Report · computed live from the catalog

The Spec-Sheet Honesty Report

Our pipeline stores a first-class “unstated” wherever a spec sheet goes quiet — which means we can count the silence. This page does exactly that across all nine categories we track, per category and per brand, recomputed from the live catalog.

The short version

Of the 481 speakers in our catalog with a published sensitivity figure, 87% record no measurement conditions. Only 169 of 562 speakers disclose minimum impedance — the number an amplifier actually fights. 5 of 47 receiver power ratings omit channels-driven, and 34% of subwoofer depth claims come without a dB tolerance. On the video side, 146 of 196 TVs state whether VRR is supported, against 11 of 188 projectors — the same question, answered at very different rates.

Why we can measure this at all

Most spec databases silently drop what a manufacturer doesn’t state. Ours can’t: every value is stored with its measurement context as an atomic pair, and a missing context is recorded as the literal value “unstated” — never guessed, never defaulted (our methodology). One honest caveat about what’s counted: “not stated” below means our normalized record carries no value for the field — either the sources we track don’t state it, or no source we track covers it. We count our records, not manufacturers’ intent; the pipeline’s refusal to guess is what makes the two line up.

Speakers: sensitivity numbers without conditions

481 of the 562 speakers we track publish a sensitivity figure, and 418 of those 481 record no measurement conditions. That matters because an in-room figure reads higher than an anechoic one for the same speaker — a “90 dB” with no conditions attached is a softer claim than an anechoic 87 (how to read sensitivity and impedance). Minimum impedance is quieter still: 169 of 562 disclose it, even though the dip — not the nominal rating — is what your receiver has to drive.

BrandSpeakers trackedTrackedState sensitivitySensitivity…with conditions…w/ conditionsState min. impedanceMin. imp.
ELAC7770 of 770 of 7020 of 77
KEF6867 of 680 of 6725 of 68
Bowers & Wilkins6633 of 660 of 3333 of 66
Aperion Audio5251 of 520 of 510 of 52
Polk Audio4944 of 490 of 4411 of 49
Wharfedale3939 of 390 of 3933 of 39
Definitive Technology3030 of 300 of 300 of 30
Arendal Sound2626 of 260 of 260 of 26
Q Acoustics2019 of 200 of 1920 of 20
Emotiva99 of 90 of 90 of 9
Fluance88 of 80 of 80 of 8
JBL66 of 60 of 60 of 6
RSL44 of 40 of 40 of 4
Klipsch4111 of 412 of 110 of 41
Power Sound Audio1212 of 1211 of 120 of 12
PSB2725 of 2725 of 2527 of 27
SVS1313 of 1313 of 130 of 13
Ascend Acoustics1212 of 1212 of 120 of 12
Sorted by share of stated values missing their context, worst first. Brands with fewer than 3 tracked models are left out — too few rows to characterize a brand honestly.

Receivers: the watts-per-channel game

47 of 64 receivers publish a power rating; 5 of those ratings omit how many channels were driven, and only 42 state distortion (THD). The conditions are the claim: a single channel driven into an easier load can read roughly double a two-channel 8Ω rating for a comparable amplifier — which is why we refuse to compare bare wattage numbers anywhere on this site (how much power you actually need).

BrandReceivers trackedTrackedState powerPower…with channels driven…w/ ch. drivenState THDTHD
Anthem31 of 30 of 10 of 3
NAD32 of 30 of 20 of 3
Yamaha66 of 65 of 66 of 6
Onkyo98 of 97 of 87 of 9
Denon1717 of 1717 of 1717 of 17
Marantz116 of 116 of 66 of 11
Sony72 of 72 of 22 of 7
Emotiva41 of 41 of 10 of 4
Pioneer44 of 44 of 44 of 4
Sorted by share of stated values missing their context, worst first. Brands with fewer than 3 tracked models are left out — too few rows to characterize a brand honestly.

Subwoofers: how low, at what tolerance

143 subwoofers we track claim a low-frequency extension, and 48 of those claims arrive without a dB tolerance. The tolerance is the whole claim: “20 Hz” at −10 dB and “20 Hz” at ±3 dB describe very different subwoofers (sealed vs ported, and how low is low enough).

What the sheet is askedState itStatedShare%
Do they publish a low-frequency extension figure?
Asked of 184 subwoofers tracked
143 of 18478%
…and of those, do they state the dB tolerance it was measured to?95 of 14366%
Every subwoofer we track that is not delisted. The tolerance row is asked only of the products that make a depth claim — a subwoofer that publishes no extension figure cannot omit its tolerance.

TVs and projectors: the gaming specs, side by side

This is the comparison the catalog was always able to make and this page could not, until the report learned all nine categories. VRR and ALLM are asked the same way of both kinds of screen and stored in the same shape, so the two columns are genuinely like-for-like. Of the 196 TVs we track, 146 state whether VRR is supported and 103 state ALLM. Of the 188 projectors, 11 state VRR and 22 state ALLM. Note what “state” means here: a sheet that says “VRR: not supported” has answered the question and counts as stated. Only silence is silence.

