FFT Scaling

GNU Radio
1

There is an internal scaling effect of fftlen (the length of your FFT)
in
the FFT block. How do I unscale or remove the scaling factor in the
python
generated by thr GRC? I tried the one suggested in the faq,

… forward FFT --> multiply_const_cc(fftlen*[1.0/fftlen,]) --> reverse
FFT
…

but it doesn’t seem to work. I know I am missing something. How do I
insert
this into this line of code in my python?

  1. self.fft_vxx_0 = fft.fft_vcc(1024, True, (window.rectangular(1024)),
    True, 1)
    or here?
  2. self.blocks_complex_to_mag_squared_0 =
    blocks.complex_to_mag_squared(1024)

Thank you in advance.

Attached is my flow graph.

Best,

*Gerome Jan M. Llames *

“Design is not just what it looks like and feels like. Design is how it
works.” - Steve Jobs

2

On Wed, Aug 12, 2015 at 10:51 AM, Gerome Jan L
[email protected]
wrote:

  1. self.fft_vxx_0 = fft.fft_vcc(1024, True, (window.rectangular(1024)),

Best,

*Gerome Jan M. Llames *

“Design is not just what it looks like and feels like. Design is how it
works.” - Steve Jobs

What do you expect the constant to be? fftlen*[1.0/fftlen,] = [1],
unless
I’m having a really slow day.

3

On Wed, Aug 12, 2015 at 12:31 PM, West, Nathan
[email protected]
wrote:

but it doesn’t seem to work. I know I am missing something. How do I
Attached is my flow graph.

What do you expect the constant to be? fftlen*[1.0/fftlen,] = [1], unless
I’m having a really slow day.

It’s been pointed out to me that I am having a slow day. Taking a step
back… when you say “it doesn’t seem to work”, what exactly does that
mean? What do you observer that is different from what you expect?

4

On 08/12/2015 05:39 PM, West, Nathan wrote:

    |or here?

Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page

Try multiply_const_cc( 1.0/fftlen, vlen=fftlen )

  • Jeff