a quick overview of my experiences. See the below discussion for an original overview of my journey. I have moderate supine based sleep apnea. I have been undertaking sleep positional therapy (essentially using a pillow to stop me sleeping on my back, where I snore most) since Apr/May 2024. I’m also currently using NightShift to help track things for a 3 week trial period.
As suggested below, I purchased this watch.
I am concerned though, with the findings of this watch below. Must be noted though that it didn’t test the blood oxygen level as well., which is the main reason for purpose along with the sleep tracking component.
Would love to know why Petr loves this one so much, other than that it allows full open source sensor access. Poor performance though would somewhat negate any value in sensor access IMHO. Thoughts, Petr?
Is it a case that combined with SaA data/tracking it somewhat improves the reliability of the data? I’m not convinced, as it’s my understanding in my chats with Lenka that the app does’nt track a lot of data, as it relies on what it pulls from the watch. Do I have that correct, Lenka?
I can say, have had a lot of help from Lenka over the past few months understanding the data/graphs for all this. Thank you, Lenka.
I must say, I find the SPO2 very unreliable during the day, but seems very consistent in showing readings during the night. Obviously less activity during the night helps with that. I recently had a medical grade device for 3 nights, whilst also using this watch and SaA, and hope to get the report next week, so I can compare what it shows to this setup of watch/SaA. I’ll update this post once I do.
Thank you @mgs1968 for your contribution. I look forward to your future updates. SpO2 tracking was one reason, among several others that led me to buy this watch about a month ago. I have somewhat of a snoring issue. Have you also tried the snoring alert sounds? If so, have they helped? Both @lenka-urbandroid and @petr-urbandroid have been helpful over the past number of years. They’ve been the most responsive team I’ve encountered of any app I’ve ever used.
Hello @mgs1968 … I love the Pebble watches as they were perfect combination of features, great display readability on direct sun and battery life. Over the years nothing did come close only the TW3 70% and then TW5 90%.
A very good test of the SPO2 sensor is Wim Hof’s breathing technique I have tested the sensor and it seems the react as expected. Of course the SPO2 sensor on a wrist heavily depends on the tightness. So when my band is loose my SPO2 drops to 70 many times during the night which is a false positive. But when I tight the band completely at endge of comfortability I get 95-100 readings from my night.
And exactly as you say, to track SPO2 realibility the arm has to be still. You cannot track SPO2 with this type of sensor if you hand is moving!
Of course as an engineer I appreciated that Mobio opened the sensors for 3rd parties and it is a big pitty nearly no nody else is doing this - Pixel Watch, GW etc…
Remember he is using the algorithm that the watch itself uses to determine sleep phases, not what Sleep as Android finds using the data from the same sensors. The results vary a lot on my TicWatch Pro 3 between the native TicSleep app and what Sleep as Android finds. I wish I knew which was more accurate. Even the Apple Watch SE, which is now as cheap as the TicWatch Pro 5, he finds to be as accurate as his wearable EEG. This makes me want to get one, however there is no app as good as this one on iOS. Hopefully Sleep as Android comes to more accurate conclusions with the TicWatch sensor data. I am inclined to get an Apple Watch and compare.
Hi @urbanhym, we explain our approach here and here, where we also explain our method to translate the stages of N3, N2, N1 and REM into sleep phases (deep, light , REM). Each app can take different approach, so the results may vary. It does not mean one is better than other. We recommend comparing your progress in long term in every app rather than comparing one night in all apps.
Is there a source explaining how the wearable measures the brain activity for EEG data? Normally, this is measured by electrodes placed on the scalp.
Hi, we compared our data to the data from Sleep labs (medical facilities). So we hope our source of data was reliable enough.
The standard hypnogram does not use deep, light, REM. It uses Non-REM (N3, N2, N1), and REM. The method to translate N3 and N2 into deep sleep vs light sleep may differ.
The approach we use to place a threshold for deep sleep and light sleep is probably different from other apps, we use this method consistently for 13+ years. Maybe the other apps are also using different approach than the EEG headband he was using?
It would be interesting to see more deep review, 3 days is quite a short test.
Well we can’t be flippant about accuracy if we’re trying to wake in a light phase, which is the main reason I use the app. Also to increase REM specifically, which also requires accuracy of phase recognition. Using TicWatch 3 Pro the measurements of the internal program TicSleep were much different last night than Sleep as Android. I trust your app more because I’m sure more development is going in to it. But with these sensors, and your app data, I’d like to compare it to Apple Watch. Was thinking of buying one, and now my TicWatch screen is failing after only 9 months, so it would’ve been the perfect time, however it is still under warranty, so I may get a TicWatch 5 as a replacement.
I tried Apple Watch just for a few days and it was not bad, the REM estimation was ± matching (it was overestimating my awakes, though).
If you will have any feedback on the comparison, you can share it here
I have tested the sensor and it seems the react as expected. Of course the SPO2 sensor on a wrist heavily depends on the tightness. So when my band is loose my SPO2 drops to 70 many times during the night which is a false positive. But when I tight the band completely at endge of comfortability I get 95-100 readings from my night.