I am getting a lot of random viewer (1.0.2 and 0.6) crashing when reading an EDF file.
It may be the EDF file! A recent convert from Somnologica, where the 16 bit AD conversion is done on a scale of +/- 2.5 Volts (EEG is only pre-amplified by 10). The signal "gains" are tweaked up for display.
To get these signalsto look right on XDF viewer, I have to set the Digital Min/Max parameter to +/- 8! (from +/- 32767).
Cheers.
OpenXDF.org
Discussion forum for the OpenXDF polysomnography digital storage standard
crashes galore
3 posts • Page 1 of 1
crashes galore
by derek.eder on Wed Aug 23, 2006 10:27 am
Derek N. Eder
SU/Sahlgrenska
Utvecklingslab 1, Med
Gröna stråket 8
SE 413 45 Göteborg (Gothenburg)
Sverige (Sweden)
+46 (031)* 342 8261 (28261 inom Sahlgrenska)
+46 0709 721 283 (mobile)
+46 (031) 25 97 07 (home)
SU/Sahlgrenska
Utvecklingslab 1, Med
Gröna stråket 8
SE 413 45 Göteborg (Gothenburg)
Sverige (Sweden)
+46 (031)* 342 8261 (28261 inom Sahlgrenska)
+46 0709 721 283 (mobile)
+46 (031) 25 97 07 (home)
- derek.eder
- Posts: 2
- Joined: Thu Dec 23, 2004 6:56 am
- Location: Gothenburg, Sweden
by rwidell on Wed Sep 06, 2006 2:49 pm
Derek,
Sorry it took me so long to respond. I have been out of the office for a while.
The best thing to do in the case of a crash is post the exception number and address if Windows gives them to you when XDFviewer crashes. With the exception number and address, I can usually find the problem pretty quickly.
In pre-XP versions of Windows, you usually get a message like: "Unhandled exception 0xC00000005 at 0x00004000 in XDFviewer.exe" I just need that information.
In XP, you usually get a dialog telling you the program had a problem and needed to close. You have to click the link that reads something like "View technical information about the problem" to get the above information.
In the next version, I will add a feature to automatically send the information to us when it crashes.
As far as the incorrect signal scale...
The EDF spec provides a field for the signal's dimension. Since there is no strict rule on how this field should be used, XDFviewer just looks for "V", "mV", or "uV". It sounds like there is either no unit specified or it is not something XDFviewer recognizes.
When you open an EDF file, XDFviewer creates a XDF header file. If you backup the existing one and open the EDF file again, you will have a clean header file. Once you do that, you can open the new XDF header in notepad and search for "<xdf:Source>". From there, you should be able to tell how XDFviewer is interpreting the EDF signals. The physical min/max or unit tags may have incorrect values.
Sorry it took me so long to respond. I have been out of the office for a while.
The best thing to do in the case of a crash is post the exception number and address if Windows gives them to you when XDFviewer crashes. With the exception number and address, I can usually find the problem pretty quickly.
In pre-XP versions of Windows, you usually get a message like: "Unhandled exception 0xC00000005 at 0x00004000 in XDFviewer.exe" I just need that information.
In XP, you usually get a dialog telling you the program had a problem and needed to close. You have to click the link that reads something like "View technical information about the problem" to get the above information.
In the next version, I will add a feature to automatically send the information to us when it crashes.
As far as the incorrect signal scale...
The EDF spec provides a field for the signal's dimension. Since there is no strict rule on how this field should be used, XDFviewer just looks for "V", "mV", or "uV". It sounds like there is either no unit specified or it is not something XDFviewer recognizes.
When you open an EDF file, XDFviewer creates a XDF header file. If you backup the existing one and open the EDF file again, you will have a clean header file. Once you do that, you can open the new XDF header in notepad and search for "<xdf:Source>". From there, you should be able to tell how XDFviewer is interpreting the EDF signals. The physical min/max or unit tags may have incorrect values.
- rwidell
- Posts: 43
- Joined: Fri May 21, 2004 8:22 am
- Location: Gainesville, FL
by rwidell on Tue Sep 26, 2006 10:47 am
Derek,
I received a crash report this morning (maybe from you), and I think I see the problem. I examined the crash report, and the EDF file the user was trying to read had a data record length of 0.0039069s (256 Hz).
This is a valid value according to the EDF spec. I guess I overlooked that. The good news is, this is a pretty simple fix. I am in the process of validating the fix now, I will try to get 1.0.4.0 out by the end of the week.
The OpenXDF spec fixes the frame length to integers between 1s and 30s, so I probably need to relax that requirement to account for fractional frame lengths.
Hopefully, 1.0.4.0 will solve your problems.
I received a crash report this morning (maybe from you), and I think I see the problem. I examined the crash report, and the EDF file the user was trying to read had a data record length of 0.0039069s (256 Hz).
This is a valid value according to the EDF spec. I guess I overlooked that. The good news is, this is a pretty simple fix. I am in the process of validating the fix now, I will try to get 1.0.4.0 out by the end of the week.
The OpenXDF spec fixes the frame length to integers between 1s and 30s, so I probably need to relax that requirement to account for fractional frame lengths.
Hopefully, 1.0.4.0 will solve your problems.
- rwidell
- Posts: 43
- Joined: Fri May 21, 2004 8:22 am
- Location: Gainesville, FL
3 posts • Page 1 of 1
Who is online
Users browsing this forum: No registered users and 1 guest