Releasing a Suno Song on Streaming: Checklist
A Suno song goes out to streaming services without rework if eight checks pass before upload: rights under your plan, the distributor's rules on AI content, sound without artefacts, loudness and true peak, mastering, AI-trace cleanup, stems, and metadata with the cover. Each item below says what to check, and a summary table sits at the end.
Why the checks come before upload
A published track can be fixed, but the fix is slower and more awkward than doing it before release. Replacing a file, changing metadata and correcting answers in a distributor's form all follow their own route, and any of them can delay the release. So it helps to run the checklist before upload and note what was checked and on what basis.
The order is deliberate. First come the items that decide whether the track can be released at all: rights and rules. Next comes the sound. Last are the items that go into the form and the metadata.
1. Rights under your Suno plan
Commercial rights depend on the plan under which the track was created. Terms change, so do not rely on summaries from chats or videos. Check the terms of your plan on the Suno website and keep a copy of what they say on the release date, as a screenshot or a PDF.
This is not a formality. If a distributor or a platform asks what basis you release the track on, the answer should point to a source, not to memory.
Keep the generation history together with those terms: the prompt, the creation date and the version of the track. If the question comes up a month later, you have something to show rather than only recollections.
If the track has several versions, such as a separate cut for a video or a remix, decide in advance which one goes to the platform and keep the terms under which that version was made. If you asked the service's support about something, keep their reply with the terms: a written confirmation is more reliable than a memory of a conversation. If other people took part, for example a lyricist or a vocalist, record in writing who did what and on what terms. Do not upload someone else's track to any service without the rights to it, not even for measurement.
2. The distributor's rules on AI content
Many distributors ask in the upload form whether the track was made fully or partly with AI, and expect you to say so. The wording, the order of the tick boxes and the metadata requirements differ from one distributor to another, so open the rules of your distributor and check your answers against them.
Do not tick a box for something that did not happen, and do not hide something that did. The answers in the form are part of the release, and correcting them after publication is awkward.
Typically the form asks whether the track was generated fully or partly by AI, which service was used, whether the track includes someone else's samples or fragments, and who wrote the lyrics and the music. Your answers should match what actually happened. If you used cleanup or mastering tools in the process, check the rules on whether they need to be mentioned. Those are audio processing tools rather than music generation, and the rules may treat such cases differently.
3. Sound: artefacts, hiss, a flat top end
Generated music gives itself away through three typical flaws: a glassy or metallic top end, a muddy lower middle, and smeared transients. Before release, listen to the track on at least three systems: headphones, a phone speaker and a car, and if you hear hiss, ringing or a flat chorus, take the track back into work. For how to fix the sound and master the track for streaming, see how to master Suno songs.
4. Loudness and true peak
The targets for streaming are an integrated loudness of about −14 LUFS and a true peak no higher than −1 dBTP. Measure the track with the free analyser, which reports LUFS, true peak and loudness range. Chasing −9 LUFS is not worth it: streaming services level playback, and the dynamics suffer.
5. Mastering
If the track came out of the generator as it is, mastering brings the finished mix to the target loudness and peak, and the steps from export to upload are in how to master Suno songs. Base mastering in Magic Master runs within quotas: guests get 3 runs per calendar month, and accounts without tokens get 1 per day. Mastering with PRO modules costs one token. After mastering, compare the result with the original at the same integrated loudness, and check the top end and the chorus.
6. AI trace and cleanup
Generators leave traces in the signal itself: markers in the high frequencies, artefacts at the boundaries between blocks, and a repeating envelope pattern. In Magic Master the cleanup runs on the AI-trace cleanup page. It works on three markers: high frequencies above roughly 15.5 kHz, block boundaries, and the periodicity of the envelope. It does not change the loudness or the character of the track. The cleanup falls within the mastering quota, and no token is charged for it.
An important caveat: the cleanup addresses the sound, not the distributor's rules. Item 2 has to be checked separately, and no processing replaces it.
After cleanup, compare the result with the original on the same passages. If the timbre has changed noticeably, or clicks appear at the joins, do not upload that version. Check the result again and compare it with the original at the same loudness.
