Updated: Comparison: Phone alarm vs sunrise alarm | Result: Positive

[cb0806151]

Update: the slider feature allowed a better examination of this and it's going to move the result to positive.

The top two are with just the sunrise alarm (no sound, just light) whereas the bottom two show the phone alarm going off at 7AM and causing a massive heart rate spike as a result. I only recently disabled the phone alarm so it's going to take a bit more data to see the full effects of waking up with just a light alarm, but its clear that it results in a more calm start to the day.

One of the things I really enjoyed about the whoop reddit was people sharing their before/afters of different things they did to optimize sleep and such so I'm starting that here.

Recently got a Hatch sunrise alarm:
https://www.hatch.co/hatch-sleep-clock?utm_source=google_shopping&utm_medium=paid_search&utm_campaign=20987488079&utm_adgroup=199754585198&utm_content=819023140933&utm_term=&utm_device=m&gad_source=1&gad_campaignid=20987488079&gbraid=0AAAAADh7rJN75RnmbBSrxmUio6VBbF99s
and was able to run a comparison with just a plain old phone alarm to see how it compares.

Phone alarm on the left and sunrise alarm on the right:

While it's clearly making waking up easier it doesn't seem to change the bpm jolt that an alarm does. Based on this data my bpm is still spiking ~40 bpm upon waking. The bpm spike might also be tied to movement so I might just need to lay in bed for ten minutes and see how the two compare. Based on all that, I'm giving a result of: mixed.

[abdulsaheel]

insightful man, honestly

[mikenrafter]

I'd be curious to see how your sleep phase aligns with that. I find that tying my wake to a dynamic window of ~45m with some application that estimates my sleep phase drastically improves the experience of waking up. I am so much less stressed.

I've tried a bunch of the aforementioned "natural wake" solutions.

• I first used SleepOnDroid, but they charge you an arm and a leg in subscription fees, so I first tried abusing their free trial before deciding that was too much effort to continually technologically circumvent. I never did fork over that cash. Can't say how well the pro version with its numerous features work 🤷...
• I also tried using a PineTime watch with WaspOS and a little SleepTk app that someone made. It worked OK, but given the hardware constraints, it never lasted long enough to do anything but estimate and die halfway through the day. Even micropython is expensive.
• I have used an airgapped version of an app some random startup made. It was rugpulled not long after. I have the old APK though, and I just blacklisted it from my updates. It doesn't estimate my sleep phases very accurately though. So, that's hit or miss.
• My personal (for now) fork of edge is currently the best at this so far. Hoping to get it PR ready sometime here shortly. I'd be curious to know if it makes any difference in your setup.

[cb0806151]

That feature sounds awesome! Can't wait to see the PR, I'll definitely be testing it out asap once it releases.

I remember people clamoring for a feature like that on the whoop reddit for ages (and I don't believe they ever got it but I've been off the reddit for over a year)

[cb0806151]

Would you say it's like 25% drop in stress or more?

[mikenrafter]

Hard to quantify, but between 20% and 40% when it works right. Though, I suppose I could quantify it with HR spike differences... Though there are many variables to control for there.

[mikenrafter]

Unfortunately, the phenomenon seems to be understudied in open-access scientific literature. 'Sleep Inertia' does have some research behind it, but that isn't a stress response.

[cb0806151]

That's a substantial stress drop in either side of the range! Really looking forward to testing this out.

Mmm, that is unfortunate; I wonder if edge could eventually have a 'participate in study' feature where someone could follow the study criteria and then submit their data within that time range. Itd be difficult to collect a lot of people, but (I'm going out on a limb here) a study of 10 over two weeks testing out the feature should be enough to create at least a small study that might trigger larger followup studies down the road.