We read the spec sheets for 2547 pieces of home theater gear. Here’s what the makers won’t tell you. (2026-08 edition)

I run a small site that does one unglamorous thing: we read the official spec sheet for every product in our catalog, record exactly what the manufacturer states, and refuse to fill in the blanks. No estimates, no “probably.” If a maker doesn’t publish a number, our database stores “unstated” and the page says so.

When I first counted, on July 18, 2026, the catalog held 340 products. It now holds 2547, across nine categories, and the blanks have kept their shape. Two things are worth saying before the numbers. A stated no counts as stated: a projector sheet that says “VRR: not supported” has told you something, and only silence is counted as silence. And “not stated” means our record carries no value — either the sources we track don’t state it, or no source we track covers it. We count our records, not intent; the pipeline’s refusal to guess is what makes the two line up.

1. If you game, the projector sheets still abandon you

In July this was the study’s hardest line, and it was drawn from a few dozen models: 0 of 38 projectors stated whether VRR works, against TVs managing it on almost every model (July 18, 2026 edition). The catalog is several times larger now, and the asymmetry is still there. Of the 188 projectors we track, 11 of 188 state whether VRR is supported and 22 of 188 state ALLM. Of the 196 TVs, 146 of 196 state VRR and 103 of 196 state ALLM. Same feature, same question, two industries answering it at completely different rates — so nobody can claim the field is too obscure to print.

Input lag is the one number I refuse to put in that comparison, and the reason is the whole point of this site. Our schema stores a TV’s lag at 4K/120 and a projector’s at 4K/60, because those are the modes each category publishes — two different measurement conditions, which means the counts are not like-for-like and anyone who lines them up is doing the thing this study is about. Separately, then: 5 of 196 TVs state a 4K/120 figure, and 56 of 188 projectors state a 4K/60 one. Both are the latency number a console owner actually needs, and neither industry is generous with it.

And when the sheets did state it, it wasn’t always true. In 2020, an entire generation of Denon, Marantz, and Yamaha receivers shipped with an HDMI 2.1 chip that couldn’t actually pass 4K/120 from an Xbox Series X. Spec sheet: compliant. Reality: black screen. The fix required a free external adapter box, and in some cases a physical board swap. Every current model we track is a later generation and unaffected, but it is the clearest proof that “HDMI 2.1” on a sheet is a version number, not a promise.

2. The watts game, and the rule that was just tightened

Of the 64 receivers we track, 47 of 64 publish a power rating. That part is easy. The conditions are the claim: 42 of 47 of those ratings say how many channels were driven, and 42 of 47 state the distortion the rating was measured at. A single channel driven into an easier load can read roughly double a two-channel rating for a comparable amplifier, which is why we refuse to compare bare wattage anywhere on this site (how much power you actually need).

There is a regulator in this story, and the popular version of it is wrong. The FTC’s Amplifier Rule was not quietly dropped in 2024. It was strengthened: the Commission issued final amendments in June 2024, effective 12 August 2024, adding uniform testing methods and multichannel guidance — and in October 2024 the Consumer Technology Association petitioned to have the amendments apply only to products going forward. So the rule got tighter and the industry pushed back, which makes the obvious question “who states their power honestly?” a question with an answer. The counts above are that answer, and the per-brand version is in the honesty report. (Dates and status per our research file, backlink-research-assets-pr-2026-08-05, §1.3, as of 5 August 2026; the rule is federal and the amendments are on the FTC’s public record.)

The July edition’s receiver finding — 14 of 40 receivers stating an HDMI 2.1 port count, and a $1,299 model disclosing less than its own $550 sibling — is preserved in that edition with its date. Port-count disclosure is not yet one of the fields the live engine counts, so I am not going to restate it as if it were current.

3. The sensitivity number you’re comparing is not one number

481 of 562 speakers publish a sensitivity figure, and it looks wonderfully comparable: 85 dB, 88 dB, 94.5 dB. Then you ask how it was measured, and only 63 of 481 say. The two common methods differ by roughly 3 dB, and 3 dB is the entire audible gap between “efficient” and “hungry” (how to read sensitivity and impedance). Those numbers get compared side by side in a thousand forum posts as if they came from the same ruler.

