Who publishes this, and how to catch us out

Black Ops Suppression Systems makes suppressors, and most of the products ranked here are our competitors. That is the obvious objection to this tool, so here is the answer to it. Every product tested is shown; none is withheld. The default sort favours nothing of ours. Our own results are ranked by the identical formula and land where the measurements put them.

The scoring is published in full below and the raw measurements are Thunder Beast’s own, downloadable from their site. Anyone can take that file and check every number on this page. That is the point: a ranking that hides its own bad results is not worth reading, and one you can audit is.

How your pick becomes a ranking

There are no sliders, because nobody knows what "weight: 3 of 5" means, and there is no ordering to set either. You pick one thing and the list reorders around it. Clicking it again clears it.

An earlier version let you rank a first, second and third choice against each other. It worked, but it asked you to hold an ordering in your head to read the result, and almost every question people actually bring to this data is a single one — quietest, lightest, cheapest. One pick answers that without the arithmetic.

Most picks map to a single measurement, and when they do, that measurement's column is tinted in the table so you can see what is driving the order after the header has scrolled away. A few read more than one, because they are really one question. Smaller the better is mostly bulk but partly length and diameter (0.4 and 0.3 of its weight); Suppression for the size is the form-factor figure alone — blending plain volume into it made it agree with smaller the better 85% of the time, which defeated the point of having both.

What “Overall value” actually does

It is the broadest pick on the page and the only one that touches seven measurements at once, so here are its exact weights rather than a description of them:

Street price	1.00	29% of the total
Muzzle sound	0.50	the other 71%
Weight	0.45
Volume	0.45
Sound at the ear	0.35
Mount freedom	0.35
Suppression for the size	0.30

Price is the largest single share at 0.29 of the total, but the other six together outweigh it more than two to one. That is deliberate. Someone asking for value wants a can that is decent everywhere and does not cost much — not the highest decibels-per-dollar figure, which a loud cheap can wins outright. That ratio is a real thing people want, so it has its own separate button (Most sound suppression for the money); it is not folded in here.

Two consequences worth knowing before you use it. A product with no confirmed selling price cannot be ranked on this pick at all and drops below the ones that can — {{NO_PRICE_N}} of {{SKU_N}} products are in that position, and the list says so where it happens. And because it spreads across seven axes, it rewards the all-rounder: a can that is excellent at one thing and poor at the rest will rank higher under a single-measurement pick than under this one.

Your pick is the only thing that orders the list. An earlier version kept a small residual weight on sound even when you had not asked for it, so that a sound-comparison tool never entirely ignored sound. In practice that put a $407 suppressor above a $370 one under a button labelled Cheapest first, because the cheaper one was louder — the tool quietly answering a question you had not asked. It is gone. The sound columns are still on the table whatever you pick, so nothing is hidden; they just no longer move the order unless you choose them.

The formula

Within the selected host class, each axis you weight is scaled 0–1 against the best and worst value present in that class, with direction applied so 1 is always better. The score is the weighted mean of those scaled values: score = Σ(wᵢ × nᵢ) / Σ(wᵢ), over the axes a product actually has data for, where the weights come from the pick above. It is then shown out of 100.

Value: suppression per dollar

Cheapest and best value are different questions, so they are different buttons. On the 5.56 class the cheapest can is the {{CHEAP_NAME}} at {{CHEAP_PRICE}}, and it buys {{CHEAP_DB}} dB; {{STEP_EXTRA}} more buys the {{STEP_NAME}} at {{STEP_DB}} dB. Value is dB below bare muzzle ÷ price, shown per $100.

Checked like every other derived figure here: it correlates {{CORR_VPD_PRICE}} with price and {{CORR_VPD_SUPP}} with suppression. That leans on price harder than form factor leans on volume, so treat it as "cheap, weighted by how well it works" rather than a wholly independent axis — which is why picking it also carries a little plain price.

The striking number is the other one. Suppression correlates just {{CORR_SUPP_PRICE}} with price on 5.56 and 0.23 on .308: across this field, paying more barely buys quieter.

Form factor: suppression per cubic inch

How much quieting you get for the bulk you carry — dB below bare muzzle ÷ cubic inches, using published factory dimensions. A big can that is quiet is unremarkable; a small one that is quiet is the interesting result.

This is the muzzle figure only. Sound at the ear is not in it. That is deliberate rather than an oversight: on a gas gun the two are very nearly independent — within the 5.56 16″ AR class they correlate just {{CORR_MUZZLE_EAR_AR}}, because what reaches your ear is dominated by gas out of the ejection port rather than by the muzzle. Averaging them into a single "suppression" number would blend two things that measure different phenomena and hide both. If sound at the ear is what you care about, that has its own button and its own column.

