Greetings readers,
I have the following software: WSJT-X v2.7.2 b4f9a4
I would like to know if there is a mode where I can present the above app with a recorded *.wav file and watch the WJST-X decode the file transactions.
I have recorded an audio file 0f 30m duration. On inspection the SNR in the signals is about +5db - the signal tones are clearly visible in the band.
This file is sampled at 96 ksps and I have tried various sample rates down to 8 ksps
BUT
On playback the signal registers on the applications’ spectrum display.
No decoded output is presented.
. Questions:
Can this software work in the monitor mode?
If so can you offer advice as to how to set up the app.
I think if you follow our guide on Audio Setup, you can use a wave file playback program like Audacity to play back the wave file into a virtual audio device. Connect the other side of the virtual audio device as the input to WSJT-X.
Remember to start playback on the minute changeover mark.
Yes the FT-8 mode (and some others) require the sender and receiver to have synchronized codes, and this is done every 30 seconds for either odd or even transmissions. I forget if the code repeats at 30 seconds or if it is different. I think it’s the same code for each slot.
I actually posted a video on YT of an odd signal on 144 MHz SSB, which a viewer was able to feed in and decode as WSJT-X from Mexico to the US, pretty cool.
I made another recording and it depends where one starts the play back on the block boundaries. On play back I have never so many call signs " in one place" !
Now all I have to do is get that virtual cable to connect your app with the SDR (IC - R8600) controll ap. This is not a problem for this forum.
This FT8 signal really intrigues me. My next exercise is to estimate the signal SNR. My main problem is finding parts of this signal where there is noise only. It looks as if I’ll have to get a noise sample by choosing small bands of noise between signal blocks and adding them.
SNR will be highly dependent upon the bandwidth used for both the signal and the noise measurement. Also keep in mind that it is a lot easier to measure (S+N)/N. If I were you, I would measure the noise by looking at the same bandwidth and the same frequency during the short 1 second break between FT-8 transmissions.
There may also be a way to measure the BER (bit error rate), which might be more meaningful for FT-8 transmissions.
There may be a way to do this with GNU Radio.
You can connect wfview directly to the IC-R8600 over ethernet, and then connect from wfview to any other applications needed.
My noise estimate will be taken from the spaces between the signal blocks. The main assumption being that the noise properties are the same for each noise block. I will test this first. The next thing to do is compute the noise PSD and get an estimate for the noise PSD (W/Hz). The next it to look at the signal energy in each baud interval (?) Please Note: I am still learning about this signal
About BER : Gee whiz! That would be a lot of work, for me, to estimate the BER … huge amount to learn about the signal, modulation, error control coding etc.
Initially I’ll stick my SNR estimates ie. Eb/N0 to improve my understanding.
As I would say " You chaps have invented one hell of a signal "