search

fact-project / fact-tools

The fact-tools are an extension to the streams framework to analyse the data of the First G-APD Cherenkov Telescope.
http://sfb876.tu-dortmund.de/FACT/
GNU General Public License v3.0
6 stars 1 forks source link

UnixTimeUTC and BoardTime are repeated for 100 times #389

Closed maxnoe closed 5 years ago

maxnoe commented 5 years ago

When reading zfits files, 100 consecutive events have the same UnixTimeUTC and BoardTime.

This smells like a bug related to zfits tiles.

maxnoe commented 5 years ago

Happens e.g. for file 20131011_222.fits.fz

maxnoe commented 5 years ago

It's not the file, I checked with https://github.com/fact-project/zfits

maxnoe commented 5 years ago

From https://docs.oracle.com/javase/8/docs/api/java/nio/ByteBuffer.html#asShortBuffer

The content of the new buffer will start at this buffer's current position. Changes to this buffer's content will be visible in the new buffer, and vice versa; the two buffers' position, limit, and mark values will be independent.

So the buffer does not get advanced.

maxnoe commented 5 years ago

fixed in #390

maxnoe commented 5 years ago

Now I came to see that this also affected StartCellData, so all our DRS calibrations also must have been wrong since that change was made.

maxnoe commented 5 years ago

So this should affect all files with ZTILELEN > 1

dneise commented 5 years ago

Now I came to see that this also affected StartCellData, so all our DRS calibrations also must have been wrong since that change was made.

Was this not realized, since the data in the fits files are already DRS-amplitude calibrated a little bit, and then applying a wrong DRS-calibration is not so easily visible anymore?

But still. Does this mean, the currently running analysis should better be restarted?

maxnoe commented 5 years ago

Was this not realized, since the data in the fits files are already DRS-amplitude calibrated a little bit, and then applying a wrong DRS-calibration is not so easily visible anymore?

But still. Does this mean, the currently running analysis should better be restarted?

I jsut checked the ZTILELEN of the files at ISDC, only files from 2013 until mid October have ZTILELEN == 100, after that, most have ZTILELEN == 1, a few have small integer numbers.

This is probably why it never was realized. Only files before 2013-10-22 are affected heavily. All other files are ZTILELEN == 1, 2, 3, so not or only not as badly affected.