Set the reference up properly first
Most reference comparisons are broken before they start.
Route it past your master chain. A reference on a normal track runs through your master bus, so you are hearing someone else's finished master through your compressor, your EQ and your clipper. Put it on a track that bypasses all of it, or use a reference plugin that sits outside the chain.
Pick one with the same arrangement density. A sparse ballad and a wall-of-guitars mix need completely different amounts of clipping. Matching genre is not enough: match how much is going on at once.
Accept that it is already mastered. Your mix is not. The reference will be louder, and louder wins, which is the whole problem described in how to judge a clipper without fooling yourself. Match the levels before you form any opinion. In HardStop Pro, the LUFS readout shows your mix's loudness; meter the reference on its own track and match the two; in Basic, match by ear until switching between them stops feeling like a volume change.
Listen to transients, not tone
A clipper does one thing: it shortens peaks. So when you compare against a reference, the useful question is not "is mine as bright" or "is mine as wide". Those are EQ and imaging problems and clipping will not fix either.
The useful questions:
- Does the snare hit with the same weight, or does mine sound like a snare that has been sat on?
- Does the kick have the same front end, the click before the body?
- When the chorus arrives, does mine get bigger, or just louder?
That last one is the important one. A well-clipped mix gains density when the arrangement thickens. An over-clipped mix stops responding: the chorus arrives and nothing happens, because there is no headroom left for it to happen in.
The loop
Four steps, repeated on the loudest section of the song rather than the intro:
- Drive up until the snare stops moving air. Not until it distorts, until it stops pushing. That is earlier than you think.
- Back off two or three decibels. You are looking for the point where it recovers.
- Check the delta. Peak tips only. If you can recognize the performance in it, you are still too far. What a healthy delta sounds like is covered in judging a clipper.
- A/B against the reference at matched loudness. Preset A/B is in Basic, so store both and switch rather than trusting your memory of the last setting.
The signs a meter will not show you
Your meters will look fine long after your mix has stopped working. These are the things to listen for instead:
- The snare lost its depth. It still cuts, but it sounds like it is coming from a smaller room.
- The bass pumps against the kick. Two loud low-frequency events competing for the same shrinking headroom.
- Cymbals turned to fizz. The harshness described in why it distorts on some sources.
- It got louder and smaller at the same time. The most reliable sign of all, and the hardest to hear without a reference, which is exactly why you set one up.
When to stop clipping and limit instead
Clipping is a peak tool. When you find yourself adding drive and the mix is not getting louder any more, only denser, you have reached the end of what a clipper can do for this material.
That is the point to let a limiter after HardStop take the last stretch, or to go back to the mix itself. Two or three decibels of clipping followed by a limiter almost always beats six decibels of clipping alone, and it leaves the transients you were trying to keep.
For the wider pre-release checklist this fits into, including tonal balance and mono compatibility, see the mixing checklist pros use before every release. For where to put your ceiling before export, see it measured under the ceiling but still distorted.