discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] grc file -- display problem


From: CEL
Subject: Re: [Discuss-gnuradio] grc file -- display problem
Date: Fri, 15 Mar 2019 10:24:29 +0000

Hi Parvathi!

> 1.  why the screen shot is displaying x-axis and y-axis as dB only
i.e.  2.03dB, -8.83dB.

Displaying the frequency (x-axis) in dB is most definitely a bug. Which
version of GNU Radio are you using (`gnuradio-config-info --version`
has the answer)? We've fixed that a while ago.

Displaying the y-axis in dB is a pretty common thing to do for PSD
estimates.

> 2.   How to interpret the magnitude of the output and on what basis

Not quite sure what you're asking here – that plot is a power density
estimate over frequency. The dB are relative to "full scale".

You're making mistakes when configuring the Qt GUI frequency sink –
your center frequency is simply what the center of your x-axis labeling
is shifted to; since you're dealing with baseband only, it needs to be
zero.

> 3. We would basically want to write the peak of the signal into a
file to record continuously and request you to give an example for the
same. 

That sounds like you want a spectral estimator, not a visualization!
So, you wouldn't do that with the Qt GUI frequency sink – which is just
something to look at, not something to generate data.

The operation done by the Qt GUI frequency sink is really just

stream to vector (length=1024) ->
FFT (length=1024) (including windowing) ->
complex to magnitude^2 ->
convert to logarithmic scale

You could do the same, write a small Python block to find the bin with
the maximum value, and write that bin's index to a file. Should be
pretty easy, if you know Python!

Generally, from a scientific point of view, it's questionable whether
this FFT-based route is an appropriate estimator for whatever you want
to do – in the end, you're quantizing your frequency in to FFT-length
number of steps, and you get horrendous leakage effects if a frequency
isn't exactly N·f_sample/L_FFT for some integer N.

But: the frequency estimator suitable for your problem can not be
advised on without knowing which problem you're trying to solve, or
which question to answer in the end.

Best regards,
Marcus

On Fri, 2019-03-15 at 09:48 +0530, Mrs.  Parvathi Piduparthi wrote:
> hello
> 
> We have installed gnu radio and working with ubuntu .  Attached is the grc 
> file and its output which basically gets a signal from source and tries to 
> display the fft  of the same as a spectrum. W.r. to these , kindly clarify 
> the following
> 
> 1.  why the screen shot is displaying x-axis and y-axis as dB only i.e.  
> 2.03dB, -8.83dB. 
> 
> 2.   How to interpret the magnitude of the output and on what basis
> 
> 3. We would basically want to write the peak of the signal into a file to 
> record continuously and request you to give an example for the same. 
> 
> thanks &  regards
> Parvathi 
> Scientist/Engineer 
> National Atmoshperic Research Laboratory , Department of Space
> Gadanki, Tirupathi
> Andhra Pradesh 
> India
> 
> Ph. 91 877 2500583
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

Attachment: smime.p7s
Description: S/MIME cryptographic signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]