THE_CLIPPER 1.1 Is the Cleanest Clipper You Can Buy — Here's the Measured Proof
THE_CLIPPER 1.1 produces 34–73 dB less aliasing than two of the best-known clippers on the market at identical clipping depth. Same harmonics, none of the digital garbage — measured in THE_LAB and reproducible.
Today I'm releasing THE_CLIPPER 1.1, and I pushed the DSP quality even further. So I'm not going to just tell you it's clean — I'm going to show you, with numbers you can reproduce yourself.
I put the new 1.1 engine head-to-head against two of the best-known clippers on the market in a fully controlled measurement. Same signal. Same clipping depth. Factory defaults on all three. The result: THE_CLIPPER produces 34 to 73 dB less aliasing than both, while making the exact same harmonics.
Same sound. None of the digital garbage.
First: what "clean" actually means for a clipper
A clipper is a hard nonlinearity. When it flattens a waveform, it manufactures a whole stack of new harmonics — that's the point, that's the sound you want. But some of those harmonics land above half the sample rate (Nyquist). In the digital domain they can't exist up there, so they fold back down into the audible band as inharmonic tones that have no musical relationship to the note.
That folded-back junk is aliasing, and it's exactly what makes cheap clippers sound harsh, grainy and "digital" on bright material — cymbals, vocals, distorted guitars, anything with a lot of high-frequency energy.
So there are two very different things coming out of any clipper:
- Harmonics (THD) — musically related to the note. Intended. This is the sound.
- Aliasing — inharmonic, unrelated to the note. Never intended. This is the flaw.
A great clipper maximises the first while eliminating the second. That's the whole game. Aliasing is the only honest measure of how clean a clipper actually is.
How the test works (and why it's fair)
Everything below was measured in THE_LAB — my free plugin measurement tool — using its HARMONICS analyzer. I built the test so there's nowhere to hide:
- Identical signal. A pure sine, generated on an exact FFT bin so the analysis is fully coherent. That means folded aliases can never sneak onto a harmonic bin and disappear — they show up naked. 48 kHz, 65,536-point FFT.
- Identical clipping depth. Every plugin left at factory defaults, fed the same sine driven +6 dB past 0 dBFS, so all three clip the same amount of the waveform. I did not touch the rivals' settings.
- Proof the fight is fair. At 1 kHz all three land within 0.1 % of the same THD (~23.1 %). Identical harmonic distortion means identical clipping. The only variable left is aliasing.
That last point is the whole ballgame: when two plugins produce the same THD but one produces 40 dB less aliasing, that difference is entirely the quality of the anti-aliasing. Nothing else.
For the comparison I'm using two rivals — I'll call them Competitor 1 and Competitor 2. Both are well-known, well-respected clippers you'll recognise instantly. The point here isn't to dunk on anyone; it's to show what the measurement reveals, on tools people actually trust.
Round 1: 1 kHz — same harmonics, different garbage between them
At 1 kHz the harmonics fall inside the audible band, so you can see both the intended distortion and the aliasing in one picture. THE_CLIPPER is drawn upward; the rival is mirrored downward. Look at how perfectly the harmonic spikes line up — that's identical sound. Now look at the floor between the spikes.
THE_CLIPPER's floor is flat and silent. The rival's is packed with aliasing.
| Plugin | THD (the sound) | Worst alias (the flaw) | THE_CLIPPER advantage |
|---|---|---|---|
| THE_CLIPPER | 23.17 % | −121.4 dB | — |
| Competitor 1 | 23.19 % | −84.7 dB | 37 dB cleaner |
| Competitor 2 | 23.09 % | −87.4 dB | 34 dB cleaner |
Round 2: 8 kHz — the torture test
Push the tone up to 8 kHz and every harmonic a hard clipper makes lands above Nyquist. There's nowhere for the distortion to go except aliasing. This is where clippers get exposed.
This next view is "spurs-only" — the harmonics are masked out so you're looking only at the aliasing floor. Anything you see here is pure error.
| Plugin | Worst alias @ 8 kHz | THE_CLIPPER advantage |
|---|---|---|
| THE_CLIPPER | −136.9 dB | — |
| Competitor 1 | −74.2 dB | 63 dB cleaner |
| Competitor 2 | −75.9 dB | 61 dB cleaner |
THE_CLIPPER sits ~60 dB below the competition across the entire band. And the same story holds against the second rival — here's Competitor 2:
Round 3: slam it — the drive-invariance finding
Here's the part that actually separates a great anti-aliaser from a merely good one. Drive the input harder and harder and watch what happens to the aliasing floor at 8 kHz:
| Drive into clip | THE_CLIPPER | Competitor 1 | Competitor 2 |
|---|---|---|---|
| +6 dB | −136.9 dB | −74.2 dB | −75.9 dB |
| +12 dB | −137.9 dB | −64.9 dB | −74.9 dB |
| +18 dB | −137.3 dB | −63.1 dB | −74.0 dB |
THE_CLIPPER's aliasing does not move. It's pinned at ~−137 dB no matter how hard you hit it. Competitor 1 gets audibly worse the harder you push — which is exactly when you need a clipper most: loud masters, aggressive limiting, slamming drums. Against Competitor 1 slammed, the gap widens to 73 dB.
You reach for a clipper hardest on the loudest, brightest, most aggressive material — the exact situation where the other clippers alias the most and THE_CLIPPER doesn't budge. Drive it as hard as you want. The aliasing never rises.
The honest tradeoffs
Cleanliness this deep isn't free, and I'd rather tell you the cost than pretend there isn't one:
- CPU. THE_CLIPPER runs heavy oversampling and steep anti-alias filtering: about 4 % of one CPU core at 48 kHz. Competitor 1 is much lighter (~0.5 %) — because it does far less anti-aliasing, which is why it aliases 60 dB more. Competitor 2, which oversamples similarly to THE_CLIPPER, actually costs slightly more CPU while still aliasing ~60 dB more. So against the plugin doing comparable work, THE_CLIPPER is both cleaner and cheaper on CPU.
- Latency. The anti-alias filtering adds latency — about 3 ms versus ~1.3 ms for the others. It's reported correctly to your host and fully compensated, but it's there.
- True peak. Like every clipper here, THE_CLIPPER can produce inter-sample peaks slightly above 0 dBFS. It ships a true-peak mode to catch them; the rivals behave the same by default.
Bottom line: you spend a few percent of a core and a couple of milliseconds to buy 30 to 70 dB less aliasing. On anything bright, that's the difference between "clean" and "digital."
Reproduce it yourself
I'm not asking you to take my word for it. Every number here was measured in THE_LAB, which is free at philspeiser.com: all plugins at factory defaults, identical coherent-sine stimulus, matched clipping confirmed by matched THD. Load your clipper of choice, run the same test, and look at the floor between the harmonics.
That's the whole promise of THE_CLIPPER 1.1 in one sentence: same sound, none of the digital garbage — 34 to 73 dB less aliasing than the clippers everyone else is using.
