In 0.8.6, I have been trying to use preferences to downsample and reduce the statistical data in order to perform a long battery test. There appears to be certain combinations of reduction and sample rate that do not work. I started recording with 1Hz reduction and 20Hz sampling, but when I stopped the recording the next day I only got a 2K file. It’s not clear to me how this is supposed to work, because with some settings I can start and stop recording and the live data continues to display. Other times when I start a data log file, after I OK the file name(I get the busy cursor for about 5 seconds and then the acquisition stops. When I start it again, the data is apparently not being logged. All I get is the header information when I stop recording. Not sure if this is a bug or if there needs to be additional instruction on downsampling and logging from the UI in 8.6.0.
Hi @wired,
Would you mind posting a Joulescope log where this happened? You can open Help → View logs… The most recent log file is for the current run of the Joulescope UI, and the next most recent is for the previous run.
Node that reduction should have no effect on the JLS recording.
Thanks!
I just did a quick Joulescope run this morning. Please see attached log file.joulescope_20200320_135850_54592.log (112.1 KB)
This log does contain some interesting information with a good lead. I will investigate.
Hi @wired - I found and fixed the issue. The root cause was that at 20 Hz, the new data was “empty”.
The recording module did not appropriately handle this case. The code is fixed in:
- Improved handling of empty JLS files and zero length data. · jetperch/pyjoulescope_ui@37c6df6 · GitHub
- Fixed DownsamplingStreamBuffer.samples_get zero length handling. · jetperch/pyjoulescope@695b175 · GitHub
The next UI release (hopefully this coming week) will contain the fix.
Thank you for the log file which allowed me to easily duplicate and trace down this issue!