Can you tell me which flaw is wrecking my track?

No. Nothing here listens to your files, and there is nowhere to upload one. What you get is the vocabulary, and nothing else: the names, and how small a dose of each one your own ears still catch.

Because three is what the clip pipeline can render as a controlled dose with a right answer behind it: pitch drift, timing smear and compression damage. Plenty of other things go wrong in a mix; they are absent because we cannot measure them yet, not because they do not matter.

If I catch these in the trials, will I catch them in my own work?

Unmeasured, so it is not claimed. The instruments report what you caught in these trials, on these recordings, in physical units. Whether that transfers to your own sessions is a question no data here answers.

REFERENCE · WHAT THE GYM CAN MEASURE

Naming what went wrong

You can hear that a render is wrong and have no word for it. This page is the word: three kinds of damage the gym can measure, what each one sounds like, and the machines that find how small a dose you can still catch.

Pitch drift

“It sounds sour or slightly seasick, and nothing you can point at is off-key.”

The whole take slides out of tune while it plays. It starts where it should and ends somewhere else, so what is wrong is never a note — it is a slope.

Measured in cents by The Delicacy Trials and The Threshold Test. A cent is a hundredth of a semitone, so a hundred of them is the step from one key to the next.

Timing smear

“It feels rubbery and unanchored. The groove will not lock, however much you nudge the drums.”

The beat wanders off the grid and back again in slow waves. No single hit is late enough to notice on its own; all of them together are.

Measured in ms by The Delicacy Trials and The Threshold Test. A millisecond is a thousandth of a second; the figure is how far a typical beat has wandered off the grid.

Compression damage

“It sounds cheap, underwater or brittle — like a good idea saved one too many times.”

Low-bitrate compression throws away quiet detail. Cymbals turn grainy and reverb tails go swishy and airless, while the loud middle survives intact.

Measured in kbps by The Delicacy Trials and The Threshold Test. Kbps is how much data each second of audio was allowed, so a lower number means more damage.

These three are what the pipeline can render as a controlled dose with a right answer behind it. They are not a list of everything that can go wrong with a piece of audio. A render can fail in ways nothing here measures, and this page would rather be short than pretend otherwise.

Questions, answered straight

Can you tell me which flaw is wrecking my track?: No. Nothing here listens to your files, and there is nowhere to upload one. What you get is the vocabulary, and nothing else: the names, and how small a dose of each one your own ears still catch.
Why only three?: Because three is what the clip pipeline can render as a controlled dose with a right answer behind it: pitch drift, timing smear and compression damage. Plenty of other things go wrong in a mix; they are absent because we cannot measure them yet, not because they do not matter.
If I catch these in the trials, will I catch them in my own work?: Unmeasured, so it is not claimed. The instruments report what you caught in these trials, on these recordings, in physical units. Whether that transfers to your own sessions is a question no data here answers.