• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 98 subscribers
  • Views 1392 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Right way to call multi pass DSPF_sp_fftSPxSP()?

Rob Pira
Prodigy 80 points

I'm using the c6657 dev board.  spra947a.pdf has an example for N=1024:

/* stage one */

DSPF_sp_fftSPxSP(N, &x[0], &w[0], y, brev, N/4, 0, N);
/* stage two */
DSPF_sp_fftSPxSP(N/4,&x[2*3*N/4], &w[2*3*N/4], y, brev, radix, N*3/4, N);
DSPF_sp_fftSPxSP(N/4,&x[2*2*N/4], &w[2*3*N/4], y, brev, radix, N*2/4, N);
DSPF_sp_fftSPxSP(N/4,&x[2*1*N/4], &w[2*3*N/4], y, brev, radix, N*1/4, N);
DSPF_sp_fftSPxSP(N/4,&x[0], &w[2*3*N/4], y, brev, radix, 0, N);
The result in 'y' is always incorrect. The correct result is shown by calling DSPF_sp_fftSPxSP() with 1024 in the first and last parameters:
DSPF_sp_fftSPxSP(1024, &x[0], &w[0], y, brev, radix, 0, 1024);
I'm calling the twiddle factor generator defined in the DSPLIB DSPF_sp_fftSPxSP() example in both cases. Is this correct? I'm also double word (8) aligning the input and output and twiddle arrays. What am I doing wrong in the multipass instance?
  • 0
    TI__Guru* 81335 points

    Welcome to the TI E2E forum. I hope you will find many good answers here and in the TI.com documents and in the TI Wiki Pages (for processor issues). Be sure to search those for helpful information and to browse for the questions others may have asked on similar topics (e2e.ti.com). Please read all the links below my signature.

    We will get back to you on the above query shortly. Thank you for your patience.

    Thank you.

    Note: We strongly recommend you to create new e2e thread for your queries instead of following up on an old/closed e2e thread, new threads gets more attention than old threads and can provide link of old threads or information on the new post for clarity and faster response.

  • 0
    TI__Guru* 81335 points

    Hi Rob,

    Please check whether below threads answers your query or not.

    Thank you.

  • 0 in reply to Raja
    Prodigy 80 points

    None of those threads deal with multi pass. 

    As I said, I have implemented DSPF_sp_fftSPxSP() and it works for the single pass case with 1024 bins. It fails with an incorrect result when tried with the multi pass use described in TI documentation. Is there a documentaion error? Am I calling it correctly? I did copy the SPRU's example directly for 1024 yet it produces incorrect results.

  • 0 in reply to Rob Pira
    TI__Genius 12805 points

    Rob, here is what I suggest, if you already did it we will continue

    1. Run the simple 256 complex FFT example from the release and make sure that it works for you. Look at directory dsplib_c66x_3_4_0_0\examples\fft_sp_ex of the release

    2. Do search on FFT decimation in frequency.  This will give you the basic understanding of how the multi-path is working. The first pass uses the global twiddle factors (generated with the "real" N) but the number of passes is not log(N) (or half of it when Radix 4 is used) but rather until the blocks are of 256 complex elements.  The Web is full of materials about Decimation in Frequency.

    Again, understand decimation in frequency and how it works

    Best regards

    Ran

  • 0 in reply to ran35366
    TI__Genius 12805 points
    I close the thread