Reviews
Part of Which silent switch type makes sense for which typist, in plain terms
How to Read a Keyboard Switch Spec Sheet: What the Numbers Do and Do Not Tell You
Learn which switch spec sheet numbers, like actuation force and travel, predict typing feel, which to ignore, and why noise is not listed on a sheet.
Switch listings publish more numbers than almost any other component at a desk, and most of them are measured under conditions nobody states. Here is which ones carry information.
What to take away
- Actuation force and total travel are the two numbers that predict feel.
- Rated keystroke life is a manufacturer figure with no common test behind it.
- Nothing on a switch specification describes noise, because there is no standard test for it.
- Your own proportions change the answer more than a ten gram difference does.
The numbers that mean something
- Actuation force, in grams force: how hard you press before the key registers. This is the number to match to your hands.
- Total travel, in millimetershow far the key goes. Longer travel generally means a gentler stop.
- Pretravel or actuation pointhow far down the key registers. Shorter means faster and more accidental presses.
- Bottom-out forcewhat you actually press, if you are a hard typist, and often absent from listings.
Bottom-out force is the omission worth noticing. Most people bottom out most keys, so the force they experience is not the actuation figure printed on the box.
Switch spec numbers: useful vs less useful
Useful
- Actuation force
- Match to hands
- Total travel
- Longer = gentler stop
- Pretravel
- Shorter = faster, accidental
- Bottom-out force
- What hard typists feel
- Rated keystrokes
- —
- Noise figure
- —
Less useful
- Actuation force
- —
- Total travel
- —
- Pretravel
- —
- Bottom-out force
- Often absent
- Rated keystrokes
- No public test method
- Noise figure
- No comparable standard
The numbers that mean less
- Rated keystrokes, usually quoted in tens of millions. No common public test method sits behind it, and switches rarely fail by wearing out. They fail by contamination.
- Any noise figure. There is no comparable standard, so two numbers from two makers are not comparable.
- Tolerances quoted without a sample size or a method.
Why the sheet cannot describe your experience
The specification describes the switch. It does not describe you at the desk. One laboratory study built prediction models from twenty-five people performing computer trials across a range of keyboard and mouse mixes. It found that the proportion of keyboard and mouse use explained much of the variability in wrist velocities and accelerations.
Individual factors including hand length, shoulder width and age were each significant predictors of at least one measured exposure. Two people with the same board have different exposures, and no specification sheet contains that.
The laptop exception
If the machine is a laptop, most of this is beside the point. Cornell's ergonomics group gives the reason in its tips for laptop use. The design puts the keyboard and the screen in one unit, which breaks a basic requirement that the two be separable.
If the keyboard is right the screen is wrong, and the reverse. Laptops, it notes, are excluded from current ergonomic design requirements for that reason. Comparing internal laptop switches is a smaller decision than deciding whether to add an external keyboard at all.
How to use a sheet in practice
- Write down the actuation force of the board you type on now, if you can find it.
- Move by no more than about ten grams force in either direction on the first change.
- Prefer longer travel unless you have a specific reason not to.
- Ignore rated keystrokes entirely.
- Judge noise by listening, never by a number.
For the mechanical families those numbers describe, see options compared, and for a plan that survives a change of machine, see planning a switch choice you will not redo.