The other half of the amplifier question is quieter still: 169 of 562 speakers state a minimum impedance, the number that decides whether a budget receiver can drive them safely. The rest publish a nominal rating and let you find out.

4. Every measured number has a companion, and the companion goes missing

This is the pattern underneath all of it. A value without its measurement context is not a smaller claim; it is a different kind of claim, one that can’t be checked. Four categories tell the same story with different fields.

Subwoofers: 143 of 184 claim a low-frequency extension and 95 of 143 of those claims arrive with a dB tolerance. “20 Hz” at −10 dB and “20 Hz” at ±3 dB describe very different subwoofers (sealed vs ported).

Projection screens: 711 of 964 publish a gain figure; 436 of 711 of those also publish the half-gain viewing angle. A 2.4-gain screen that halves by twenty degrees off axis is a different product from one that holds to sixty, and the gain number alone cannot tell you which you bought.

Media players: 42 of 42 publish a maximum video resolution, and 29 of 42 of those say which frame rates they can actually output. “4K” without 24p is half a spec if you watch films.

Acoustic treatment: 92 of 242 publish an NRC rating, and 45 of 92 of those publish the 125 Hz, 500 Hz and 2 kHz absorption coefficients behind the average. NRC is one number standing in for a curve, and the curve is where a bass trap either works or doesn’t.

5. Nobody will tell you how many you’re buying

The July edition’s version of this was about speakers sold in pairs. The engine now counts it where the schema records it explicitly: 178 of 242 acoustic treatment products state whether they are sold as a single panel or a pack. Of the 62 that name a multi-unit pack, 30 of 62 resolve to an actual panel count. A pack of “acoustic panels” whose count you can’t determine is a price you can’t compare.

Note what we deliberately do not count here: a product sold as a single panel cannot omit its pack count, so it is never scored as a non-discloser. Treating an absent pack count as a failure would be the same absence-is-not-a-no error this whole study is about.

6. Theater seating specs are still the emptiest

Of the 105 seating models we track, 49 of 105 state a per-seat weight capacity and 83 of 105 state whether the recliner clears a wall. The recurring trick is the width one: 61 of 105 publish a total row width, and 14 of 61 of those also publish the per-seat width — the only one of the two you can plan a room with. The July edition has the example that makes it concrete: a 23.5 inch seat inside a 34.5 inch chair (July 18, 2026 edition). Plan a three-seat row off the wrong number and the mistake is measured in feet.

7. Who does it right

A league table that only shames is a worse dataset than one that also names the brands doing it properly, so: these are the brands, computed from the same live tables, that publish the headline value on every model we track, state its measurement conditions every single time, and publish the compat-critical companion field on every model. A brand needs at least 3 tracked models to appear at all, because fewer than that cannot characterize a brand honestly.

Speakers: no brand currently clears the complete-disclosure bar — sensitivity on every model, its measurement conditions every time, and a minimum impedance on every model. The brands that come closest are the condition-complete ones below.

Speakers, condition-complete: PSB, SVS and Ascend Acoustics never publish a sensitivity figure without saying how it was measured, even where the minimum-impedance column has gaps. That is the single most useful habit a speaker maker has, and it costs a line of text.

Receivers, complete disclosure: Denon and Pioneer. Power rating on every model, channels driven stated every time, and a THD figure on every model.

If your brand is in the other half of a table and you think we have the wrong source, send us the sheet. Every value we hold traces to a document, and we would rather fix the record than keep the story.

8. Silence is not a no

The deepest problem isn’t any single blank. It’s that a blank is ambiguous. Sometimes “no VRR listed” means no VRR. Sometimes it means the copywriter didn’t bother. A buyer can’t tell the difference, which is why we store absence as “unstated” instead of “no,” and why our compatibility checker reports “we can’t verify this” instead of guessing.

Across 2547 tracked products the pattern is consistent: makers publish the numbers that flatter and omit the numbers that decide. The fix isn’t more marketing. It’s boring: state the measurement method, state the channels driven, state what the price buys. Until then, the blanks are the most informative spec on the sheet.