Mix vs. Master: What’s Actually Wrong With Your Track
Mixing balances the parts against each other; mastering treats the finished file as one unit. A symptom by symptom guide to which stage is causing your problem.
Read More →
Upload a track and get the numbers that decide whether it is ready: integrated LUFS, true peak, clipping, loudness range, phase correlation and spectral balance against your genre. Measured from the file with ffmpeg, not estimated.
Two different kinds of number. We keep them apart, and we tell you which is which.
Computed directly from the audio. Objective and repeatable: the same file always produces the same numbers.
Loudness. Integrated and short-term LUFS, plus loudness range (LRA), measured with ffmpeg ebur128.
Peaks and headroom. True peak in dBTP, sample peak and crest factor, so you know what happens after lossy encoding.
Clipping. Count of samples pinned at full scale, the percentage, and exactly where in the track they occur.
Stereo image. Mean and lowest phase correlation, plus mono compatibility of the low end.
Spectral balance. Energy across eight bands, compared to reference targets for your genre family.
Tempo and key. BPM, musical key and scale, and tuning deviation in cents.
Judgement, not measurement. Indicative estimates that can vary between runs, useful for spotting weaknesses.
Mix quality judgement. How the measured values read together, and which of them matter most for this particular track.
Genre appropriateness. Whether a deviation from the reference curve reads as character or as a fault.
Priority. Which correction to make first when several are outstanding.
Improvement suggestions. Concrete moves for each problem found, phrased as actions rather than observations.
Loudness, true peak, clipping, loudness range and phase correlation are computed, not judged: the same file always produces the same numbers. Everything in the right-hand column is an assessment, and it can vary between runs.
A report you can act on the same evening, not a score to look at.
Integrated LUFS, true peak, clipping, loudness range, phase correlation and spectral balance, each against the target for your genre.
The corrections to make, already sorted into the correct processing order: clipping first, limiter ceiling last.
Timecodes for clipping events and the lowest correlation moments, so you can go straight to the bar that needs fixing.
Your spectral curve against the target shape for your genre family, normalised so you are comparing balance and not volume.
Whether anything technical is still standing between the track and distribution, stated plainly.
Exportable, so a mixing or mastering engineer can see the same numbers you did.
What the analysis does, and what it does not.
The guides behind what the analysis measures.
Mixing balances the parts against each other; mastering treats the finished file as one unit. A symptom by symptom guide to which stage is causing your problem.
Read More →
The six technical checks that decide whether a track is release-ready: LUFS, true peak, clipping, loudness range and mono compatibility, with the order to fix them in.
Read More →
The loudness target for streaming is around -14 LUFS integrated. What that number means, what each platform does with it, and why mastering louder gains you nothing.
Read More →