OSCAR is Released and Live

[Tubaman]

Your suggestion worked!  I deleted the last 2 days and then re-imported them into (OSCAR).  That solved it.  OSCAR shows my nap on today's graphs and SH shows my nap on yesterday's graphs.  I don't care, either way is OK -- as long as everything is displayed.

Your question: "Was last night's sleep data imported in two different import sessions?"
I import the data once or twice each day whenever I have any new sleep data.  Yesterday's I imported twice: when I got up in the morning and after my 4 hour afternoon 'nap'.  SH had no problem dealing with it.  OSCAR was OK yesterday but when I got up this morning and imported the last night's (this morning's) data, OSCAR ignored last night's == this morning's data and only showed yesterday's nap.  SH correctly updated its graphs showing the nap on yesterday's graphs.  I've never had this problem (entire sleep session's data not displayed) with SH and I've had many strange sleep times (naps) since I started using SH. 

Your question: "I notice that SH and OSCAR appear to be reporting as if in different time zones.  Is one of them correct?"
Since SH allowed me to set the daily divider to 1500, the afternoon 'nap' was correctly shown on yesterday's graphs.  They both show today's (5/21/2019) sleep graphs as 5/20/2019 even though my entire (today's) sleeping session happened on 5/21/2019.  That's the way the graphs have always been displayed in SH and now OSCAR.  I don't know if this answers your question -- please let me know?

I noticed OSCAR setup gives a reason for not allowing the daily divider to be switched -- apparently this does not work for Resmed machines in OSCAR.  My SH daily divider time adjustment to 1500 has been working OK for my Resmed machine for a long time (at least that's how it appears to me).  Occasionally the SH 'Overview' display page has shown the pressure chart blank but I've solved that by deleting and re-importing that day's data (the pressure chart is then correctly displayed).  I haven't experienced this yet with OSCAR but then I've only been using OSCAR since yesterday (5/20/2019).

Thank you for helping me!  I've used the delete page -> re-import page function many times in the past for the pressure display problem but I did not think of it this time ('senior moment').
Carl / Tubaman

[GuyScharf]

Your suggestion worked!  I deleted the last 2 days and then re-imported them into (OSCAR).  That solved it.  OSCAR shows my nap on today's graphs and SH shows my nap on yesterday's graphs.  I don't care, either way is OK -- as long as everything is displayed.

Good.  Glad it worked.

Your question: "Was last night's sleep data imported in two different import sessions?"
I import the data once or twice each day whenever I have any new sleep data.  Yesterday's I imported twice: when I got up in the morning and after my 4 hour afternoon 'nap'.  SH had no problem dealing with it.  OSCAR was OK yesterday but when I got up this morning and imported the last night's (this morning's) data, OSCAR ignored last night's == this morning's data and only showed yesterday's nap.  SH correctly updated its graphs showing the nap on yesterday's graphs.  I've never had this problem (entire sleep session's data not displayed) with SH and I've had many strange sleep times (naps) since I started using SH.

I've seen problems in SH when importing a day in two "pieces" and one of the sessions being at least somewhat ignored.  Summary data was not correct if that happened.  If you do that again, keep an eye on the summary data (left column on Daily page) to see if anything odd happens there.  I have not tested this scenario in OSCAR.

Your question: "I notice that SH and OSCAR appear to be reporting as if in different time zones.  Is one of them correct?"
Since SH allowed me to set the daily divider to 1500, the afternoon 'nap' was correctly shown on yesterday's graphs.  They both show today's (5/21/2019) sleep graphs as 5/20/2019 even though my entire (today's) sleeping session happened on 5/21/2019.  That's the way the graphs have always been displayed in SH and now OSCAR.  I don't know if this answers your question -- please let me know?

My question was whether the times shown on the graphs were correct to your local time.

I noticed OSCAR setup gives a reason for not allowing the daily divider to be switched -- apparently this does not work for Resmed machines in OSCAR.  My SH daily divider time adjustment to 1500 has been working OK for my Resmed machine for a long time (at least that's how it appears to me).  Occasionally the SH 'Overview' display page has shown the pressure chart blank but I've solved that by deleting and re-importing that day's data (the pressure chart is then correctly displayed).  I haven't experienced this yet with OSCAR but then I've only been using OSCAR since yesterday (5/20/2019).

I believe the issue is that if OSCAR were to split the day other than at the noon that is built in to the ResMed machines, that calculated AHI etc. would not agree with the summary AHI in the left column, since that data is from the machine which always splits the days at noon.  However, I have not investigated this myself so may have the details wrong.

[Tubaman]

Today I obtained a relatively 'normal' sleep session (started and stopped within a single day, 5/22/2019).  Two things appeared to be problems with OSCAR:

1) Insertion of oximitry data does not work correctly (nothing new there);
2) Daily summary data does not agree with SH, especially in 'Max' column (see attached, OSCAR on the left, SH on the right).

I'm unsure of the importance of the 'Max' column but the differences compared to SH are significant, especially in my respiration rate and tidal volume statistics (which I'm watching closely for obvious reasons).  Which one is correct (or closest to being correct)?

