Shop Noise by Machine: Decibels, NRR Ratings and What Protection Really Costs

I can still picture the first shop I set up in a corner of the basement: a contractor saw, a small jointer, a router table I'd built out of plywood scraps. The noise never bothered me much. I wore earmuffs when I remembered, which was maybe half the time, and I figured that was plenty. What I didn't have, and what most of us don't have, was any real sense of what those machines were actually doing to my ears.

That's the gap this piece aims to close. Not with scare tactics, but with the numbers that exist, the ones that don't, and an honest look at what a pair of plugs or muffs buys you against the 85 dB line that hearing researchers keep pointing to.

The Line in the Sand: 85 Is a Recommendation, 90 Is the Law

Here's the confusing part right up front. NIOSH, the federal health research agency, sets its recommended exposure limit at 85 dBA as an eight-hour time-weighted average, with a 3 dB exchange rate (CDC/NIOSH, "Understand Noise Exposure"). OSHA, the enforcement agency, sets its legal limit at 90 dBA over eight hours with a 5 dB exchange rate (WoodworkingCove, "How Loud Is Woodworking, Really?"). Two different numbers from two different agencies, and neither one is wrong. NIOSH is telling you where hearing damage starts. OSHA is telling you where the fine begins.

The 85 dBA figure is the one worth carrying in your head. NIOSH also offers a non-instrument test that any woodworker can apply: noise is likely hazardous if you have to raise your voice to speak with someone at arm's length (CDC/NIOSH). No meter needed.

And every 3 dB matters disproportionately. Sound energy doubles with each 3 dB increase, per Fine Woodworking's David Heim, writing in issue #189. That doubling compresses your safe time dramatically, as the NIOSH exchange rate table makes plain:

Level	Safe daily allowance
85 dBA	8 hours
88 dBA	4 hours
91 dBA	2 hours
94 dBA	1 hour
97 dBA	30 minutes
100 dBA	15 minutes

That table comes from WoodworkingCove applying the NIOSH 3 dB exchange rate, and it's the single most useful thing in this article. A router at 94 dBA uses a full day's allowance in about an hour, and that's an hour of the tool actually running, not an hour of you standing in the room.

What the Stations Actually Measure

Now the hard part: honest numbers are harder to come by than you'd think. The most citable figures come from the UK Health & Safety Executive's research report RR1223 (2025), which compiled operator-position levels from real industrial operation. Table, bench and ripping saws land at 100–110 dB(A). Planing and moulding machines at 100–105 dB(A). Band saws at 95–100 dB(A). Circular saws generally, citing WorkSafe Victoria, at 80–120 dB(A). Those are summarised by Snap Decibel Meter, which cites RR1223 directly.

Against that, your table saw's manual probably says something like 90–94 dB(A). Makita's 2704 variants declare 90, 91 and 94 dB(A) at the operator with 107–108 dB(A) sound power, measured to EN 62841-3-1 with a stated uncertainty of K = 3 dB (Snap Decibel Meter, quoting Makita manuals). The manual isn't lying. It's measuring a different thing under a defined test cycle, and that declared figure is not a prediction of what cutting a particular board will sound like.

The gap matters. How you cut matters more than what you cut.

A NIOSH woodshop measurement reported by WoodworkingCove put a table saw, miter saw and router in the range of roughly 86–94 dBA, with the handheld router loudest at 92–94 dBA.

For other common stations, the honest answer is that I couldn't verify primary figures. A safety-supply vendor publishes a table listing thickness planers at 100–108 dB, router tables and CNC routers at 95–105, jointers at 85–95, band saws at 80–90, dust collection at 82–95 and shop vacs at 85–100. But the page names no survey, no meter and no standard, and its planer figure sits above HSE's planing class with no way to reconcile the two. I'll take the HSE and NIOSH ranges as the figures you can stand behind. Treat the vendor table as directional only, and don't build a purchase decision on a number with no method behind it.

The NRR Math Has Three Answers

Here's where most shop articles go wrong. They quote the "subtract seven" rule and stop. But OSHA's mandatory Appendix B to 1910.95 offers four methods, and they give different answers.

