringtones and alerts — 20–40 seconds from a song you own.
- Podcast clips — an intro, a quote, a segment for social media.
- Samples — a drum hit, a chord, a spoken phrase from your own session.
- Cleanup — removing silence, a false start, a cough, dead air.
- Delivery — sending only the useful minute of a longer recording.
Two ways to cut an MP3
| Method | What happens | Quality | Where |
|---|---|---|---|
| Frame-level cut | The compressed frames are copied; only whole frames can be removed. | Bit-perfect, no generation loss. | Desktop tools with bitstream access. |
| Decode + re-encode | The selection is decoded to PCM and encoded as a new MP3. | One extra lossy generation. | Browsers, mobile apps, most online cutters. |
| Export to WAV | The selection becomes uncompressed PCM. | Lossless, but no longer an MP3 and much larger. | Editors and DAWs. |
Braise uses the second method. It is the only approach a browser can perform reliably, and it gives sample-accurate edges. For a ringtone or a podcast clip, the extra generation is inaudible at 192 kbps or higher. For mastering, use a lossless source and a DAW.
Why MP3 frames matter
MP3 is not a sequence of independent samples. It is a sequence of frames, each covering 1,152 samples — about 26 ms at 44.1 kHz. A lossless cutter can only remove whole frames, so its edges land on those boundaries. It also has to deal with the bit reservoir: a frame can borrow data from previous frames, which is why naive frame deletion can produce glitches.
A decode-and-re-encode avoids all of that. The browser decodes the whole file into PCM, takes the exact sample range you selected, and encodes a new stream from scratch. The trade-off is the extra lossy generation, not the precision.
Trim a file in the browser
-
01
Load one audio file
MP3, WAV, M4A, FLAC or OGG. It is decoded locally; nothing is uploaded.
-
02
Drag the selection
The waveform shows where the music starts and ends. Drag the handles or type exact times.
-
03
Preview it
Play the selection before exporting. It is much faster than exporting and checking afterwards.
-
04
Choose a bitrate and fade
192 kbps with a 10 ms fade is a good default. Use 0.5–1 s for a musical fade-out.
-
05
Export the clip
You get a new MP3. The original file is never modified.
Fades, clicks and edges
A waveform that is cut at a non-zero value produces a sudden step to silence. You hear it as a click. A short fade — 5 to 20 ms — removes the step without changing the perceived timing. Longer fades (0.5–1 s) are a musical choice, useful for ringtones and social clips that should not end abruptly.
The right bitrate for a short clip
| Ringtone on a phone speaker | 128 kbps mono or stereo |
|---|---|
| Podcast clip for social media | 128–192 kbps |
| Music clip for headphones | 192–256 kbps |
| Sample for further editing | 320 kbps, or export WAV |
A 30-second ringtone at 192 kbps is about 720 KB. At 320 kbps it is about 1.2 MB. Neither is large; the difference is mainly how much of the source detail you keep. For a phone speaker, 128 kbps is honest.
One rule about rights
Trimming a song does not create a new right to it. A ringtone from a song you own is a personal copy; publishing a clip of someone else’s recording is a different act. If you plan to publish, use audio you created, audio under a permissive licence, or audio you have cleared. See legal music sources.