HR accuracy

[rbittman]

We all know the heart rate accuracy on the Fitbit devices is poor. While at rest, it’s accurate, but during exercise, especially where there are quick jumps in HR, accuracy is terrible. We were all hoping the new Sense 2 would have an improved HR sensor and/or algorithm. It doesn’t seem any different from the Sense, or the Versa 3. 

The newly released Google Pixel watch brags that it has the best HR sensing of any Fitbit/Google device. This claim remains to be tested. But for the moment, let’s assume it’s great. 

My question for you all who seem pretty knowledgeable about tech is this: is there a trade off between HR accuracy and battery life? The Apple Watches have great HR accuracy too. But both the Apple watch(es) and this new Pixel watch have woefully short battery life. For my personal use, their battery is absurd. I need at least 3 days of battery when out hiking. I’m thinking that the quicker sampling rate just naturally eats up battery energy.

Is that the reason?

[t.parker · Accepted solution]

@rbittman there is a trade-off. It's not only sample rate (I think I have measured once the sample rate of Fitbit devices and compared with PolarOH1 which is very accurate despite being optical). I wrote about it here . In short it's Sense 24Hz vs PolarOH1 120Hz. I don't remember Garmin's sampling rate but it was also higher than Sense. It doesn't mean that Sense/Sense 2 would be inaccurate because of that but it's more likely to skip some beats (although it could detect 1440bpm and I bet there is no human being with heart that fast). Turning off HR sensor indeed saves battery. Turning off HR on my Garmin Fenix 7 predicts +4 days. When I turned off HR sensor on my Suunto Spartan the battery life doubled that could last nearly 2 weeks on single charge (I was using chest strap for activities and no daily HR at all as it was horrible anyway). So there is a connection between HR and battery life. Sampling rate is one thing, another is brightness of lights. The Sensor has work harder to detect HR on certain types of skin and (depends on implementation and hardware) when it's struggling it will increase brightness which will consume more power. This will increase accuracy, too and may easier get through darker skin or tattoos (and on some watches people with tattoos or darker skin do notice higher battery drain and Garmin also confirmed that indeed that's the case). So it really depends how the hardware and software works. Fitbit tries to squeeze as much as it can from battery life so it cuts all the corners sacrificing smoothness of animation, framerate, HR sampling frequency etc. This indeed increases the battery life. Galaxy Watch 5 has a setting to limit HR readings. At rest, it may take a reading every 10 minutes, use normal sampling for activities and sleep and that also is supposed to reduce power consumption.

[rbittman]

@t.parker Thank you. Your explanation of answers a lot for me. We all have different expectations for these devices and long battery life is one of my requirements. It’s good to understand the trade offs involved!