Vivoactive Respiration Rate - So inaccurate it cant be used

It's not about adapting long-term, it's that it can't actually see your breathing, just the side effects of it on your heart rate, and just like it takes (at least) a few steady beats for most heart rate sensors to pick up on your heart rate, it takes the respiration 'sensor' algorithm (at least) a few steady breaths to pick up on your breathing rate.

Thanks for the reply.  Understand now that the algorithm takes a few steady breaths to pick up your breathing rate.  I just doubt that the algorithm actually works?  I’ve returned the Vivoactive 4 and are now trying out a Venu, in case it was just a problem with the watch.  The Venu reads the same.  Today I did a 2 hr breathing session and maintained <3 bpm throughout using the  Garmin Yoga app to record.  The average Respiration Rate read 16 bpm?  I would have thought that 2 hrs was long enough to pick up a few steady heart beats ?

If I test by doing a 12 min session with the Garmin app coherence, which has already predetermined timed  Breath rates of 3-4 bpm, the Respiration Rate reads the same at 3-4 bpm average.  If I then continue at the same breathing rate and then use the yoga app for 12 mins, the  respiration rate in the app reads 16 bpm when I am only breathing at 3 bpm ???  I can’t see that the algorithm works at all.  I’d appreciate if others would test with the watches Garmin apps coherence and then yoga and let me know if they get something different (I’d be very surprised ).

I don’t think the Respiration Rate feature works.

Oof, the Vivoactive 4/4S and the Venu are mostly identical except for the display size/type -- I'd expect the same sensor behavior in both.

Honestly the RR algorithm may just be ineffective. I get the sense that Garmin's raw sensor tech is way behind Fitbit and they somewhat improve things using context about what activity you're doing. I stopped using my VA4S a while back because all the neat health stats depended on solid heart rate/SpO2 data and it just wasn't getting that.

The same thing happens to me.  Using the built in breath work activity will show 3-4 brpm but outside of the breathwork activities it never detects breaths lower than about 9-10 per minute.  I've tried it for longer durations of 30+ minutes at about 4 breaths per minute and lower but it reads 12-13... Total bummer.  Makes me doubt the accuracy of the other features as well

Yeah, just a Gimmick.  A $600 step counter!

Yes, I've had the same problem when I do yoga. It seems to count breaths fine during the day and when I'm asleep: when breathing is above 12 bpm most of the time (I go below a little during sleep, but not much, and not for long), but yoga, where I breathe roughly 1 breath 10-12s (5-6 bpm) it counts as about 3x that. Stress is usually high too. (Mind you, this week it told me I was stressed and needed a break-- when it was charging. I wasn't even wearing it!) It's got better at stress over time. Perhaps, after four months, it is learning something. (I suspect the breath thing is an algorithm which overrides observed breathing as implausibly low, and it therefore multiplies what it measures by any integer to get bpm into the 12-20 range.)

Yeh I've got an issue currently logged with Garmin for same thing. I can be sitting down and physically timing 4 - 5 breathes per minute resting and the Fenix 6 is counting 14. An interesting experiment is to do their box breathing exercise (4 sec In, 4 sec hold, 4 sec Out, 4 sec hold) while in normal mode and ti will count that at around 14 -15 bpm for me even though it is 4 bpm. Then, while maintaining the 4 bpm box breathing, enter the breathwork activity and it will instantly drop the reading down to 4 - 6bpm, then, when the breathing activity has timed out, it will instantly start timing exactly the same breathing rate at 14 or so.

So if it can measure reasonably accurate during the breathing activity but be like 350% out during normal operation, then they have some issues with their algorithm. That they are using two different algorithms say's there's a problem anyway. It's like they are locked in on the 12 - 20 range so the algorithm is looking for interpretations that give it around that range. And then when the breathing activity is started, the algorithm is looking for 4 - 5bpm because that's what the activity is guiding you too.

If you're happy to pant away every 3 - 6 seconds when you're at rest, then this is probably a good tool for measuring that. If you've gone to the trouble of getting your abdominal muscles active in the breathing process, and perhaps even using nose rather than mouth, then the current software in the watch doesn't seem able to interpret the signals anywhere near close to accurate.

Disappointing, because like Journeyman, the respiration feature they advertised was a key reason for selecting this watch. Will see what word comes back as they check my data files and comments. Seems fixable if they want to.

It is the same issue with heart rate. Depending on activity, the watch reports different values with the same sensor input, because different algorithms are used. This should not come as a surprise. To convert a sensor signal into a real world reading, some probability based calculations needs to be done. If the watch thinks it knows what you are doing, that knowledge will be taken into account. I guess all watches are doing that. Over the next years, both algorithms and sensor technology will improve, but as of today, we must live with these limitations.

Haven't noticed any issues with heart rate measurement with the watch. It's also pretty much the same reading as the Fitbit it replaced.