Release Your AI Song Before New Year: A Checklist
A checklist for releasing an AI song before New Year: check your plan's rights and the distributor's AI rules, plan the dates, then clean and master the file.
A holiday release can be stopped by three things: the rights your generator plan gives you, the rules your distributor applies to AI music, and a master that misses basic loudness and true-peak targets. Check them in that order, before you master the final file. The checklist below follows that order and gives the numbers and the Echoteka steps for each part.
The order also reflects what is hard to undo. Rights and the AI answers you give a distributor are the hardest parts to change after upload, so they come first. A master can always be redone.
1. Check the rights first
Rights come from the plan you made the song on. Suno's help center says that songs made on the free plan are for non-commercial purposes only. It also says that songs made on Pro or Premier keep their commercial rights even after you end the subscription (Suno help center, checked 12 October 2026). So the question "can I release this?" depends on the plan you had when you generated the song.
Whether you may release the song depends on your Suno plan. Read Suno's current terms before you upload, because they change. Other generators, Udio included, have their own terms, so read the terms of the plan you used there too. Do not rely on a summary written by someone else.
2. Check the distributor's AI rules
Distributors set their own rules for AI-generated music. Some ask you to declare AI use when you upload, and some check or label tracks after release. The rules change, so read your distributor's current policy on the day you submit, and answer its questions truthfully. Keep a short note of the generator, the plan and the date you made the song. If a question comes up later, you will have the facts.
3. Work back from the release date
Choose the day you want the song live, then plan backwards: the final master, the measurements, the upload and the review. Holiday weeks slow reviews and deliveries at many services, so do not count on a fixed number of days. Upload with a wide margin instead of on the last evening before the date you need.
A simple plan works well here. In the first week, check the rights and the distributor's rules. In the second, clean and master the song and measure it. In the third, upload and leave the review time it needs.
4. Clean the artifacts
Generated audio often carries marks that show up as metallic shimmer or watery highs. Clean the file before you master it, not after. Echoteka's AI artifact cleanup removes narrow tones above 15 kHz, smooths seams between generated blocks and corrects fixed spectral peaks. It is designed to leave loudness and tone where they were. Cleaning first means the mastering step works on the file you actually want to release. If the cleaned version sounds worse to you, keep the original and compare the two on the same speakers. Cleanup is optional, and the decision is yours.
5. Master to a target, not to the limit
Integrated loudness is measured with ITU-R BS.1770, the standard behind LUFS. EBU R 128 uses the same method with a broadcast reference of −23 LUFS and a true-peak ceiling of −1 dBTP. For streaming, the reference points are these:
- Spotify. Its Normal setting plays at −14 LUFS. Keep the true peak below −1 dBTP, and below −2 dBTP if the track is louder than −14 LUFS.
- Apple Music. Apple does not publish a loudness target for music. Third-party measurements put Apple Music around −16 LUFS, so treat that as a reference, not a rule.
Our data on 248 Suno exports shows where AI songs usually start: a median of −14.3 LUFS, a median loudness range of 5.8 LU, and 40% of exports below 5 LU of range (Suno loudness report). Loudness range measures how much the level moves between quiet and loud parts. As a rule of thumb, if the master loses more than 1 to 2 LU of range, it is probably squeezed. Check the true peak again after you set the loudness, because gain pushes the peaks up as well.
Louder does not win on streaming. Spotify turns down tracks louder than −14 LUFS in Normal mode, and the dynamics you gave away do not come back.
6. Metadata, artwork and format
- Metadata. Use the same title, artist name and genre on every service, and give the same answers to any AI questions. Small differences, such as a featured artist credited one way on one service and another way on a second, can make one release look like two.
- Artwork. A common spec is a square image of at least 3000 × 3000 px. Some distributors reject artwork that carries URLs, prices or streaming-platform logos. Check your distributor's specification.
- Audio. If the distributor asks for lossless audio, deliver 24-bit WAV. Echoteka's master comes out as 24-bit WAV, FLAC or MP3 320.
7. Final checklist
| Check | What to confirm |
|---|---|
| Rights | Your plan allows the release, and you read the current terms on the upload date |
| Distributor | AI questions answered, policy checked on the day you submit |
| Dates | Upload has a margin before the date you need |
| Artifacts | Cleaned, then checked on headphones and on speakers |
| Loudness | About −14 LUFS integrated, measured with ITU-R BS.1770 |
| True peak | At or below −1 dBTP |
| Loudness range | Not squeezed compared with the original |
| Files | 24-bit WAV if the distributor asks for lossless |
| Artwork | Square, at least 3000 × 3000 px for many distributors |
| Metadata | Same title and artist name on every platform |
Common mistakes
- Assuming a song is cleared because it was made with AI. The plan decides.
- Treating the upload date as the date the rights were set. The rights follow the plan the song was made on.
- Mastering to −9 LUFS to sound louder than other releases. The platform turns it down, and the dynamics suffer.
- Uploading an MP3 when the distributor asks for WAV.
- Leaving the upload until the last week.
- Skipping the listening check after cleanup and mastering.
How to do it in Echoteka
- Measure the song with the free LUFS checker. It needs no sign-in and checks against −14 LUFS for Spotify and YouTube and −16 LUFS for Apple and podcasts. It accepts files up to 50 MB and allows 20 checks per hour per IP address.
- Clean the AI artifacts with AI artifact cleanup. Cleanup does not use a credit, and guests get one run per UTC day.
- Master with the free master: one full master per UTC day, no account needed. Every master comes with a before-and-after report on integrated loudness, true peak, loudness range, crest factor and stereo correlation. If a target was not reached, the report says so.
Echoteka makes a track sound finished and measures the result; it does not hide where the music came from. For the Suno-specific steps, see the Suno mastering page.
FAQ
Can I release a Suno song before New Year? It depends on your Suno plan. Suno's help center says that songs made on the free plan are for non-commercial purposes only. Read the current terms of the plan you used before you upload.
Do distributors accept AI-generated music? Many do, but each sets its own rules for disclosure. Read your distributor's current AI policy and answer its questions truthfully.
What LUFS should an AI song be for Spotify? About −14 LUFS integrated, with true peak below −1 dBTP. Spotify's Normal setting turns louder tracks down anyway, so a hotter master gains nothing.
Is there free AI mastering for a song? Echoteka gives one full master per UTC day without an account, with a before-and-after report for every master.
Should I remove AI artifacts before a release? If you hear metallic shimmer or watery highs, yes. Cleanup is designed to leave loudness and tone alone, and it does not use a credit.
Master your track free
1 full master every day, no account needed. Upload a WAV or MP3 and get a streaming-ready master with a before/after report.
Master a track free