Cubase project to WAV

Drop a .cpr to bounce it as a WAV, or select a file — it opens in the studio, ready to export.

CPR · ZIP

How to get a WAV out of a Cubase project

A collaborator asked for WAVs, or a brief did, and what you have in front of you is a .cpr. A project file is a description of a performance — parts, positions, tempo, names — and a WAV is a recording of one, which is the gap nothing can transcode across. The parts themselves travel. So they are played here, and that performance is what gets written down.

  1. Add your .cpr: The project file, or a zip with it inside — the reader finds the .cpr in the archive and opens it the same way. Only the project is read; whatever else is in the folder stays where it is.
  2. It is read here: Parts, positions, tempo and part names are parsed on your machine. Nothing is uploaded, and no account is needed to get this far.
  3. Choose a depth and bounce: Instruments first — a part with nobody assigned to it is silent. Export audio then writes 16- or 24-bit integer PCM, or 32-bit float, at 44.1 or 48 kHz, as the master mix, as one selected track, or as every track in a zip of stems.

More ways to work with your music

Try our online tools to separate, transcribe and convert — Doseedo does it all.

Questions about the cubase project to WAV converter?

We have answers.

Can I convert a .cpr to WAV?

No. A project file holds no audio, so there is nothing in a .cpr for a converter to transcode — the parts are read and performed instead, and the WAV is a recording of that performance. A WAV of your actual mix exists only once Export Audio Mixdown has run in Cubase.

Do I need an account?

Yes, for the render and the bounce — the instruments are played on a GPU, and a zip of stems comes out of that same step. The read does not: a .cpr is parsed on your own machine, signed out, and never uploaded.

Can I get stems instead of one file?

Yes. Every audible track that has rendered audio comes back as its own WAV in a zip, all of them starting from the same timeline zero so they drop back into line in one drag. A track with nothing rendered on it yet is not in the zip, and the export says so before you start.

Is a WAV bigger than the project?

Yes, by a long way. A .cpr is a description, a few megabytes at most; four minutes of stereo 24-bit audio is around seventy, because every sample of every second has to be written down.

Try our free cubase project to WAV converter

  • One tempo, and what stays in Cubase

    A .cpr stores note positions as ticks at 480 per quarter, and what turns ticks into seconds is the project's own tempo — one number, read once and applied to the whole file. A tempo track that ramps or steps is not followed, so a project that pulls up into the last chorus bounces at the speed it started at. That is worth knowing before you send a WAV somebody will line up against a grid. The rest of what stays behind is simpler to state: everything in the pool, because the parts are the only thing read. What is written is what the parts play.

  • Every track becomes one modelled player

    A part says what to play and never what plays it. Each track takes one of thirty-six modelled instruments, rendered on a GPU behind a free account. That decides what a WAV holds: one track, one instrument, no plugin output in the signal, which makes a stem a clean source and the mix a re-performance. The master bounce follows the studio's own mixer rather than Cubase's: track and bus gain, pan, mute and solo, region gains and fades, then the master fader. Your Cubase patches are not in it. The notes are. The bounce reads what the session holds, not the render that made it: a session survives a reload and a render does not, so the writer gets samples the store already encoded. That is what a session you can close costs. Pay it on a working file.

  • Lossless describes the file, not the resemblance

    Lossless is a property of the container: what the writer puts in the file comes back out of it unchanged. It promises nothing about whether the result sounds like your project, and on this page it does not. What the writer does at the last step is round — the render works in floating point, and a 16- or 24-bit integer is the nearest value on a fixed scale, clamped at full scale in both directions. Nothing else is applied on the way: no dither, no soft limiter, and the one loudness move in the dialog is a checkbox called Normalize peaks to −0.1 dBFS that stays off until you tick it. Left off, a bounce that clips in the studio clips in the file and a quiet one stays quiet — which is the behaviour a source ought to have, and the reason to move off the 16-bit the dialog opens on before this goes anywhere.

Try these Doseedo tools