I'm looking at the DSPF_sp_fftSPxSP function description in the TMS320C67x DSP Library Programmer's Reference Guide (spru657c). This appears to be a copy of the comments in the DSPF_sp_fftSPxSP.h file from the TMS320C67x DSP Library. There is a section in both that reads:
DSPF_sp_fftSPxSP(512, &x[0],&w[0],y,brev,2, 0,512); is equivalent to: DSPF_sp_fftSPxSP(512, &x[2*0], &w[0] , y,brev,128,0,512) DSPF_sp_fftSPxSP(128, &x[2*0], &w[2*384],y,brev,4,0,512) DSPF_sp_fftSPxSP(128, &x[2*128],&w[2*384],y,brev,4,128,512) DSPF_sp_fftSPxSP(128, &x[2*256],&w[2*384],y,brev,4,256,512) DSPF_sp_fftSPxSP(128, &x[2*384],&w[2*384],y,brev,4,384,512) The four 128 point sub FFTs in the "equivalent to" section are called with a radix of 4. I'm wondering if this is a typo in the documentation and should be calling these with a radix of 2 or does the DSPF_sp_fftSPxSP function allow a non "power of 4" point FFT to called with a radix of 4.