Derived metrics are easy to fool yourself with, so this one was checked the same way we checked and then rejected a backpressure score. Within the 5.56 16″ AR class it correlates {{CORR_EFF_VOL}} with volume and {{CORR_EFF_SUPP}} with suppression — so it is neither an inverted size measurement nor a restatement of how quiet the can is, but a third thing. (The backpressure index we threw out correlated −0.93 with the muzzle number; that is what a metric that is secretly one of its own inputs looks like.)

Because it divides by a small number it is very sensitive to a wrong dimension, so volume is density-checked first: a suppressor is a hollow tube, and real ones fall between 0.6 and 1.5 oz/in³. Anything implausible gets no figure rather than a spectacular false one.

How each measurement is scaled

Every measurement is scaled 0–1 within the class before weighting, and only the losing tail is clipped — to the 5th or 95th percentile depending on which direction is better.

Clipping the bad end matters: scaling to the true extremes puts 82% of the 5.56 field into the top 40% of the scale, because one result at 158 dBA stretches a range whose middle half spans 6 dB. Everything bunches and the ranking stops responding to what you asked for.

Clipping the good end turned out to be a mistake, and it is not done. Diameters cluster between 1.49″ and 1.75″, so a genuinely slim 1.00″ can was being clamped to the same score as an average one and ranked below it on the tie-breakers. Standing out is the thing worth finding, so the winning tail keeps its true extreme.

What is on the table

Every measurement the ranking is currently using gets its own column, so the inputs are visible rather than buried — a weighted list that will not show its working is asking to be taken on trust. Columns follow your picks; both microphone positions are always there, labelled, because they are the two numbers the summit exists to produce. Show every measurement puts all of them on at once.

Why muzzle and ear are separate columns

On gas guns, muzzle sound and sound at your ear are close to independent — within the 5.56 16″ AR class they correlate at just 0.25, because what reaches your ear is dominated by gas out of the ejection port rather than by the muzzle. Ranking them separately is the only way to express "I want it quiet for the neighbours" versus "I want it quiet for me". On bolt guns there is no port and the two move together (0.77), so the ear figure largely follows the muzzle one.

Missing data is shown, never invented

A product is scored only on the axes it has values for, and its data coverage is shown on every row. No value here is estimated, averaged or inferred.

That rule has a sharp edge: a product missing an axis is not penalised on it. If you say weight matters most, a can whose weight nobody publishes would be scored on its other axes and could outrank a genuinely light one — a free pass on the very thing you asked about. So the priority you pick also decides who is eligible: only products we have that figure for are ranked, and the count above the table tells you how many that is. Switch it off in the filters if you want to see everything, knowing the incomplete rows are flattered.

Where a figure comes from

Weight, length and diameter are the manufacturer's published specification wherever one has been found. Where none has, the summit's own measurement stands in, marked with an asterisk. Those are taken as submitted, with whatever mount came in the box — the same model has arrived up to {{SPREAD_OZ}} oz apart across years — so they are weaker for comparing one product against another and are replaced as soon as the published number turns up. They are shown because a measured figure beats an empty cell.

Sorting by hand

Sort manually stands the priorities down and orders the list by a single measurement — click any column heading, click again to reverse. Choosing a priority again hands the ordering back to the weighted score. Products with no figure for the column you sorted on go to the end, not the top.

The old note on factory specs

Thunder Beast weighs and measures each suppressor as submitted, with whatever mount the manufacturer supplied, so those figures are not comparable between products — the same model has arrived up to {{SPREAD_OZ}} oz apart in different years. Size and weight filtering uses published factory specifications; the as-tested figures are shown in the detail view for cross-checking only.

What is deliberately not here

There is no backpressure score. The mechanism is real, but no single number in this dataset measures it: sound at the ear minus muzzle sound is mathematically an inverted muzzle number, and it ranks known-loud cans as low-blowback. Acoustic energy at the ear is not the same thing as gas flow to your face.

2023 results are excluded. Thunder Beast ran a proposed 2 m 225° microphone at the 2023 summit and replaced it with milspec-right from 2024 onward. Muzzle and ear readings are on the same two microphones throughout, so those would carry over — but port asymmetry is milspec-left against milspec-right, and no 2023 test can produce it. Rather than publish a year that silently supports a different set of columns, the table stays on the consolidated 2024–2026 dataset that Thunder Beast publishes as one set.