
Technical discussions about the TI C54x DSPs (including the c5401, c5402, c5402a, c5404, c5407, c5409, c5409a, c5410, c5410a, c5416, c5420, c5421, c5441, c549, c5470 and c5471).
|
Hello Torgeir TI has a optimized Q31 FFT - but I has not been released yet so you can't find it on their homepage. TI has sent me te routine a few weeks ago - you?ll find it in the attachment. Documentation is available at TI?s homepage - look for SPRA554 (sorry - I have only a printet version - the pdf file got somewhere lost ...). I have not testet the routine yet (still working on approved selfmade C / ASM - Mixed Q31 FFT) - please give me a short info if the routine realy work! Best regards Thomas -----Ursprungliche Nachricht----- Von: Torgeir Jakobsen [SMTP:] Gesendet am: Freitag, 9. Juli 1999 16:07 An: Betreff: [c54x] Real FFT for C54x Hi I need an optimized double precision ( fixed point Q31 format ) 1024 point real FFT routine for the TMS320C54x DSP family. The DSPLIB from Texas only contains single precision ( Q15 format ) FFTs. Best regards Torgeir Jakobsen e-mail: << Datei: ATT00000.html >> ------------------------------------------------------------------------ eGroups.com home: http://www.egroups.com/group/c54x http://www.egroups.com - Simplifying group communications | |||
|
|
Hello Thomas Thank you for your response. I have already got these files from Texas. I hoped somone had a real FFT. I imagine that a real FFT is faster than a complex FFT. I can use the routine from Texas an set the imaginary values to zero. I have not testet the routines yet. I have only tried to link them to the existing DSPLIB ( version 1.0 ). I have found a few faults. In addition to the files cfft_32 and cifft_32.asm you also have to include the cfft1024 and cifft1024 to do a 1024 point fft and inverse fft ( or the other cfft???.asm and cifft???.asm files depending on how many points you have ). One of the .include statements in the cifft1024.asm ( and cifft???.asm files ) is wrong. They have written .include cfft_32.asm but the correct is cifft_32.asm. Texas told me that I should condsider the files as a BETA-version, since they have not testet all the files yet. Best regards Torgeir Jakobsen e-mail: -----Original Message----- From: Thomas Hupp <> To: ' <> Date: 9. juli 1999 20:18 Subject: [c54x] AW: Real FFT for C54x >Hello Torgeir > >TI has a optimized Q31 FFT - but I has not been released yet so you can't find it on their homepage. >TI has sent me te routine a few weeks ago - you?ll find it in the attachment. > >Documentation is available at TI?s homepage - look for SPRA554 (sorry - I have only a printet version - >the pdf file got somewhere lost ...). > >I have not testet the routine yet (still working on approved selfmade C / ASM - Mixed Q31 FFT) - please give me >a short info if the routine realy work! > >Best regards >Thomas >-----Ursprungliche Nachricht----- >Von: Torgeir Jakobsen [SMTP:] >Gesendet am: Freitag, 9. Juli 1999 16:07 >An: >Betreff: [c54x] Real FFT for C54x > >Hi > >I need an optimized double precision ( fixed point Q31 format ) 1024 point >real >FFT routine for the TMS320C54x DSP family. The DSPLIB from Texas only >contains single precision ( Q15 format ) FFTs. > >Best regards >Torgeir Jakobsen >e-mail: > > > << Datei: ATT00000.html >> > >------------------------------------------------------------------------ >Just Tell Us What You Want... >Respond.com - Shopping the World for You! >http://clickhere.egroups.com/click/390 >eGroups.com home: http://www.egroups.com/group/c54x >http://www.egroups.com - Simplifying group communications ------------------------------------------------------------------------ eGroups.com home: http://www.egroups.com/group/c54x http://www.egroups.com - Simplifying group communications |