The Generator Noise Normalization Index
There is no enforced standard distance for measuring a generator's dB(A) noise rating — only an informal 23 ft (7 meter) norm that some manufacturers state and many don't. This index normalizes every claim we could verify a test distance for to that common 23 ft reference using the inverse-square law, and flags every model where the distance is undisclosed as non-comparable, rather than pretending a number means more than it does.
The one thing to know: A generator's dB(A) rating is only meaningful alongside the distance it was measured at, because sound intensity falls off with distance under the inverse-square law: moving twice as far away roughly halves perceived sound energy, a drop of about 6 dB. An informal industry norm has settled on 23 feet (7 meters) as a comparison distance, but it is NOT an enforced standard — some manufacturers state it, some test closer (making their number look louder than it would be at 23 ft), and many don't disclose a distance at all, which makes their number impossible to compare fairly to one that does.
License: CC BY 4.0 — free to use with attribution.
The index
| Model | Claimed dB(A) | Claimed distance | Load condition | Normalized @ 23 ft | Flag |
|---|---|---|---|---|---|
| Honda EU2200i | 57 dB | 23 ft | rated load | 57 dB | baseline — already stated at the 23 ft industry-norm distance |
| Yamaha EF2000iSv2 | 61 dB | 23 ft | rated load | 61 dB | baseline — already stated at the 23 ft industry-norm distance |
| Westinghouse iGen2200 | 52 dB | 20 ft | 25% load | 50.8 dB | tested slightly closer than 23 ft — normalized value is ~1.2 dB quieter than the raw claim |
| Cummins Onan (RV gasoline generator) QG 2800i | 65 dB | 10 ft | half load | 57.8 dB | tested much closer than 23 ft (10 ft) — the headline '65 dB' overstates how loud it is at the industry-norm distance by ~7 dB |
| Champion 3400 Dual Fuel (inverter) | 64 dB | not disclosed | rated load (per aggregator) | not computable | not computable — no test distance found in manufacturer materials; treat the 64 dB figure as non-comparable to distance-stated models |
| Generac GP2200i | 58 dB | not disclosed | ~30% load (per aggregator) | not computable | not computable — third-party measurements for this model range from ~50 dB to ~70 dB depending on load and source, with no consistently stated test distance |
| WEN 56235i | 51 dB | not disclosed | quarter load | not computable | not computable — WEN's own spec materials state the load condition but not a test distance, making this figure impossible to normalize or fairly compare |
| DuroMax XP4400E | 69 dB | not disclosed | full throttle | not computable | not computable — no test distance published; this is the loudest claimed figure in this index and should be treated as a rough indicator only |
| Predator 3500i | 57 dB | not disclosed | not stated | not computable | not computable — retail/aggregator sources report 57 dB with no test distance or load condition specified |
How the normalization works
Sound intensity from a small source falls off with distance under the inverse-square law: doubling the
distance drops sound pressure level by about 6 dB. The normalization formula used here is
dB_at_23ft = claimed_dB + 20 × log10(claimed_distance_ft ÷ 23). A generator tested
close up (say, 10 ft) and reported at 65 dB will normalize to a lower, quieter figure at 23 ft — and a
generator tested farther away than 23 ft would normalize to a louder figure, which is exactly how a
manufacturer could make a unit look quieter than it really is at a typical listening distance by testing
farther from the machine.
Caveats — read before citing a normalized figure
- We did not measure any of these generators ourselves. Every claimed dB(A) figure and distance comes from a manufacturer spec sheet or, where noted, a secondary aggregator source with no verifiable primary distance.
- Normalization assumes free-field, point-source conditions. Real placement (near a wall, in a partial enclosure, on hard pavement vs grass) changes perceived loudness and isn't captured here.
- Load condition is not standardized either. Inverter generators are commonly tested near 25% load and conventional generators near 50% load, but this convention isn't universally disclosed, so even normalized figures at different load conditions aren't perfectly apples-to-apples.
- Normalization does not capture tonal character. A whining or tonal engine note can be more annoying than a steady hum at an identical measured dB(A).
- "Not computable" does not mean "loud" or "quiet" — it means the claim cannot be fairly compared to a distance-stated model. Treat those raw figures as rough indicators only.
Methodology & versioning
Version 1.0 — published July 30, 2026. Claimed dB(A) figures and distances are drawn from manufacturer spec sheets (Honda, Yamaha, Westinghouse, Cummins Onan) where available, and from aggregator/ review-site reporting (topnotchoutdoor.com, generatorbible.com, generatormag.com, eeeproject.com, generatorgrid.com, bobistheoilguy.com) where a primary spec sheet with a stated distance could not be found. Normalization applies the standard inverse-square-law formula above; we do not sound-test generators ourselves. Corrections: if a manufacturer publishes a previously-undisclosed test distance or we find an error, this page and the CSV are updated in place and the version/date bumped — see How We Evaluate for our full editorial policy.
How to cite: Generator Decoded, "Generator Noise Normalization Index," version 1.0, July 30, 2026. https://generator-decoded.pages.dev/generator-noise-normalization-index/
Machine-readable download: the full table is available as CSV at /generator-noise-normalization-index.csv (version-pinned: /generator-noise-normalization-index-v1.0.csv), and a condensed digest is maintained in /llms-full.txt.
This page reports published noise specifications and a documented physics normalization, not independent acoustic lab-test results.
Frequently Asked Questions
A reference table that takes published dB(A) noise claims for popular generator models and, where a test distance is actually stated, normalizes them to a common 23 ft reference distance using the inverse-square law. Where no distance is disclosed, we mark the figure as not computable rather than guess.
23 feet (7 meters) is an informal norm some manufacturers (notably Honda and Yamaha) already state for their dB(A) claims, which makes it a useful common reference point. It is not an enforced industry standard, and many manufacturers don't use it or don't say what distance they used at all.
The inverse-square-law math requires knowing the original test distance. For models where the manufacturer or a primary spec sheet doesn't publish that distance (several of the most popular budget models in our review), there's nothing to normalize from — we report the raw claim and flag it as non-comparable rather than invent a distance.
Not necessarily. Normalization corrects for distance only, not for tonal character. A high-pitched, whining engine can be more irritating at the same measured dB(A) than a low, steady hum. Two units at an identical normalized figure can still sound different to a neighbor.
It makes the generator look quieter than it would at the industry-norm distance, because sound intensity falls off with distance. This is the documented complaint behind this dataset: comparing a manufacturer that tests at, say, 50 feet against one that tests at 10 feet produces two numbers that are not measuring the same thing, even though both are presented as a simple 'dB' figure.