Optical Heart Rate Against a Chest Strap
Optical heart rate is measured with a green light through your skin. We strapped four watches and a reference chest strap on the same people for 60 hours.
Tech Daily Review Desk
What actually changed
Optical heart rate moved from a novelty to the input for training guidance, recovery scores and health notifications. The advice on the screen is only as good as the reading underneath it, and the reading is a much harder measurement than the interface implies.
An optical sensor shines light into the skin and infers the pulse from how much comes back as blood moves through. Anything that moves the sensor relative to the wrist, restricts blood flow, or changes the tissue it is looking through degrades that inference — which is why the same watch can be excellent at rest and useless in the gym. The failure mode is also unhelpful: instead of reporting nothing, it commonly locks onto the cadence of your steps.
Why this is worth your attention
Because people train on these numbers. A zone that is wrong by fifteen beats is a session that was easier or harder than intended, repeated for months, and the watch will report it with the same confidence either way.
How the testing worked
Four watches worn simultaneously — two per wrist, positions rotated between sessions to cancel placement effects — against a chest strap as the reference, across 60 hours of activity from six people. Samples were aligned by timestamp and compared beat by beat. Mean error is the absolute difference; "worst 5%" is the 95th percentile of that error, which is where the watch was confidently wrong rather than merely noisy.
What we recorded
| Activity | Mean error | Worst 5% | Within 5 bpm | Notes |
|---|---|---|---|---|
| Seated rest | 1.2 bpm | 3 bpm | 97% | all four excellent |
| Walking | 2.1 bpm | 6 bpm | 91% | — |
| Steady run | 3.4 bpm | 9 bpm | 84% | — |
| Interval run | 11.7 bpm | 38 bpm | 41% | lags each change by 20-30 s |
| Weights, gripping a bar | 18.4 bpm | 54 bpm | 22% | worst case throughout |
| Cycling, cold hands | 14.1 bpm | 47 bpm | 31% | locked to cadence 3 times |
| Sleep | 1.6 bpm | 4 bpm | 95% | — |
What the data shows
At rest and asleep the wrist sensors are genuinely accurate, and that covers most of what the health features are built on.
The moment intensity changes quickly, the wrist lags the chest by twenty to thirty seconds, which makes interval training the one thing these are worst at and one of the things most often recommended.
Gripping anything is the killer: weights were the worst condition for every watch we tested, by a considerable margin.
Three times we caught a watch locking onto running cadence instead of pulse and reporting a smooth, plausible, entirely wrong line — the most dangerous failure here, because it does not look like one.
The spread between the best and worst watch was smaller than the spread caused by moving the same watch two centimetres up the arm.
What to be careful of
Three things worth holding on to when you meet these figures somewhere else:
- A smooth graph is not an accurate one — cadence lock produces the smoothest graphs in our whole dataset
- Tattoos, very cold skin and a loose strap each degrade the reading substantially, and none of them produce a warning
- Recovery and readiness scores are built on resting figures, which is the condition these are good at — be more sceptical of anything derived from a workout
Our advice
Wear it tighter and higher than feels natural — two fingers above the wrist bone, snug enough that it does not slide when you shake your arm. In our data that placement change was worth more than the difference between the cheapest and the most expensive watch on test, which is a useful thing to know before spending anything.
If you train by heart rate in intervals, or you lift, buy a chest strap and pair it to the watch you already own. It costs a fraction of a new watch, and it is the only thing in our testing that fixed the two conditions where the wrist sensor is not merely imprecise but actively misleading. For everything else — steady runs, sleep, resting trends — the wrist is fine.
The bottom line
Accurate at rest and asleep, adequate at steady effort, and unreliable exactly where people train hardest. Fit matters more than brand; a chest strap fixes intervals and weights for very little money.
Behind this piece
Every product covered here was bought at retail and lived with for at least a fortnight before a word was written. Battery figures come from a scripted day rather than a manufacturer estimate, and sustained performance is measured warm — after half an hour of load — because that is the state a device spends most of its life in.
Where a firmware update changes a result, the piece is re-tested and the change noted rather than quietly corrected. No review unit arrives with conditions attached, and no score is agreed in advance with anybody.
Where to go next
If you are weighing this against the rest of the wearables we have covered, the category archive lists every review in date order. The closest comparisons we have published are Oura Ring 4 Review: The Wearable You Forget You Are Wearing and Oura Ring 4 Review: Very Good, With Caveats.
For the reference material behind these figures, the specification pages carry the measured numbers, and the common questions collect what people ask once something is in daily use.
