I and Q file recording format

[Andrew Rich]

Can I ask what the file format is for the I and Q recording in gqrx

Why ?

I then go on to use inspectrum and I am interested to know what format gqrx records I and Q as opposed to other I and Q recording software

I am recording 1090 MHz data and viewing it in inspectrum

I am also wondering then does the dynamic range of the SDR in use impact on what I am seeing

For example the rtl hardware returns 8 bit resolution vs my sdr play device that returns 12 bits.

I was watching aircraft on my flightaware dongle on 1090 MHz in gqrx but I am not really seeing a great amplitude in gqrx spectrum

Yet the reciever can see some 100 kms range

I would have expected the spectrum view to return some more distinct signal strength

When run with dump 1090 it decodes nicely

Andrew

Sent from my iPhone

[Robin Gape]

Andrew,

I suspect that you'll find what you need here:
https://github.com/csete/gqrx/issues/301

Good luck,

Robin, G8DQX

[Andrew Rich]

Thanks robin

I read the gqrx docs but could not find it

There is no meta data. Just raw samples in float complex format, i.e. 32 bit floats interleaved IQIQIQ... samples. It's the same format you get with a gnuradio file sink.

One thing I don't get is when you replay,

How does the sdr program know that you should go back to the start of the frequency spectrum

Does that make sense ? 

IQ................IQ...................IQ

IQ................IQ....................IQ

And if you don't know the origanal sampling rate does that bugger up the spectrum alignment ?

I am looking for OOK so I guess I just need to bother with the I values ? 

I am hoping I have this right that the I values represent the instaneous values in time of a spectrum like a fast sweeping spectrum analyser 

A 

Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "Gqrx SDR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gqrx+uns...@googlegroups.com.
To post to this group, send email to gq...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gqrx/631fc8be-80b3-bddb-f2bc-7ca97de5c501%40googlemail.com.
For more options, visit https://groups.google.com/d/optout.

[Robin Gape]

Andrew,

it really helps to be familiar with the theory first! [Carpe textbook or equivalent!]

That said, the IQ samples represent data at a centre frequency, plus/minus, ±, half the sampling frequency. So the spectrum represented is:

    (centre frequency - (sampling frequency)/2) to (centre frequency + (sampling frequency)/2)

The (sampling frequency)/2 arises from the Nyquist criterion, and the pairs of samples (I and Q) give rise to the plus/minus offset from the centre frequency. That centre frequency is the frequency of the hardware oscillator that your dongle appears to posses. [In the case of an RTL dongle, there is some trickery involved, but the principle remains.]

It is up to you, as the user, to tell the program that consumes (analyses or demodulates) the sample file what the sample rate and centre frequency are, since there is no metadata in the file to provide that information. If that information is not recorded by you, then the experimenter will be sitting in the corner with the Dunce's cap on!

Inspectrum and GQRX will analyse pairs of IQ samples to resolve the plus/minus spectra. The program does this as part of its essential function. If it did not, the spectrum above the centre frequency and the spectrum below the centre frequency would be folded in to one another. The display of frequency would be ambiguous, and two frequencies would be received at once.

The analysis is independent of the modulation method used.

The I and Q values mathematically represent, after Fourier analysis, a spectrum whose width is that of the sampling frequency. [There is a much longer and more complete version of this analysis, of course!]

HTH,

Robin, G8DQX

PS: Depending on the use case, there may be other ways of achieving what you're looking for.