cbassa / strf

Radio Frequency Satellite Tracking
GNU General Public License v3.0
163 stars 36 forks source link

Problem with large number of bins when passing 00 UTC #49

Closed zanco closed 1 year ago

zanco commented 1 year ago

have this recording with 1463 bin files

I run plot for the specific frequency range with :

rfplot -p 2023-05-01T11:33:41 -z 0.0006 -s 360 -l 72000 -f 436025000 -w 170000

2023-05-01T11:33:41_001464.bin does not exist Read spectrogram 1700 channels, 72000 subints Frequency: 436.025 MHz Bandwidth: 0.17 MHz Loaded 7503 orbits Traces for 52 objects for location 2671

So, my recording / logging started on the 1st of may around 11:33 UTC.

This is the plot result, because starting from s 360 the start time is just before 18:00 UTC

plot example

I walk through the plot with backspace

That shows at a certain moment this output:

plot zoomed

Which has to be from 2023-05-02 unknown satellite starting at 11:12 UTC

So, running s and f exits in segmentation fault

But if I plot the same file starting at -s 1320

plot partly

The UT Date at the bottom is 2023-05-02

s and f does not terminate in an segmentation fault

match plot

Astrocast-0105 , no segmentation fault

So, the big question is, should the plot switch to the new date if you pass the 00 UTC timestamp when "tabbing" throug and observation, and if yes, would that solve the segmentation fault I ran into and got solved by only plotting a smaller part of the strf obs ?

Thanks Ben

cbassa commented 1 year ago

Indeed, when panning through the waterfall, the x-axis label doesn't update when you cross into a later date. This is a minor issue which I'll think about if it needs fixing. As for the segmentation fault, I do not think this is related to this date issue, so may be something else. That's impossible for me to check without the same data.