usage of keep_one_in_n in fftsink_nongl

View: New views
2 Messages — Rating Filter:   Alert me  

usage of keep_one_in_n in fftsink_nongl

by adib_sairi :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Dear All,
I am exploring the fftsink_nongl.py file because i want to modify it to be the sensing method of my project. in the fft_sink_c class, there is a signal processing block call keep_one_in_n. I think that this block is use to slowing down the sampling of the spectrum sense, am I correct? if i remove this block from the flow graph, the output value will be unexpected (unreliable). can any one tell me what is actually happen? what did i need if i want to remove this block? because this block make my sensing speed extremely slow..

best regards,
Adib

Re: usage of keep_one_in_n in fftsink_nongl

by Josh Blum-2 :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

sample_rate/fft_size/N = fft_rate

N = sample_rate/fft_size/fft_rate

The fftsink_nongl.py wants to decimate the stream of fft vectors by N to
achieve fft rate (aka frames per second).

If you tell fftsink_nongl.py a sample rate that is greater than your
actual sample rate, then N will be too large, and the frame rate will be
very slow.


-Josh


adib_sairi wrote:

> Dear All,
> I am exploring the fftsink_nongl.py file because i want to modify it to be
> the sensing method of my project. in the fft_sink_c class, there is a signal
> processing block call keep_one_in_n. I think that this block is use to
> slowing down the sampling of the spectrum sense, am I correct? if i remove
> this block from the flow graph, the output value will be unexpected
> (unreliable). can any one tell me what is actually happen? what did i need
> if i want to remove this block? because this block make my sensing speed
> extremely slow..
>
> best regards,
> Adib


_______________________________________________
Discuss-gnuradio mailing list
Discuss-gnuradio@...
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio