A Channel Strip Can't Be As Clean As Chaining Premium Plugins. So I Measured All 15 of THE_STRIP 3's Modules.

Eight plugin windows open on one vocal bus. A high-pass you trust, a boutique EQ, the compressor everyone swears by, a de-esser, a saturator, a limiter. It sounds good. It also took forty minutes, and you have thirty more tracks.

So you try a channel strip, and the old thought shows up:

"There's no way fifteen modules in one plugin are as good as fifteen real plugins."

I had that instinct too, and it comes from somewhere real — stock DAW strips genuinely were afterthoughts. But it's not something you have to take on faith. "Pristine" is a number. So I measured it.

The result, before anything else: 666 automated checks at 44.1 and 48 kHz, 664 at 96 and 192 kHz, zero failures — 321 parameters, all 15 modules, on the exact engine that ships in the installer.

THE_STRIP 3 measured DSP performance — the short version

The rest of this article is the receipts — including the two findings that didn't pass, because a report that can only make its product look good isn't a report.

Why I measured any of this

When I started work on THE_STRIP 3, one thought wouldn't leave me alone:

What if every single module in the strip could match — or even outperform — the most expensive, most acclaimed plugins in the world?

The compressor. The dynamic EQ. The limiter, the filters, the warmth — all on that level. THE_STRIP 2 was already strong, but "strong" is something you say. I wanted to promise you, as a fact, that the processing in here stands with the very best tools available anywhere.

Ears are king in audio — they always will be. But processing quality isn't only taste. You can put a number on it. So I started measuring, benchmarking, and coding — and I hated myself more than once, because all of it had to fit inside the streamlined, AI-driven body of the plugin.

But now I can finally say it: I did it. 🍾

What "pristine" actually means

When someone says a strip can't be as clean as a chain, they mean four testable things:

1. The modules are approximations. A "24 dB/oct filter" that isn't really 24 dB/oct. A "+6 dB bell" that gives you +5.4 and a tilt you didn't ask for.
2. The strip taxes you even when you're not using it. Fifteen modules colouring your track before you turn a single knob.
3. The saturation is cheap. Whatever anti-aliasing a dedicated distortion plugin does, a strip surely does less.
4. It hides its costs. Latency it reports wrong, CPU that balloons, behaviour that changes at 96 kHz.

Everything below was measured in THE_LAB, my plugin measurement tool, hosting the exact engine that compiles into the VST3, AU and AAX you install — not a simulation, the shipping code. Apple silicon, Release build, 48 kHz unless noted; aliasing tests at 44.1 kHz, because fold-back is worst there.

Claim 2 first, because it's the easy one: off is off

With every module disabled, THE_STRIP 3 is a wire: −0.0016 dB flat from 20 Hz to 20 kHz across 1998 measured points, −153.02 dB THD+N, −240 dBFS self-noise on digital silence, 0 samples of latency, normal polarity.

All fifteen modules off. The shaded band is the ±0.10 dB acceptance threshold; the gold line is the measured envelope.

Measurement	Result	Acceptance threshold
THE_STRIP 3 magnitude flatness, 20 Hz – 20 kHz	−0.0016 to −0.0000 dB	≤ 0.10 dB
Gain at 1 kHz	−0.00 dB	0.00 ±0.05 dB
THD+N at 1 kHz	−153.02 dB	≤ −110 dB
Self-noise on digital silence	−240.00 dBFS	≤ −140 dBFS
Reported latency, all modules off	0 samples	—
Measured points in band	1998	—

Envelope, not a swept curve — the report publishes the bounds (−0.0016 dB to −0.0000 dB) rather than per-point data, so both bounds are drawn. At this scale they are one line.

And a module at zero isn't a module. I set the MOD stage to mix 0 and compared it sample-by-sample against the stage being fully disabled: bit-identical, a delta of 0.000000000000, in all four modes. In a chain, every plugin you left inserted "just in case" is still running, still resampling, still reporting latency. Here, an unused module is arithmetically absent.

Stability is the same story. All 321 parameters swept to both extremes stay finite and bounded. The delay at 0.95 feedback peaks at −1.41 dBFS and decays to −64.51 dBFS within thirty seconds. The reverb tail falls monotonically to −100 dBFS. Nothing explodes, nothing rings up, nothing pushes past the limiter's ceiling.

That's the floor — it isn't an achievement, it's just the part the objection gets wrong first. Now the parts that took the actual work.

Claim 1 — is a module in a strip a real module?

For the linear parts of a mixing chain, there is no "better." There's correct. A 24 dB/oct Butterworth high-pass has exactly one right answer — the maths was settled in 1930: −3.01 dB at the corner, maximally flat below it. A plugin can miss that number. It cannot beat it.

So the only question that matters: does the strip hit the number?

Corner at 200 Hz, resonance 0. Dashed grey traces are the Butterworth magnitude response, computed. Gold marks the measured corner values. Hover for figures.

Slope	Measured	Textbook ideal	Error	Slope measured
6 dB/oct	-3.00 dB	-3.01 dB	0.01 dB	5.8 dB/oct
12 dB/oct	-2.98 dB	-3.00 dB	0.02 dB	12.0 dB/oct
24 dB/oct	-2.95 dB	-2.99 dB	0.04 dB	24.1 dB/oct
48 dB/oct	-2.89 dB	-2.98 dB	0.08 dB	51.4 dB/oct

The dashed curves are computed, not measured — the report gives measured values at the corner, not a full swept curve, so the gold point is the measurement and the dashed lines are the ideal it is checked against. All four corners land within 0.12 dB of each other, so at this scale they stack into a single point; the table carries the individual figures.

Worst corner error across 6, 12, 24 and 48 dB/oct: 0.08 dB. Passband maximum: −0.000 dB at every slope. No ripple, no lift, no "character" smuggled in below the corner.

The EQ and the dynamics hit their numbers the same way:

The row nobody thinks to test is editor curve vs rendered audio: I compared the curve drawn on screen against the audio that actually came out. Worst disagreement across the whole band, 0.09 dB. The picture isn't a decoration. It's a promise the audio keeps.

One more detail worth a sentence: the ducker ducks the same 12 dB whether the trigger hits at −3 dBFS or −20 dBFS. That's the difference between a control that means something and a control you re-learn every session.

These aren't approximations of a real EQ and a real compressor. They're the numbers real EQs and compressors are aiming at.

Claim 3 — how clean is the saturation?

Fair suspicion, because saturation is the one place a strip could genuinely be worse. A nonlinearity manufactures harmonics above Nyquist that fold back into the audible band as aliasing — inharmonic junk that makes distortion sound digital. It's the same failure I went after in THE_CLIPPER, and the same measurement exposes it here.

An 8003 Hz tone at −1 dBFS, drive at 0.7, 44.1 kHz, limiter disabled. Decibels below the fundamental — more negative is cleaner:

8003 Hz at −1 dBFS, drive 0.7, 44.1 kHz, limiter disabled, CLEAN filters. Every point measured. Lower is cleaner.

Saturation model	1x	2x	4x CLEAN	8x CLEAN
Console	−12.9 dB	−41.4 dB	−97.0 dB	−142.0 dB
Transformer	−11.3 dB	−32.4 dB	−72.2 dB	−129.0 dB
Tube	−16.5 dB	−33.4 dB	−57.1 dB	−76.1 dB
Tape	−12.3 dB	−31.7 dB	−61.2 dB	−111.8 dB
Diode	−12.8 dB	−33.9 dB	−54.8 dB	−73.3 dB
FET (hard clipper)	−7.7 dB	−14.4 dB	−32.5 dB	−44.5 dB

Console at −142.0 dB and Transformer at −129.0 dB put their worst artefact below the noise floor of 24-bit audio. There's no headroom left to improve into — you can't make an artefact quieter than inaudible.

Then the test that separates good anti-aliasing from great: hold the drive, sweep the input level over 60 dB.

Drive held at 0.7 while the input is swept 60 dB. All ten points measured. Hover to compare the two settings at any level.

Input level	2x Fast	8x CLEAN	Difference
-60.0 dBFS	-153.27 dB	-171.97 dB	+18.70 dB
-53.3 dBFS	-139.85 dB	-171.03 dB	+31.18 dB
-46.7 dBFS	-126.55 dB	-170.80 dB	+44.25 dB
-40.0 dBFS	-113.20 dB	-171.05 dB	+57.85 dB
-33.3 dBFS	-99.88 dB	-170.71 dB	+70.83 dB
-26.7 dBFS	-86.62 dB	-171.52 dB	+84.90 dB
-20.0 dBFS	-73.62 dB	-171.90 dB	+98.28 dB
-13.3 dBFS	-61.67 dB	-172.44 dB	+110.77 dB
-6.7 dBFS	-51.68 dB	-148.82 dB	+97.14 dB
0.0 dBFS	-30.16 dB	-125.71 dB	+95.55 dB

At the fast setting, the alias floor climbs right along with your signal. At 8x CLEAN it doesn't move: pinned near −171 dB across a 47 dB range of input level. Hit it harder and the garbage stays exactly where it was.

Notice I didn't hide FET. It's a hard clipper, and a discontinuous derivative produces a harmonic series no finite oversampling fully resolves — that's inherent to the shape, not a defect in the filtering. If I only showed you the good columns, you'd have no reason to believe the good columns.

Claim 4 — what does it cost you?

Latency first, because this is the one that quietly ruins mixes. I measured the actual delay through the plugin by correlation and compared it against the number the plugin hands your DAW:

Delay through the plugin measured by correlation, against the figure handed to the host. Plotted as absolute disagreement in samples.

Configuration	Reported	Measured	Error
1x	72 smp	72.00 smp	0.00 smp
2x	121 smp	121.00 smp	0.00 smp
4x	132 smp	131.16 smp	0.84 smp
8x	136 smp	136.02 smp	0.02 smp
Fast	126 smp	126.08 smp	0.08 smp
CLEAN	213 smp	213.53 smp	0.53 smp

Worst disagreement anywhere in the engine: 0.84 samples. 17.5 microseconds. And the reported figure never changes after audio starts — which matters more than the accuracy, because a plugin that revises its latency mid-session is how phase relationships silently rot between one playback and the next.

That last table row is a design decision worth naming: THE_STRIP 3 MINI, the companion you put on every track, refuses settings that would incur latency rather than quietly introducing delay. Ask it for every lookahead and 8x oversampling — it reports zero samples because it delivers zero samples.

Now CPU, with every one of the fifteen modules at full depth:

The whole rack, running at once, is 3.42 % of one core. The MINI set is a quarter of one percent — the number that lets you put it on forty tracks without thinking about it.

And none of it drifts with sample rate:

One more, because nobody tests it: stereo

Width is easy to fake and expensive to do without wrecking the mono sum — which is exactly why it's worth measuring.

Width at 0.0 doesn't reduce the side signal, it removes it. Width at 2.0 widens around a mid that doesn't move. And bass-mono collapses the bottom without narrowing the top.

So where does the chain actually lose?

I haven't measured anybody else's plugins for this article. No rival numbers, no anonymised competitors, no ranking. I aimed at the theoretical limit instead, because the limit doesn't move and a competitor's marketing does — and nothing can be more correct than +6.00 dB for a +6 dB bell.

But one structural point is just arithmetic. A chain's errors add up. Ten plugins each within a very respectable 0.05 dB of unity can hand you half a decibel of drift you never dialled in. Ten latency reports each off by a couple of samples add up to a phase relationship that's quietly wrong. And you can't check any of it, because the number you'd need is the number for the assembled chain — and nobody publishes that.

The strip is one measured path. Those 666 checks don't run on fifteen modules in isolation — they run on the engine you actually insert, stages talking to each other, gain-staged to a single −12 dB reference so every threshold in the rack reads in real dB. A chain of great plugins is a sum of parts that were each measured alone. A strip can be measured whole. On this axis, it's not the compromise — it's the better instrument.

The two things that didn't pass

If I only publish the numbers that flatter me, none of the numbers mean anything. Two measurements missed the standard I hold the rest of the engine to. Neither is audible. Both are getting fixed in the next update.

1. The EQ's bells narrow near the top of the spectrum. The EQ runs at the base sample rate using the standard bilinear transform, so a bell centred at 16 kHz keeps 71.8 % of its 1 kHz shape at 44.1 kHz. It matches the textbook maths to 0.01 dB, the effect disappears entirely at 96 kHz, and on a real-material render it had no measurable effect at all. Some dedicated EQs solve this with a Nyquist-matched transform or by oversampling the EQ stage. This one doesn't yet — it will.

2. The CLIP stage's alias rejection doesn't improve with oversampling. It measures −26.9 dB at 8x; the brickwall limiter alone measures −58.0 dB, so the clip stage dominates when pushed hard. What does that cost on real material? I rendered a commercial drum loop driven +12 dB into the clipper at 100 % — far harder than anyone sane would push it — with anti-aliasing off versus maximum. The entire difference: 0.23 dB in the highs, 0.06 LU of loudness, and no change in dynamic range at all. A real finding, worth publishing, and not something you'll hear in a mix.

And a footnote, while I'm being honest: the compressor's attack and release times aren't calibrated to the 63 %-of-gain-reduction convention — a set 10 ms measures 23 ms. The knobs are monotonic and well-ordered, so faster is reliably faster, but I'd rather the printed numbers were a spec sheet too. Same list, same update.

I'd rather you heard all of this from me than found it yourself.

Test the same things on your own chain

To be clear about what's mine and what's yours: the acceptance suite that produced these numbers is my internal test rig inside THE_LAB. It hosts the exact engine that ships, and it's how a regression gets caught before it ever reaches an installer. The suite itself isn't in the box — but the tests aren't magic, and you can run the equivalent on your chain today.

Bypass everything and look at how flat a sweep comes back. Measure the actual slope of your favourite high-pass against the number printed on it. Feed your saturator an 8 kHz tone and look at what lands between the harmonics. You may find your chain is immaculate — genuinely, some of those plugins are superb, and I own several of them.

What you won't find is the thing the objection assumes: a strip that had to be watered down to fit fifteen modules in one window.

So: strip or chain?

Reach for a specific plugin when you want that plugin's character — no measurement in this article argues against taste, and I'd never tell you to give it up.

But the belief that a channel strip must be sonically second-rate because it's a channel strip? That's just not true anymore, and it isn't a matter of opinion. In THE_STRIP 3, a 24 dB/oct Butterworth measures 24.1 dB/oct. A +6 dB bell measures +6.00 dB. A 4:1 ratio compresses at 4.00:1. An unused module is bit-identical to no module. The saturation holds its worst artefact below the noise floor of 24-bit audio across 47 dB of input. The whole rack costs 3.42 % of one core, and the latency figure is honest to 17.5 microseconds.

666 checks. Four sample rates. 321 parameters. Zero failures — and two findings published instead of buried.

Load whatever gets you to the mix. Just stop paying for the assumption.

Frequently asked questions

Is a channel strip as good as chaining separate premium plugins?

It can be, and THE_STRIP 3 is measured to demonstrate it. For linear processing there is a correct answer rather than a better one: a 24 dB/oct Butterworth high-pass has one textbook response, and THE_STRIP 3 measures 24.1 dB/oct with a 0.04 dB corner error and no passband ripple. A +6 dB EQ bell measures +6.00 dB. A 4:1 compressor ratio measures 4.00:1. No separate plugin can be more correct than the target value.

Do unused modules in THE_STRIP 3 colour the sound?

No. With every module disabled, THE_STRIP 3 measures -0.0016 dB of magnitude deviation across 20 Hz to 20 kHz, -153.02 dB THD+N at 1 kHz, -240 dBFS self-noise on digital silence and 0 samples of latency. A module set to zero mix is bit-identical to the module being disabled: the worst single-sample delta is 0.000000000000.

How much aliasing does THE_STRIP 3's saturation produce?

At 8x CLEAN oversampling with an 8003 Hz tone at -1 dBFS and drive at 0.7, THE_STRIP 3's Console model measures -142.0 dB and the Transformer model -129.0 dB below the fundamental, both below the -144 dBFS noise floor of 24-bit audio. The alias floor also stays near -171 dB across a 47 dB range of input level rather than rising with the signal.