7. Stems
If you need separate parts, such as a backing track for a video or a remix, make the stems before mastering. Magic Master offers a two-track split, vocals and backing track: guests get 2 per day, accounts get 3 per day, and anything beyond the limit costs one token. A four-track split (vocals, drums, bass and the rest) requires signing in, and the result comes as a ZIP archive of WAV files.
Listen to each part on its own. A separation can leave artefacts that are barely audible in the mix but stand out in an isolated track.
Stems are not needed by everyone. To release a complete track they are usually unnecessary. They make sense when you need a separate backing track, a version without vocals for a video, a remix, or levels balanced between individual parts before the mix. If the goal is only to release the track, the finished mixed file and mastering are enough.
8. Metadata, cover and file
- Title, with no extra tags that are not part of the track itself.
- Artist name: your own name or an alias. Do not use the names of real artists, even in descriptions or tags.
- Genre, lyric language, release date and an explicit-content flag, as your distributor requires.
- ISRC codes are normally assigned by the distributor. Do not make them up.
- Cover art: a square, usually 3000 by 3000 pixels, with no third-party logos or photos you have no rights to. Distributor requirements can differ, so check them.
- File: distributors usually accept WAV at 16 or 24 bit. Check their sample-rate requirement too.
- True peak after MP3. If you upload an MP3, check the peak again, because encoding can raise it.
If the track fails a check
If even one item does not add up, do not rush the upload. The order is rights and rules first, then sound, then loudness. If the sound is wrong in one place, it is often quicker to regenerate that section than to fix the whole track. If loudness or the peak is off, mastering fixes it. If artefacts are audible, the AI-trace cleanup and careful top-end correction help. If the terms of your plan or the distributor's rules do not allow the release, no processing will fix that, and it is better to learn it before upload.
Summary table before upload
| Item | What to check | How to check |
|---|---|---|
| Rights | plan terms saved for the release date | the terms page on the Suno website |
| AI rules | form answers match the distributor's rules | the distributor's rules |
| Sound | no hiss, ringing or flat top end | listening on three systems |
| Loudness | around −14 LUFS | LUFS analyser |
| True peak | at or below −1 dBTP, including after MP3 | LUFS and true-peak analyser |
| AI trace | cleanup applied where it is needed | AI-trace cleanup page |
| Stems | every part listened to on its own | listening |
| Metadata and cover | filled in to the distributor's rules | the distributor's form |
How to do it in Magic Master
- The pre-order request for the release-ready package covers mastering and a check, delivered as a PDF. Submitting the request involves no payment, and the price is named before any payment. Sending the request requires signing in.
- The AI-trace cleanup runs cleanup and mastering in one pass, and it is free within the mastering quota.
- Mastering Suno tracks: base mastering within quotas, with PRO modules at one token per track.
- Stem splitting: the vocal-and-backing split gives 2 tracks; guests get 2 per day and accounts get 3, and anything beyond that costs one token. Four tracks require signing in.
Questions and answers
Why is there no sound in my Suno track?
If the file is silent or very quiet in every player, the problem is in the export, not in mastering. Mastering raises the level of audio that exists but cannot create sound out of silence, so it is usually quicker to export or generate the track again. If the sound is there but quiet, mastering to the platform target makes sense, and the causes and fixes are in improve Suno audio quality.
Does a Suno track need AI-trace cleanup before release?
That is your decision: the cleanup is worth running when generator traces are audible or show up in analysis. The AI cleaner changes only those traces and leaves loudness and the character of the sound as they were, and it is free within the mastering quota.
Can I release a Suno song on Spotify?
The answer depends on the terms of your Suno plan and on the rules of the distributor you upload through. Check both on their own websites rather than in summaries, because the terms change.
What is an AI fingerprint in a track?
It is a set of generator traces in the signal itself: high-frequency markers, artefacts at block boundaries and a repeating envelope pattern. They concern the sound, not the rights to the track.
Does this checklist also work for Udio tracks?
Most of the items are the same: loudness, true peak, sound quality, cover and metadata. Rights and terms are set by the service itself, so check them on the Udio website rather than against this list.
Upload a track — the same mastering engine, prices in euros.
Open Echoteka →AI mastering onlineLUFS analyzer