Input lag is the one number we will not put in that head-to-head. Our schema stores a TV’s figure at 4K/120 and a projector’s at 4K/60, because those are the modes each category publishes — different measurement conditions, so the counts are not comparable and lining them up would be the exact error this report exists to expose. Separately, then: 5 of 196 TVs state a 4K/120 lag figure, and 56 of 188 projectors state one at 4K/60.

What the sheet is askedState itStatedShare%
Do they state whether VRR is supported?
Asked of 196 TVs tracked
146 of 19674%
Do they state whether ALLM is supported?
Asked of 196 TVs tracked
103 of 19653%
Do they state input lag at 4K/120?
Asked of 196 TVs tracked
5 of 1963%
Do they state whether VRR is supported?
Asked of 188 projectors tracked
11 of 1886%
Do they state whether ALLM is supported?
Asked of 188 projectors tracked
22 of 18812%
Do they state input lag at 4K/60?
Asked of 188 projectors tracked
56 of 18830%
TVs first, then projectors. Read VRR against VRR and ALLM against ALLM; the two input-lag rows are measured at different refresh rates and are not a like-for-like pair.

Projection screens: a gain figure with nowhere to stand

711 of the 964 screens we track publish a gain figure, and 275 of those publish no half-gain viewing angle to go with it. Gain is a directional measurement: a 2.4-gain material that halves twenty degrees off axis and one that holds to sixty are different products with the same headline number, and only the angle tells you which you bought. Acoustic transparency — whether speakers can live behind the screen — is stated on 333 of 964.

What the sheet is askedState itStatedShare%
Do they publish a gain figure?
Asked of 964 projection screens tracked
711 of 96474%
…and of those, do they state the half-gain viewing angle?436 of 71161%
Do they state whether the material is acoustically transparent?
Asked of 964 projection screens tracked
333 of 96435%
Screen rows are series/material sheets rather than individual sized SKUs, which is how manufacturers publish them.

Media players: “4K” is half a spec

42 of the 42 media players we track publish a maximum video resolution, and 13 of those list no output frame rates at all. A box that does 4K but cannot output 24p will judder every film you own, and the resolution claim alone will never tell you.

What the sheet is askedState itStatedShare%
Do they publish a maximum video resolution?
Asked of 42 media players tracked
42 of 42100%
…and of those, do they state which frame rates they output?29 of 4269%
Do they state whether Dolby Atmos is supported?
Asked of 42 media players tracked
30 of 4271%
Frame rates are asked only of the players that publish a resolution claim to qualify.

Acoustic treatment: one number standing in for a curve

92 of the 242 acoustic treatment products we track publish an NRC rating, and 47 of those publish no complete 125 Hz / 500 Hz / 2 kHz set of absorption coefficients behind it. NRC is an average, and the average is precisely where a panel’s low-frequency behaviour disappears. Then there is what the price buys: 178 of 242 state whether they are sold as a single panel or a pack, and of the 62 sold as a multi-unit pack, 32 resolve to no panel count at all. Products sold as a single panel are deliberately not counted as non-disclosers — a single panel cannot omit its pack count, and scoring it as silent would be the same absence-is-not-a-no error we are measuring.

What the sheet is askedState itStatedShare%
Do they publish an NRC rating?
Asked of 242 acoustic treatment products tracked
92 of 24238%
…and of those, do they also publish the 125 Hz / 500 Hz / 2 kHz absorption coefficients?45 of 9249%
Do they state whether they are sold as a single panel or a multi-panel pack?
Asked of 242 acoustic treatment products tracked
178 of 24274%
Do they state how many panels the pack contains?
Asked of 62 acoustic treatment products sold as a multi-unit pack
30 of 6248%
The pack-count question is scoped to the products that say they are a multi-unit pack.

Seating: the width you can’t plan a row with

61 of the 105 seating models we track publish a total row width, and 47 of those publish no per-seat width. Total width includes the arms, so it is the number that flatters and not the number you lay out a room with. 49 of 105 state a per-seat weight capacity, and 83 of 105 state whether the recliner clears a wall.

What the sheet is askedState itStatedShare%
Do they state a per-seat weight capacity?
Asked of 105 seating models tracked
49 of 10547%
Do they publish a total row width?
Asked of 105 seating models tracked
61 of 10558%
…and of those, do they also publish the per-seat width?14 of 6123%
Do they state whether the recliner clears a wall?
Asked of 105 seating models tracked
83 of 10579%
Seating rows must state three of the four planning numbers — per-seat width, fully-reclined depth and seat count — to be listed at all; per-seat weight capacity may be absent, in which case it is stored as unstated rather than guessed, which is why it can still be counted here.

Method, and using these numbers

Counts are recomputed from the live catalog (cached hourly), so they change as coverage grows — a brand’s row can improve the day it publishes better sheets, or the day we ingest a better source. Delisted products are excluded throughout, from the totals as well as the numerators. Brand tables require at least 3 tracked models, and exist only for speakers and receivers, where one headline number carries one companion condition and one compat-critical extra; the other seven categories are published at category level rather than given an invented third column. Per-product provenance is visible on every product page, spec by spec. Please quote us: take the number you need, note the date you retrieved it, and link back to this report — the figures moving as the catalog grows is a reason to date your quote, not a reason to skip it. Downloads, license terms and pre-written citation blocks are on our data page. Wrong somewhere? Every value traces to a source; tell us and we’ll fix the record, not the story.