The trap
Turn the drive up and the track gets louder. Louder sounds better, every time, to everyone, for about the first ten seconds. Then you bypass the plugin, the track drops in level, and the bypassed version sounds thin and lifeless by comparison.
It probably isn't. You just made it quieter.
This is the single most common way to over-clip a mix: every comparison you run tells you the clipper is helping, because every comparison is also a volume test. Two controls in HardStop exist to take that variable away, and one of them is on by default for exactly this reason.
Auto Gain: compare tone, not level
AUTO GAIN sits after the clipper and the TP safety limiter, just before Mix and OUT GAIN. As you add drive, it pulls the output back down by the same amount, so the plugin's output stays roughly where it started.
With it on, bypassing HardStop compares what the clipping did to the sound. With it off, bypassing compares how loud you made it, which you already know.
- Judging? Auto Gain on. Toggle bypass as much as you like, and listen for transients flattening, low end thickening, cymbals fizzing.
- Committing? Auto Gain off, and use OUT GAIN to set the level you actually want.
If you only take one thing from this page: leave Auto Gain on while you are deciding how much drive to use.
Delta: hear exactly what is being removed
LISTEN solos what the clipper took out: the difference between what went into it and what came out. Not a guess at it, the real thing.
What you are listening for:
- Healthy. Short, intermittent ticks. Peak tips only. It should sound like almost nothing, and it should stop when the music does.
- Too far. You can recognize the performance. If you can hum along with the delta, or hear the snare's tone rather than its very tip, you are clipping body rather than peaks.
- Aliasing. A layer of grit that carries on after the transient and sits at pitches unrelated to the source. That is a quality-tier problem, not a drive problem: see Normal, HQ and Ultra.
In HardStop Pro, Listen also lets you pick the source, so you can hear the clipper and the true-peak limiter separately and find out which stage is doing the work.
A procedure that survives your own ears
- Auto Gain on. Non-negotiable for this part.
- Drive up until you hear it, then back off. Push until something audibly changes, note where that was, then reduce until it stops. That point is your ceiling for this source, and it is usually lower than the point where it "sounds louder".
- Check the delta. Peak tips only? Good. Recognizable performance? Back off further.
- Bypass twice, ten seconds apart. First impressions are unreliable on the second listen; if the engaged version still wins with levels matched, it is genuinely better.
- Auto Gain off, set OUT GAIN, move on.
Two things that will mislead you anyway
Switching quality tiers mid-comparison. With the TP safety limiter off, Normal reports zero latency and the oversampled tiers do not, so your DAW re-compensates when you switch. Give it a moment before judging. HQ and Ultra report identical latency, so those two are directly comparable with no settling at all.
Comparing across sessions. Preset A/B is in HardStop Basic. Use it rather than trying to remember what the last setting sounded like, because you will not.
For the wider version of this discipline, applied to a whole mix rather than one plugin, see the mixing checklist pros use before every release.