Back in November we published an article How Good Is Logic Pro Mastering Assistant? Take A Listen in which we processed a track with the four alternative flavours of AI assisted mastering which had come to Logic and comparing them with the unmastered version, asked the community which one they preferred. If you look at the results below you’ll see that the Mastering Assistant clearly produces results which listeners prefer to the original, but what else, if anything can we infer from the results?

The clear winner is the Valve option, favoured by nearly a third of respondents with Transparent and Clean roughly level in second place. Punch was significantly less popular choice and, reassuringly considering the purpose of mastering, the unprocessed version coming last.

This is of course only one example on one track so we can’t conclude that Valve is the best option more generally. But if we look at the recommendations for the character options of mastering from Apple we see:

• For a clean yet punchy sound, choose Clean, which is designed for EDM and acoustic music.

• For deep low end and refined high end, choose Valve, which is a good option for both acoustic music and hip-hop.

• For an aggressive sound with a subtle emphasis on mid-range frequencies, choose Punch, which is ideally suited for rock music.

Considering the recommendations and the fact that the example audio is a piano ballad, the fact that the transient-focused Punch option was least popular is understandable. Of the two options which might be the natural choices for this style of music Valve was clearly preferred. This might be because the emphasis on the lower frequency content is pleasing or because harmonic distortion adding a sheen to the top end. However another reason might be listener expectations and as the examples were named the poll did represent a blind test. It would have been interesting to know whether the results would have been the same if the examples were labelled A, B, C. Use the button below to visit the original post and hear the examples.

Do you use AI based mastering? If so do you find it has complemeted or replaced your previous mastering process? Share your thoughts int he comments.