How much latency does THE_STRIP 3 add, and is it reported accurately?

THE_STRIP 3 reports 0 samples with all modules off, 72 samples at 1x oversampling and 213 samples in CLEAN filter mode. The largest disagreement between the reported figure and the delay measured by correlation is 0.84 samples, or 17.5 microseconds at 48 kHz, and the reported figure never changes after audio starts. THE_STRIP 3 MINI reports and delivers 0 samples even with every lookahead and 8x oversampling requested, because it refuses latency-incurring settings rather than introducing delay silently.

How much CPU does THE_STRIP 3 use?

On an Apple silicon Release build at 48 kHz, THE_STRIP 3 uses 0.16% of one core with all modules off, 2.39% for a typical mix chain and 3.42% with every one of its 15 modules enabled. THE_STRIP 3 MINI's module set costs 0.26% of one core.

What are the known measurement limitations of THE_STRIP 3?

Two findings did not meet the standard applied to the rest of the engine, and both are published rather than omitted. First, the EQ uses the standard bilinear transform at the base sample rate, so a bell centred at 16 kHz retains 71.8% of its shape at 44.1 kHz; this disappears at 96 kHz and had no measurable effect on a real-material render. Second, the CLIP stage's alias rejection does not improve with oversampling, measuring -26.9 dB at 8x. On real programme material driven 12 dB past normal use, the total difference is 0.23 dB in the highs. Neither is audible, and both are scheduled for a fix.

Does THE_STRIP 3 behave the same at every sample rate?

Yes. THE_STRIP 3 passes 666 automated checks at 44.1 and 48 kHz and 664 at 96 and 192 kHz, with zero failures. Neutral flatness measures 0.00 dB, a low cut at 100 Hz measures -3.00 to -3.04 dB and a +6 dB bell at 1 kHz measures +6.00 dB at all four rates.

Does THE_STRIP 3's stereo widening stay mono-compatible?

Yes. At width 2.0 THE_STRIP 3 adds +6.02 dB of side signal while leaving the mid channel at -0.00 dB, so the centre of the mix does not move. At width 0.0 the side signal collapses to -180 dB rather than merely being reduced, and the mid is still untouched. The bass-mono filter set to 200 Hz removes 41.87 dB from the side at 60 Hz while leaving the side at 4 kHz at -0.00 dB.

Are THE_STRIP 3's delay and reverb stable at maximum feedback?

Yes. At 0.95 feedback THE_STRIP 3's delay peaks at -1.41 dBFS and has decayed to -64.51 dBFS after 30 seconds. The reverb tail decays monotonically from -29.5 to -90.3 to -100.0 dBFS with no ring-up. All five modulation modes at full depth with 0.95 feedback produce no non-finite samples and stay between -6.03 and +0.80 dBFS.

Is THE_STRIP 3 numerically stable at extreme settings?

Yes. All 321 of THE_STRIP 3's parameters were swept to both extremes with every output verified finite and bounded, 321 of 321 passing. The full chain under loud noise, impulse, DC and digital-silence stress produced no non-finite samples, and the output stayed at -0.83 dBFS against the limiter ceiling.

Is THE_STRIP 3's EQ minimum phase, and does it add latency?

THE_STRIP 3's EQ is minimum phase and adds 0 samples of latency. Its high-pass corner measures -3.01 dB against a -3.00 dB target with a 12.03 dB/oct slope against a 12.00 dB/oct nominal, and the curve drawn in the plugin's editor agrees with the rendered audio to within 0.09 dB across the whole band.

What is measured in THE_STRIP 3's DSP report?

The report covers 666 automated checks over 321 parameters and 15 modules at four sample rates: neutral-path flatness and noise, bypass bit-identity, Butterworth filter conformance, EQ gain and phase accuracy, dynamics transfer law, stereo width behaviour, time-domain stability, saturation alias rejection, latency reporting accuracy, numerical stability, sample-rate independence, CPU cost and a real programme-material render, plus two open findings published in full.