You approve the final WAV of a single. The vocal stays clear when the chorus opens up, and the drums have the impact you wanted. Then you listen to an encoded preview and hear a brief rasp on one drum hit. Is it in the original mix, is it appearing during encoding, or is the preview simply louder than the file you approved? You can answer that before the release if you compare the right files.
For an independent artist, approving a master and auditioning a delivery encode are two different checks. The first is a decision about the finished source. The second asks whether a particular conversion has introduced anything you would want to fix. You do not need to master the song again for every player. You do need to know which version you listened to and what the comparison actually tested.
Ask for a peak check and a listening check
The ITU’s BS.1770-5 recommendation, approved in November 2023, specifies methods for measuring programme loudness and true peak. A sample peak describes the highest stored sample. True peak estimates how high the reconstructed signal rises, including between samples. Those are different from overall loudness. A file can have sample values below the digital ceiling and still have a reconstructed peak that reaches higher than the sample meter suggests.
A peak reading gives you a place to investigate. It cannot tell you what a rasp sounds like or whether it was already part of the recording. In the hypothetical chorus, mark the time of the drum hit and check the same moment in both files. If the sound appears in the source too, look upstream before blaming the encode. If it appears only in the encoded copy, repeat the test from the same source and make sure the two files are reaching the same playback chain.
Compare the master with a specific encode
Apple’s Digital Masters resources provide tools to audition an Apple AAC encode. Its guide to auditioning audio on a Mac describes switching between the source and the encoded signal and inspecting peak and clipping information for each. That is a useful example of a controlled check. An Apple AAC audition is evidence about the encode you tested; it cannot stand in for every other codec, setting or playback path.
Start with the exact source file you intend to send to your distributor. Make a test copy using the relevant encode settings when they are available. Keep the source untouched and name the test copy clearly. If you do not know the platform’s processing settings, say what you tested rather than calling the preview an exact copy of what every listener will hear.
Compare at a similar listening level. If one file sounds more exciting because it is louder, your attention may go to the wrong feature. Our guide to matching playback loudness before approving a master explains how to make that comparison more useful. Keep any gain adjustment for auditioning separate from the original files and their peak measurements. Turning down the playback does not tell you whether distortion was introduced earlier.
Give the engineer something they can reproduce
‘The stream sounds rough’ is hard to act on. A better note identifies the exact master, the encode settings, the time and what you heard. For example: ‘On the first snare hit in the final chorus, the AAC test has a short crackle that I cannot hear in the WAV at the same playback level.’ That describes an observation. It gives the engineer a place to start without assuming the limiter, encoder or listening setup is to blame.
If the engineer sends a revision, re-encode that exact file and repeat the check. Listen to the surrounding passage as well as the single problem hit. A change that removes the crackle but makes the chorus feel noticeably weaker may need another conversation. Keep the original and revision identifiable so that a release form, a link or a chat attachment cannot silently send the wrong version back into the workflow.
Agree who signs off the encode. On a small release that might be the artist, with the engineer providing a test file and a short note. If there are clean and explicit versions, keep separate notes for each master. A fix to one version is not evidence that the other has been checked.
Run the check before you deliver
- Confirm which master and which delivery encode you are comparing. Keep the preview separate from the source you will submit.
- Check the busiest sections as well as quiet details such as a vocal tail or a long fade. Mark the times of anything you want to compare again.
- If a meter reports a peak or clip, inspect that passage in both versions. Keep the reading alongside your listening note.
- After a revision, compare again and confirm which file is approved for delivery. Follow your distributor’s actual requirements rather than treating one test setting as a universal rule.
If the problem appears in only one player, try the same two files elsewhere before requesting a new master. A playback effect or an unexpected level change can complicate the comparison. The goal is to narrow down where the difference appears, not to declare any compressed version inferior.
The listener will hear a release, not the studio’s folder of versions. A short, repeatable listening check helps you carry the decision you made in the studio through to the file you actually deliver.
One thought on “Music Masters: Listen to the Encode Before You Release”