MP3 vs WAV vs MP4: picking the right export format

The one rule that covers most cases

Export the highest-fidelity audio your plan allows, keep that as the master, and generate everything else from it later. Every other decision in this article is a footnote to that sentence.

The reason is asymmetry. You can turn a WAV into an MP3 at any bitrate, any time, in seconds, with no loss beyond the one you chose. You cannot turn an MP3 back into a WAV — re-encoding just wraps the damage in a bigger container. Choosing the smaller format first is a decision you cannot undo.

What each format is

When MP3 is the right answer

• You are sending the track to someone for feedback and want it to arrive instantly.
• The destination re-encodes anyway — most streaming and social uploads do.
• You are filling a phone or a car stereo and disk space matters more than the last few percent of fidelity.
• You need a quick reference mix to listen to on a walk.
• You are archiving a rough sketch you'd never master — not everything deserves the full-size copy.

At a decent bitrate, MP3 is genuinely transparent for casual listening. The problem is never the first encode; it is the third one, after the file has been through two more platforms.

When WAV earns its size

Any time the file is going to be processed again, take the uncompressed version. Compression artefacts that are inaudible on their own become audible once you EQ, compress, time-stretch or layer the material — you are amplifying exactly the parts the encoder decided to discard.

That means: mastering chains, DAW sessions, video edits where you'll duck the music under dialogue, stem-adjacent work, and anything destined for physical media. It also means your archive. Storage is cheap; re-creating a track you can no longer regenerate is not.

What MP4 is actually for

The video export exists because several major platforms will not accept a bare audio file. That is its entire job. It is not a higher-quality audio format — the audio inside is compressed, often more aggressively than a standalone MP3, because the bitrate budget is shared with the picture.

So use it as a delivery format, never as a master. If you want a visual for a track, it is usually better to export the audio at full quality and build the video separately, where you control the image, the aspect ratio and the audio bitrate. The MV studio on this site exists for exactly that workflow.

There is one genuine exception: a quick, disposable clip for a feed, where the whole point is speed and the post will be irrelevant in a week. For that, the one-click export is exactly right and anything more is wasted effort. Just don't let that file become your only copy of the song.

A decision table you can memorise

• One lossless master per keeper track.
• Lossy copies generated on demand, never archived.
• Video exports treated as disposable deliverables.
• Prompt and lyrics text stored beside the master.

Bitrate, sample rate and the limits of caring

There is a point past which format obsession stops paying. A generated track has already been through a rendering pipeline; chasing the last fraction of a decibel in the export dialog will not recover detail that was never in the source. Get the lossless file, stop worrying, and spend the attention on the prompt instead — that is where the audible differences actually live.

The practical floor is simple: don't archive lossy, don't master from video, and don't re-encode the same file twice for no reason.

If you want a single number to anchor on: for lossy copies, pick a bitrate at or above 256 kbps and stop thinking about it. Below that you start hearing it on cymbals and reverb tails; above it you're mostly buying file size. The same logic applies to sample rate — match whatever the source is and don't upsample, because upsampling adds data without adding information.

FAQ