Rules
Part of Quiet mechanical keyboards: where the sound comes from
Quiet mechanical keyboards privacy and sound
Quiet mechanical keyboards privacy and sound call for separate checks: compare force curves, test damping, and treat software separately.
Quiet mechanical keyboards privacy and sound need separate checks. These notes cover force, damping, repeatable sound tests, and the device-policy questions raised by companion software.
What to take away
- Deskthority's switch force curves explain feel. They do not predict how loud a board will be in your case.
- RTINGS sound tests use a meter at a fixed distance. That beats any marketing claim about decibels.
- KBDfans and Geekhack damping logs show that materials change frequency content, not just volume.
- NIST SP 800-46 Revision 2 covers device rules for a keyboard running companion software.
- The open-plan noise survey of 1,078 workers explains why a quiet board earns goodwill.
1. Deskthority switch force curve archive
Deskthority's wiki collects force curves for hundreds of switches, plotted as grams against travel. A linear switch and a tactile switch with the same bottom-out weight feel different because the curve shape differs, not the number.
Switch force curves compared
Linear switch
- Bottom-out weight
- same
- Curve shape
- smooth
- Feel difference
- curve, not number
- Sound link
- none
Tactile switch
- Bottom-out weight
- same
- Curve shape
- bump
- Feel difference
- curve, not number
- Sound link
- none
Read it before you buy if feel matters to you, and read it with one caveat: force curves say nothing about sound. A stiff switch can be quiet, and a light one can be loud. It suits anyone choosing between switches on weight alone.
2. RTINGS keyboard sound tests
RTINGS publishes measured sound output for keyboards, recorded at a fixed distance with a sound meter and reported alongside frequency data. It suits a buyer who wants a repeatable number instead of an impression from a review.
Repeatable home sound test
- Same sound meter
- Same fixed distance
- Same desk surface
- Repeat after one week
- Compare whether fix held
Copy that method at home: same meter, same distance, same desk surface, repeated after a week. A single reading tells you little. A reading you repeat after the mat compresses and the foam settles tells you whether the fix held.
3. KBDfans and Geekhack damping build logs
Build logs on KBDfans, Geekhack and similar forums document foam, felt, cork and rubber sheet placed inside a case, with before-and-after impressions. Treat these as leads, not data. They suit a builder who wants a material list before ordering parts.
Nobody in a forum thread is holding a calibrated meter, and case resonance depends on the specific board. The value is in the material list and the order of operations, which you can then test yourself.
4. NIST SP 800-46 Revision 2 on telework and BYOD
NIST's guide to telework, remote access and BYOD security covers personally owned devices used for work. A keyboard with companion software is one of those devices.
The publication is written for organizations, and it gives a home worker the right question: who decided which rules apply to this device? If nobody did, that is the gap. It suits a remote worker whose board runs a utility on a personal laptop.
5. The open-plan office noise survey
A survey of 1,078 office workers compared shared rooms of two to five people with open-plan floors of more than five. Irrelevant speech was the complaint under study. It suits anyone making the case to an office manager for a quieter board.
More than 70% of open-plan respondents said they would lower their voices if a noise monitoring system signaled them with a light. That cuts both ways: people accept the feedback, and the sensor is still in the room.
A worked example: testing one board
Pick a board, a meter and a desk, then hold everything else still.
- Record the meter reading at 30 cm from the spacebar, three presses, and write down the average.
- Add one damping layer, whether foam, felt or cork, and repeat the same three presses at the same distance.
- Wait a week, then repeat both readings to see whether the mat compressed or the foam settled.
- Compare the two averages. If the gap is inside the meter's own margin, the mod did not do what you hoped.
This is the method behind what actually makes a keyboard quiet, and it is the only way to know your own board rather than someone else's.
What the readings do not settle
None of these five sources tells you whether a specific board is quiet in your room. Case resonance, desk mat thickness and keycap material all shift the result, and a spec sheet rarely reports any of them. Buy from a seller with a return window, test on your own desk, and keep the receipt.
Software is the other half. A board that stores its layers in onboard memory and types as a plain keyboard without a utility is easier to reason about than any privacy policy. The buyer checklist covers that side, and how local installers work with quiet mechanical keyboards covers the build side.
Common questions
Do I need a sound meter to compare boards?
No, but you need consistency. A phone app at a fixed 30 cm, same desk, same three presses, repeated after a week, beats a one-off reading from a review. The number matters less than whether it moves.
Are force curves useful for judging noise?
Only indirectly. They describe feel and actuation weight. Sound depends on the case, the mounting, the keycaps and the damping, so a quiet switch in a hollow case can still be loud.
Which damping material should I try first?
Start with the cheapest reversible option, usually a felt or foam sheet under the PCB, because you can remove it. Cork and rubber change the sound differently and are harder to undo. Test one layer at a time so you know which change did the work.
Is a quieter board a privacy control?
No. It changes what a microphone picks up, not who has one open. Treat it as courtesy to the room and handle software and device policy separately.
