ResScan-aligning 30 sec graph w 8 hour graph

Why doesn't the 30 sec slice of time below, align up more closely to the time in the top graph?
The time in the 30 sec slice is from 01:10:00 to 01:10:30. The vertical line(s) in the top graph are at 00:45 approx.


Oldern
Edited to show the image.

I can't help with the ResScan question but I can help with the Image thing. The only thing I use ResScan for is the red line for large leak territory. I use SleepyHead when I want to zoom in. I have always had trouble managing the graphs in ResScan when trying to zoom in.

When you view your image there is a box on the right that has 4 addresses...copy the last line in that box and paste it in the body of your message here...no need to use the IMG buttons on the forum editor because the last line contains the IMG brackets.
You will see the address in code only...but will see the image once you click submit or try preview...like this

I just looked at ResScan and the graph times match perfectly. I'm using an S8 Autoset II.

Do you have the latest version of ResScan (v4.2) installed?

And, BTW, this is not the first bug reported in ResScan.

Is that the data "skew" that was reported early on for the S9 AutoSet?
It may have been resolved in subsequent Software Versions (internal software version in the S9 itself & not the ResScan software) - but I have no documentation of that fact.

If that is your problem, you can resolve it by periodically formating the SD card (every 2 weeks seems to work - but more often certainly would not hurt).

I rotate 2 SD cards & always format each one every time I download the data (once a week) - so, I have never seen that data skew.

What is the Software Version in your S9 AutoSet?

EDITED to add:
I see you have an older Humidifier - so, I guess you DON'T have the S9 AutoSet?

Oldern wrote:Why doesn't the 30 sec slice of time below, align up more closely to the time in the top graph?

Don't know. This issue was part of the data skew problem discussed quite a bit 2 years ago. The pragmatic view, IMO, is it's not a big deal and not worth getting worked up about. If you really need to pin down a time for an event the 30 second graph is the more accurate. In any case, the S9 has variability in its time stamp writing routine and from what I have studied any time stamp you look at could be as much as 30 seconds plus or minus from true time.

The timelines in Resscan for the high resolution data, the detail data, and the events graph, can have as much as one minute difference among them. If you have a break in therapy during the night the timelines can shift so that the differences before the break are not the same as after the break. I have not been able to decipher any rhyme nor reason for the timeline shifts. In the grand scheme of CPAP therapy they don't matter.

I am using ResScan 4.2.08.5438 and my machine is the older "enhanced" VPAP Adapt SV. I download my data via cable.

I once tried the Sleepyhead S/W but couldn't get it to work. Without getting a response when I tried to find out if SH SW would work for the VPAP Enhanced, I assumed it would not.

If anyone knows whether it will or will not work on the VPAP Adapt SV enhanced, let me know.

The graph thingy I can't figure out. Has anyone communicate directly with ResMed about ResScan currently?

Thanks for the responses.
Oldern

idamtnboy wrote:

Oldern wrote:Why doesn't the 30 sec slice of time below, align up more closely to the time in the top graph?

Don't know. This issue was part of the data skew problem discussed quite a bit 2 years ago. The pragmatic view, IMO, is it's not a big deal and not worth getting worked up about. If you really need to pin down a time for an event the 30 second graph is the more accurate. In any case, the S9 has variability in its time stamp writing routine and from what I have studied any time stamp you look at could be as much as 30 seconds plus or minus from true time.

The timelines in Resscan for the high resolution data, the detail data, and the events graph, can have as much as one minute difference among them. If you have a break in therapy during the night the timelines can shift so that the differences before the break are not the same as after the break. I have not been able to decipher any rhyme nor reason for the timeline shifts. In the grand scheme of CPAP therapy they don't matter.

I just reread your opinion as to what is a big deal and not worth getting worked up about. I decided it upset me the second time as it did the first time I read it.

In the first place, I was not worked up nor was I getting worked up. I simply posed a question about a problem which has been showing up in my graphs for months and months which I didn't understand.

Just now, I selected one (1) hour in the lower graph and an 8 hour graph above. By moving the slider under the lower graph, I got the two vertical bars in the upper graph set on the time from 22:00 to 23:00. The time in the lower graph shows 22:15 and change to 23:20 and change on the right side-a difference of 15 to 20 minutes. Not much if you say it fast. But if you are trying to work the two graphs together, it can be frustrating at times-at least it is to me. I find I have to do some mental gyrations to think how to move either of the sliders if I can't quickly highlight the point in time I am interested in. ymm(must)v.

At any rate, I thank you for the other comments.

btw, my machine type hasn't been showing up in my signature. I'm checking on that.
Oldern

I don't see how the problem can be with the machine's software. ResScan is displaying the exact same
data in both graphs, the only difference is the time axis scale used in the graphs.

That must be a ResScan display problem, no?

And that's the same version of ResScan that I am using.

Oldern wrote:I just reread your opinion as to what is a big deal and not worth getting worked up about. I decided it upset me the second time as it did the first time I read it.

In the first place, I was not worked up nor was I getting worked up. I simply posed a question about a problem which has been showing up in my graphs for months and months which I didn't understand.

