[Equipment] Set clock in AirSense 11?

[Perickson]

Importing data into OSCAR from both AirSense 11 via SD and Wellue O2 ring.

The Wellue SpO2 display time is correct. After setting the AirSense to the wrong time zone in order to correct for daylight saving time and importing the data - off by a couple of minutes! Urgh. :-(

Cannot find out for to set / reset the AirSense 11 clock? Any help?


I see there is a clock shift in OSCAR for 'drift' and will use it if necessary, but seems like the stop gap it is.

FYI, I was able to check the times in OSCAR as last night I did a cycle of movement of ring, let the air flow, and repeat with each stage for a count of 10 (~7sec.)  and indeed they indicated well in OSCAR but off from each other a couple minutes. That'll be good enough for a rough setting of clock drift if I need to resort to that, but would much rather having the data correctly time stamped at the source...

I guess I'm assume the ring is correct - display time was, but probably do need to record the actual time of the synchro actions.

Since the ring O2 takes seconds to 'start up' and I'm not exactly sure if/when the AirSense starts recording and perhaps it records in arrears but with the current time so any better higher resolution synchro events would be welcome. Ring records motion, O2, pulse and of those motion is the only one that I can 'change rapidly - so ... On the airsense - lots of metrics, but airflow is easy and can be changed rapidly. The ring seems to average over a few seconds, so not sure how to get say 1 second resolution - but <5 seems pretty easy and I guess that's close enough.

[ScottZZZ]

No way to set the time on the AS11, it only allows setting the time zone.  

You can sync the Wellue to a specific time source.  It automatically syncs to iOS device when downloading data (not sure about Android).  If downloading to PC, then the sync is a manual step but you can sync time to PC. 

I'm using the Wellue SleepU wrist band.  I used a similar method as you did -- some specific breathing patterns on AS11 and some large movements on the SleepU -- to determine how far apart they were.  In my case, it was 49 seconds.  I then used the "CPAP Clock Drift" setting in OSCAR to align the two.  The movement tracking was the key (does the ring have movement tracking as well?)

I sync the Wellue's time daily to my iPhone, which gets accurate time from its cellular connection.  The two have stayed "in sync" with my 49 second offset, which implies either that (1) The AS11 is also periodically syncing with something, perhaps the Resmed server when uploading data; or (2) The AS11 was factory-set and has a high-resolution clock chip that has very little drift.  My money is on #1, but nobody aside from Resmed knows right now.  Pretty disappointing that Resmed didn't incorporate accurate time sync on the AS11, but it is what it is.

In any case -- I've been using that 49 second offset for months and it still appears to do the job:  The "movement" data from the Wellue is still in sync with the AS11 data when (for example) getting up during the night for a bathroom trip.

[Perickson]

Thanks, I was syncing regularly to ios. Tried the manual and will test tonight.

Yes, the Wellue Ring I have has motion as well as O2 and pulse. So we should find out in the morning. I will also record accurate time of sync actions so if there is disagreement, I’ll know who is doing what. I also have other devices, but without motion so probably cannot sync in such an obvious manner.

I could adjust the computer (or other) time prior to sync, but suspect it’s not the ring that is off and so I’d rather not make to one that’s accurate wrong just to get two to agree. I understand alignment is the only part that is important for OSCAR charting etc. but I’d like to calibrate and so would like the times to agree the NIST as represented by local time.

Anyway, sounds like a plan - when I gotten it working, I’ll report back.

[Crimson Nape]

Since the Wellue can report a start time over a minute after the actual start, OSCAR v1.3.1 allows you to modify the Wellue data's filename, so you can match it to a closer sync. Since I start both the CPAP and SleepU at the same time, I change the SleepU's filename to match my reported CPAP start time. This also allows for major time inaccuracies without disturbing synchronization of the two devices.

[Perickson]

Looks like it's the AirSense 11 that timestamps late. So, I'll compensate in the OSCAR > preference > CPAP page...

[Perickson]

Test last night was enlightening:

The Wellue was approximately right on as to clock time. It did however have to issue.
    one was that it didn't seem to start recording for roughly a minute after the display was showing results - so it only captured the last part of my 'sync' movement - but enough
    the view in OSCAR only plots jointed points every four seconds, so it's going to be tricky to get ~ 1 sec. sync. I have *not* yet checked the binary data etc. but...

The AirSense 11 as reported by OSCAR however was off by 2 min 38 sec so a setting in OSCAR > preferences > CPAP > CPAP time offset of 0 -2 -38 seems to align them well.

I may adjust by a second or two if I get better infor from the Wellue, but probably not as I'd really like the data to be clock time aligned and the flow rate indication has really good resolution and the offset adjusts in to the nearest second - which seems pretty good for these purposes.

If I have to 'live with' the 4 sec resolution on the Wellue, than I think a ritual of a 6 sec burst of movement, followed by 5 idle, and then 6 more of movement will give my an 'edge' every second of the 4 second 'cycle' and I should be able to tell approximately where the edge is being recorded. But first a view of the CSV and binary is in order...

[Crimson Nape]

Wellue samples every second , but only records every 4 seconds. The recorded value could be an average of the last 4 samples. To date, Wellue will not provide an answer.

[PjotterHasApn]

Just did a test and the 'Checkme O2' from Wellue takes around 11 seconds to record in my test. When you start it at 11:00:00, it has a timestamp of 11:00:11. In the same test, the Resmed Airsense 11 took around 8 seconds to begin recording. The Resmed is currently more than 11 minutes off compared to real time (and shifting continuously).

I don't prefer the method of modifying the filename of the O2ring. Even then, it's hard to synchronize them exactly since they don't start recording immediately and have different delays. Plus, it makes it harder to also synchronize other things, like sleep stages, camera images, etc. You would just have to make everything wrong that way.

I plan to create something to import Apple Watch sleep phases into Oscar, so that wouldn't work then either.

I do the following to synchronize: I start the CheckMe O2 ring at the same time as the Resmed 11, and keep my arm with the ring still, breathing normally. After approximately 30 seconds, I'll breathe out powerfully and move my arm up in the air. I repeat it one more time after about 30 seconds.
I then import the Resmed data and the ring data into Oscar and line up the movement spikes and the breath spikes. Works like a charm.

One might say: what's the big deal if it's a few seconds off? Well, that depends on what one is trying to figure out. For instance, I see CAs, and also movement. I suspected the movement caused the CAs, but it could have been the other way around. 

It turned out my breathing is irregular when I turn during light sleep stage, and right after the movement there always came a short CA, but again, only when in light sleep stage. If I am in awake sleep stage, I often get a OA instead, because it turns out that I then hold my breath during the turning. Interesting stuff.

I could only notice these things because it was very well synchronized. It's a pretty significant discovery because these are pretty innocent events that an APAP doesn't really need to do anything about. 

It's really ridiculous that a modern device like the Resmed doesn't have the actual time, and we have to think of stupid tricks to try to synchronize it.

[Perickson]

When you figure out how to get apple sleep phases into Oscar as well. Please share. I’m thinking about it but dont Have time nor understand Oscar well enough.

[PjotterHasApn]

Extra info: The CheckMe O2 ring has an sampling interval of 1 second and a display (export) interval of 2 seconds. (while the 'normal' O2 ring has an export interval of 4 seconds).

@Perickson:
Sure, I'll share it.