To compare masters, match their integrated loudness first, then switch between them in sync on the same section of the song. A master is almost always louder than the mix, and a louder file sounds fuller and brighter, so an unmatched comparison mostly tells you which file is louder. Once the levels match, check four things: the loudness each master was made to, its true peak, its tonal balance, and where in the spectrum the masters differ. Then write feedback in those terms, with times and numbers.
What you might be comparing
- Your mix against its master, to hear what mastering did.
- Masters of the same mix from two or more mastering engineers.
- Masters from AI or online mastering services, often several presets or loudness settings from one service.
- The first master against the second, after you sent notes.
- Your master against a commercial release in the same style.
The process is the same for all of them. Use the same mix file for every master, so the only thing that differs is the mastering.
Match loudness before you listen
Ears are less sensitive to lows and highs at low levels (ISO 226, the Fletcher-Munson curves). A master 1 dB louder than another sounds fuller and brighter for that reason alone. Between a mix at −18 LUFS and a master at −9 LUFS the gap is 9 dB, and the master will win every time if you don't match them.
Measure each file's integrated LUFS and turn the louder ones down to the quietest. Don't turn the quiet one up, or it may clip. Now the master and the mix play at the same loudness, and you hear the EQ, compression, limiting and width changes. If the master only sounds better because it is louder, matching reveals it. The level matching guide explains the method.
In ABSeq, load the mix and up to three masters into A, B, C and D. Auto Gain turns every track down to the quietest one, and the A/B key switches to the next track at the same moment in the song.
What to check
Loudness
After matching, go back to the readings. Note each master's integrated LUFS. A louder master isn't rewarded on streaming services. Spotify normalizes to −14 LUFS integrated by default, so it turns a −9 LUFS master down 5 dB, and that master then plays at the same level as a −14 LUFS master, with less dynamic range left. Apple Music plays at about −16 LUFS, YouTube, Tidal and Amazon Music at about −14 and Deezer at about −15. These are estimates; only Spotify publishes its rules in full.
So a −8 LUFS master and a −11 LUFS master will play at the same level on Spotify. Compare them matched, and pick the one that sounds better there. The guide to master loudness covers targets in more detail.
True peak
True peak is the highest level of the waveform between samples. A master can read 0 dBFS on a sample peak meter and still have inter-sample peaks above 0 dBTP, which can clip in the converter or when a service encodes it to AAC, MP3 or Ogg. Spotify recommends masters peak no higher than −1 dBTP, and −2 dBTP for masters louder than −14 LUFS. A −9 LUFS master peaking at −0.3 dBTP misses both. Ask for a version with a lower ceiling.
Tonal balance
Listen to the low end, the vocal presence and the top end in turn. Then look at a long-term tonal balance graph. Commercial mixes tend to follow a slope of about −4.5 dB per octave. That is a general approximation, not a genre target, so use it to spot a master that leans far from your references rather than to grade it. The tonal balance guide explains how to read the curve.
The difference by band
An A/B difference graph shows where one master has more than another, band by band, after loudness matching. It turns "B sounds heavier" into "B has 2.1 dB more at 63 Hz". In ABSeq Pro, the A/B difference graph names the biggest gap in its headline, for example "B +2.1 dB @ 63 Hz". Bands below about 125 Hz are coarse, so read small differences there with care and confirm them by ear.
Choosing between several masters
With three or four masters, start by removing the ones that fail a check: true peak above the ceiling you need, or a tonal balance far from your references. Then compare the rest at matched loudness on the same three sections, for example the first verse, the loudest chorus and the quietest part. A limiter can sound fine on a verse and pump on a chorus.
When two masters are close, name the files neutrally and test blind. If you pick the same master 9 times in 10, chance is an unlikely explanation. If your picks split evenly, either master will do, and you can choose on price, turnaround or how well the engineer handles notes. The blind A/B test guide covers how to run it.
Revision rounds
Keep every round. Compare the new master against the previous one and against the mix, all matched. A note that asked for less low end should show in the A/B difference between round one and round two. If it doesn't, the change didn't happen or went to a different band.
Save each comparison so you can go back to it. A session that keeps the files, their loudness and your blind test results also records why you approved a master.
How to give feedback
Notes like "make it punchier" or "more open" leave the engineer guessing. Notes with a time, a part and a number can be acted on.
- Say what you compared and how: "compared at matched loudness against the mix and against a reference".
- Give the time and the part: "chorus at 1:12, the kick and bass".
- Say what you hear and in which direction: "the low end is louder than the mix, and the kick loses its attack".
- Add the measurement if you have one: "about 2 dB more at 63 Hz than the mix, true peak −0.3 dBTP".
- Say what you want: "keep the low end closer to the mix, ceiling at −1 dBTP".
The same notes work for an AI or online mastering service. There you change a preset, an intensity or a loudness setting instead of sending a message, then compare the new file against the last one in the same way.
Questions
How do I compare a mix and a master?
Turn the master down to the mix's integrated loudness, then switch between them in sync on the same section. The master is usually several dB louder, and unmatched it will sound better for that reason alone.
How do I compare AI mastering results?
Master the same mix file with each service or preset, match all the results to the quietest one by integrated LUFS, and switch between them on the same sections. Check true peak and tonal balance, then test the closest two blind.
Should my master sound better than my mix?
At matched loudness it should sound better or at least no worse. If the mix sounds better once you match levels, the mastering has hurt it, and you should say where and how in your notes.
Is a louder master better for Spotify?
No. Spotify normalizes to −14 LUFS by default, so a −9 LUFS master is turned down 5 dB and plays at the same level as a quieter one, with less dynamic range left.
How do I approve a master?
Compare it with the mix and any earlier rounds at matched loudness, check its true peak against −1 dBTP, or −2 dBTP if it is louder than −14 LUFS, and listen to the quietest and loudest sections. Approve it when it sounds at least as good as the mix everywhere.