Sorry about that. I guess I read more into your question than was justified. My bad. I apologize. (Another example of the deficiency of written language vs. face to face spoken.) I really wasn't thinking in terms of you being worked up about it. I just was saying I didn't see it as a big deal.

I don't recall seeing that great a difference between the graphs last year, but maybe I didn't pay attention to it. I do know that the closer to the same scale both graphs are, the closer the timelines coincide, and the greater the time scale difference between them, the greater the time difference. I just now looked at some data from a couple of weeks ago and over a year ago. The graphs coincide quite closely, more so than I remember seeing when I was using version 4.13. I now have 4.2. Just for giggles, take a look at the properties/detail tab for Resscan.exe and see what the version number is. Mine is 4.2.8.5438 dated 8/29/2012.

I just tried the same comparison as you're mentioning, 8 hour top graph, 1 hour bottom graph. The times are within a few minutes. Makes me wonder if there is something in the XML file that contains a list of settings for Resscan that is different in your XML from my XML.

idamtnboy,

Got your message and all is good.

I am using the same build on ResScan as you show.

XML is way over my head. If there is a way or something for me to check, I'll be glad to give it a try.

When you said the times in your graphs were within a few minutes, what is a few minutes?

Thanks for comments. Further suggestions welcome.
Oldern

Oldern wrote:idamtnboy,

Got your message and all is good.

I am using the same build on ResScan as you show.

XML is way over my head. If there is a way or something for me to check, I'll be glad to give it a try.

When you said the times in your graphs were within a few minutes, what is a few minutes?

Thanks for comments. Further suggestions welcome.
Oldern

As you can see on the following pic, (you have to d/l the pic from the link below) they are just about dead on. I did a 30 min vs. 8 hr graph for this. The XML file is a configuration file that Resscan reads to set parameters according to the machine that the data is from, plus graph settings. It's been a long time since I looked at it and I don't know if there is any setting in there that would cause the graphing problems. You do have, or see, the S9 as the pertinent machine when you download the data, right? The machine ID is part of the ID file on the SD card so Resscan should be using all the correct parameters. I don't know why it wouldn't.

https://files.secureserver.net/0sbJFLxYPU7cPd

I know it doesn't help, but I'm running Ver 4.2.08.5438 and mine seem to be nearly perfectly aligned.


I did notice that if you change the size of the window after selecting the time slice, it will move the centering some.

Oldern wrote:I find I have to do some mental gyrations to think how to move either of the sliders if I can't quickly highlight the point in time I am interested in. ymm(must)v.

If you "click" on the event you are interested in in the top display, it should center that in the bottom display - A little easier than moving both sliders.

Oldern wrote:I find I have to do some mental gyrations to think how to move either of the sliders if I can't quickly highlight the point in time I am interested in. ymm(must)v.

Jay, this is a response to your comment about moving the sliders.

Sorry about the delay in responding-I had to have a minor surgical procedure performed and today I am catching up. Thanks for your remarks.

I'm looking at the graph I put in my original post-top graph 8 hours and bottom graph 30 seconds. The two vertical lines in the upper graph, very close together, naturally, in order to represent 30 seconds, lie at about 00:47 AM. The single red line in the bottom graph sits close to 01:10:16 AM, lying about half way between 01:10:00 and 01:10:30. While I understand what you are saying about moving (or not moving) the sliders for alignment, this makes me think the two graphs are out of vertical alignment with each other by about 20+ minutes.

Another example: Top graph, same date as above, set 8 hours, bottom graph at 1 minute. Two vertical lines above (enclosing a time slice of 60 seconds on an 8 hour graph) are as close as I can get them to 01:00. The red line in the bottom graph at a 1 minute interval sits at about 01:21:56, about 21 minutes from where I think it should be.

What makes this frustrating to me is that I like to be able to compare several of the parameters with each other, at the exact (well, not exactly the exact same time) but at close to the same time. I have had to resort to the sliders which you would think would make aligning two graphs to the same point in time would be easy. But not with me. There may be something obvious I am overlooking. Let me know.

Oldern

Oldern, I'm not sure but I still think it's the sliders that are causing your problem. See my example below. I did not use the sliders at all. I set the top scale at 8 hours and the bottom scale at thirty minutes. The top and bottom charts were only roughly aligned at this point. Then, I left clicked on the point of the leak in the top chart which brought the bottom chart into alignment, centered approximately at the time of the leak in the top chart. So in the bottom chart, I see approximately 15 minutes on either side of the point of interest selected in the top chart.
Again, I did not use the slider at all to produce this result. I selected the scales and clicked on the event in the top chart. That's it. I would find using the sliders to do this difficult, too. In using a different scale, like one minute, in the bottom chart, it might take several clicks on the top chart to center the bottom to my liking.

Jay

Odern, I must confess I don't know why the center of the time slice in your top graphs appears to be displaced from the centered time in the bottom graph. As you said at the start of the thread, the center of the time slice in the top should correspond to the centered time (red line) in the bottom graph. I can not make mine do anything but. Sliders or no.

Jay