The oximetry data insertion in OSCAR, as bad as it appears in the chart, always shows up in the previous day's page.  In SH it shows up in the correct page but also appears very bad (different, however).  I've tried to set OSCAR to show up in the correct page by changing the date (OSCAR does not show the calendar during the insertion dialog, SH does) but that doesn't work.  Rather than trying to explain any further, I could show you successive screen-grabs of the comparisons between OSCAR & SH oximetry insertion dialogs, I will leave it at this:

With my Win7 computer, neither OSCAR nor SH is compatible with my CMS50I oximeter even though my oximeter model is listed in both OSCAR & SH as being compatible.

Sorry I'm such a complete annoyance to this forum.  Just tell me and I shall go away.
Carl / Tubaman

[jaswilliams]

The difference in figures of the max values is buy design SleepyHead 0.1 shows the absolute max values reported by your cpap device whereas OSCAR and Sleaepyhead 1.1 display 99% values which remove the random high values. Both are right which one you need to be watching is the question I cannot answer

[sheepless]

max, whether 100% or 99% isn't often as useful as med and 95% anyway.

[Tubaman]

Thank you!  It's difficult for me to understand why the max values would be so different (regardless of their importance to the 'sleeper') since they were both selected from the same data.  Seems to me they would be identical even if the 95% and Med values differed slightly due to being determined by calculation.  Oh well, I shall leave that alone.

My next questions were the large differences in oximetry results for my allegedly compatible CMS50I oximeter.  Below are some major differences encountered when I uploaded last night's oximetry data into OSCAR & SH.

ScreenGrab #1 (left):
SH selection screen.  Everything is OK on this screen;

ScreenGrab #2 (2nd from left):
OSCAR selection screen -- Duration values for #2 and #3 are significantly incorrect;

ScreenGrab #3 (3rd from left):
SH Import completion screen -- Everything is correct, including the oximetry start time;

ScreenGrab #4 (right):
OSCAR Import completion screen -- No calendar, no oximetry start time displayed.  Must be manually entered;

I don't think my oximeter, CMS50I, is compatible with neither SH nor OSCAR based on the above.  In both programs the final results (detail data displays) are disasters so I didn't bother showing them.  In addition OSCAR refuses to insert the display into the correct date (always the previous day).  I have tried to manually enter the date and oximetry start time into OSCAR's import completion screen with no luck.  The oximetry display will start at the correct time -- in the previous (wrong) day's detail data display.

Elsewhere in this forum is a statement that the CMS50I oximeter is compatible with OSCAR and SH.  Unless there is a problem with my Win7 x 64 bit computer (I don't think so), the CMS50I oximeter is not compatible with OSCAR and SH and the online documentation should be corrected to reflect this fact.  I have spent much time trying to insert my oximetry data into SH, and now OSCAR, which could have been avoided had the online documentation indicated my oximeter was incompatible.

FYI and thank you for all the help!
Carl / Tubaman

[pholynyk]

Tubaman - Thank you for those detailed comparisons. I have found some errors in updating the OSCAR oximeter import, as well as some in the SH 1.0. I hope I have corrected them for our next release, so I would ask you to keep on using both SH and OSCAR for comparison with the new release.
I'm not sure when we will make the new release; we still have some areas to work on.

[Tubaman]

Thank you for the 'Thank you'!  I figured everybody would be angry at me for my previous posting.  My own programming experience has shown me how difficult it is to get the software doing what you want it to do and not what you tell it to do.

Attached is a FYI -- I loaded last night's oximetry data into OSCAR (I had to manually enter the date and start time as I mentioned in my previous posting).  As usual, it loaded into the wrong (previous) day.  The resulting display needs no comments from me.  SH's oximetry display looks similar except it shows in the correct day.

I also attached the oximeter's graph for last night.  Please ignore the blocked out sections, that's an edit I do after uploading the data to eliminate my sandbox visit(s) and beginning and end periods when I am not sleeping.

Thanks again for working on resolving this problem!  If I can give you any more information, please let me know.  I run my oximeter almost every night.  Probably not necessary but once in a while it gets interesting.

Carl / Tubaman

[Tubaman]

Tidal Volume max reading (in this case my "Tidal Volume" max):

I really think OSCAR is not determining / indicating the correct Max value (SH is OK and indicates 1100 Max which is correct).  The attached expanded chart shows two successive values significantly above the 880 reading displayed in the statistics summary table, 1010.59 and ~1100.  Maybe one outlier missed is OK but I don't think two high values like this should be accepted (they are therefore not outliers).  I know the Max value isn't considered very important, however ....?

And then again:  Have you programmed OSCAR to ignore the Tidal Volume Max values occurring while the pressure is ramping up (as these are)?  In that case there is a 976+ value further on (I did not display this one) that is still higher than the 880 listed and that one did occur during operational pressures, well after ramp-up was completed.

For what it's worth...
Carl / Tubaman

[Gideon]

Carl
We have not as of yet changed any calculations in OSCAR. That said we are reviewing the code and will be looking at those display calculations. Many of them need to be fixed.I'm thinking of a new setting called "Standard" as a default for a initial view for a graph. And if the setting says max it should include even the outliers.