daniestevez
commented
3 years ago Hi Jan, thanks for reporting this. The problem we find here is that the telemetry parser is getting frames that don't match the format it expects. In this cases, it helps to run with --hexdump to see what are the frames that are being decoded. It also helps to look at the SatYAML file to see what is the telemetry parser for the satellite. In the case of AALTO-1 it is the generic ax25 parser, which only parses the AX.25 header. So whatever that is arriving to the parser doesn't look like an AX.25 header.
I'll treat the two observations separately, because I think they are very different.
Observation 3315578
Here I think that the problem is that the transmitter mode is "TLM CW", so the OGG file doesn't have the format we expect for an FSK signal. I wonder if the FSK signal could be recovered from this OGG file, but probably not, because the CW mode sets a too narrow bandwidth.
If we run with --hexdump we get
$ gr_satellites AALTO-1 --samp_rate 48e3 --wavfile satnogs_3315578_2020-12-18T08-42-53.wav --hexdump
* MESSAGE DEBUG PRINT PDU VERBOSE *
((transmitter . 9k6 FSK downlink))
pdu_length = 1
contents =
0000: bd
***********************************
* MESSAGE DEBUG PRINT PDU VERBOSE *
((transmitter . 9k6 FSK downlink))
pdu_length = 12
contents =
0000: 50 a9 0d f4 42 ef 82 ab bd d6 63 8c
***********************************These are garbage decodes (not so unlikely with AX.25), and their length is shorter than an AX.25 header, which explains the problem.
Observation 3315127
Here the mode for the SatNOGS observation is set correctly to GFSK9600. However if we run gr-satellites with --hexdump we see the following garbage decode only:
$ gr_satellites AALTO-1 --samp_rate 48e3 --wavfile satnogs_3315127_2020-12-18T07-10-46.wav --hexdump
* MESSAGE DEBUG PRINT PDU VERBOSE *
((transmitter . 9k6 FSK downlink))
pdu_length = 7
contents =
0000: 40 cc f0 3e ac 18 a0
***********************************Again, this doesn't have the required length for an AX.25 header.
What is the problem with this file. If we look at it in Audacity, we'll see that it is badly saturated and clipped. Even the packets themselves (which have lower amplitude than the noise when there is no packet) have some clipping. This deteriorates the signal quality. Maybe it would be possible to tweak the decoder parameters by hand (--clk_bw, --clk_limit) to try to get a decode (this I haven't tried). This problem with clipping is something that I reported several months ago. Not sure if they've deployed the fix already.
Finally, the decodes that appear in SatNOGS look like garbage to me. The AX.25 headers doesn't seem correct (though I haven't run them through a parser to check this). I've also noticed that the mode is listed as " GFSK Rktr 9600", which seems to refer to Reaktor Hello World, which uses something quite different from AX.25.
janvgils
kng
petrinm
Triggered by the development that is being done by SA2KNG where he tries to integrate gr-satellites decoding into SatNOGS we discovered that the gr-satellites AALTO-1 decoding seems to no longer produce data.
For this test I have used the following two observations:
https://network.satnogs.org/observations/3315578/https://network.satnogs.org/observations/3315127/The SatNOGS decoder in combination with observation 3315127 gives the following output:
Using the following gr_satellites command with the audio from this observation gives the following output:
gr_satellites AALTO-1 --wavfile Recordings/OGG/AALTO-1_satnogs_3315127_2020-12-18T07-10-46.wav --samp_rate 48e3When using the other observation we have a similar output:
gr_satellites AALTO-1 --wavfile Recordings/OGG/AALTO-1_satnogs_3315578_2020-12-18T08-42-53.wav --samp_rate 48e3When I use the example from the satellite-recordings repository the decoding works fine.
There is one addition that can maybe help in solving this, the SatNOGS transmitter definition seem to rely on some "Rktr" framing.
The AALTO-1.yml file uses AX.25 G3RUH framing, I guess it is possible something has changed, maybe the following information on there website already points to a possible explanation:
AALTO-1 satellite information: https://wiki.aalto.fi/display/SuomiSAT/Radio+Amateurs