Method (iii), a sound level meter on A-weighting, is the one woodworkers quote: subtract 7 dB from the NRR, then subtract the remainder from the A-weighted TWA. Methods (i) and (iv), using C-weighted measurements, subtract the full NRR with no correction. The 7 dB is a real correction for the A-to-C weighting difference, not a fudge factor.

Then there's the second derating, from 3M itself, which recommends reducing the NRR on its E-A-R Classic plugs by 50% to better estimate typical workplace protection. And a third formula circulates on supplier blogs, (NRR − 7) / 2, which stacks both corrections.

Take a 33 NRR plug and watch the number move. Under OSHA method (iii): 26 dB. Under the 50% rule: 16.5 dB. Under the combined formula: 13 dB. That spread is the story. The rating on the package is a laboratory number under ideal conditions, and Fine Woodworking's Heim cites a government audiologist who joked that NRR means "not really relevant." He also notes each manufacturer tests under different conditions, so the NRR can't be used to compare brands.

Plugs, Muffs, or Both?

Dual protection is not additive. The classic error treats a 33 NRR plug plus a 30 NRR muff as 63 dB. NIOSH's practical guidance, per WC Safety's dual-protection reference citing the agency and the OSHA regulation, is to take the higher field-adjusted NRR and add about 5 dB for the second device. WC Safety suggests single protection usually suffices below ~95 dBA, dual becomes worth considering around 100 dBA, and above ~105 dBA a TWA reading generally points to dual.

Fine Woodworking's Heim, writing in 2007 and citing research, concluded the opposite of the modern safety-supply advice: plugs with muffs are "only marginally superior" to either alone. That contradiction deserves a flag rather than a silent pick. I'll say this much from my own bench: I wear plugs alone for most cutting and add muffs for long planer sessions or when I'm running the router table for a whole afternoon. When I do double up, the plug goes in first. A poorly rolled shallow foam plug loses 10–15 dB, says WC Safety, and the muff on top can't recover it.

On noise-canceling headphones, Heim's verdict holds up: ANC works best on steady constant sound and is less effective on the short bursts from shop machines. It's the wrong tool for a planer.

What It Actually Costs

Only one current price survived verification, and it's a plug: 3M E-A-R Classic 310-1001, uncorded foam, NRR 29, 200 pairs per box, at $70.49 from Staples. That's about $0.35 per pair. The manufacturer's copy notes the plugs are moisture-resistant, flame-resistant and latex-free, with slow-recovery low-pressure foam.

That's it. I could not verify a current single-unit price for any earmuff. The 3M PELTOR Optime 105 H10A is rated NRR 30 and meets CSA Class AL, per R.S. Hughes and Zoro; the X2A is NRR 24 and the WorkTunes Connect is NRR 24, both from listing information, but no current US single-unit price came back verified. I won't guess. If you're pricing muffs, get the number from the seller.

Heim's 2007 product test listed prices from $6 to $185, but those are 19 years old and shouldn't be read as current. His broader point still applies: "Any hearing protector on the market will cut sound by 10 db. or more... They're more than adequate for muffling the noise from machines in a home woodshop." And the best protector, he wrote, "is the one you'll actually wear." His 2007 reader survey found more than 6 in 10 respondents used earmuffs, and 1 in 10 said they usually wore no protection at all.

The Habit Beats the Rating

At 92 to 94 dBA, the question is not whether to wear protection, WoodworkingCove puts it plainly. By the time a noise feels uncomfortable, you're well past the level where damage happens. And cordless isn't quieter in any meaningful sense, as the same source notes, because the noise comes from cutting, not the motor.

So here's what I'd do with all this. Buy a box of foam plugs and keep them everywhere a machine lives. Wear them by default, not when you remember. If you want a better-set-up shop, look at tooling geometry before you spend on fancier muffs: HSE testing on a moulding machine found helical cutters cut noise 5.4 dB(A) free-running and 5.1 dB(A) with the hood closed, and an outfeed tunnel cut cutting noise by 3.6 dB(A). That's real, and it lasts.

The muff isn't the thing that saves your hearing. The habit is. Pick the protector you'll actually put on, then put it on before you flip the switch.