Top Speed of DFT

What is the highest frequency that could be handled with discrete
fourier transform software?


Bret Cahill

Bret Cahill wrote:

> What is the highest frequency that could be handled with discrete
> fourier transform software?

http://www.fftw.org/benchfft/


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

On Sep 29, 1:16&#4294967295;pm, Bret Cahill <BretCah...@peoplepc.com> wrote:
> What is the highest frequency that could be handled with discrete
> fourier transform software?
>
> Bret Cahill

Fs/2

On Sep 29, 1:16&#4294967295;pm, Bret Cahill <BretCah...@peoplepc.com> wrote:
> What is the highest frequency that could be handled with discrete
> fourier transform software?
>
> Bret Cahill

A simple diffraction grating will give you a dft for a single
frequency bin for optical frequencies. Crystal lattices will do this
for x-ray and gamma frequecies.

Clay

>
>
>Bret Cahill wrote:
>
>> What is the highest frequency that could be handled with discrete
>> fourier transform software?
>
>http://www.fftw.org/benchfft/
>
>
>Vladimir Vassilevsky
>DSP and Mixed Signal Design Consultant
>http://www.abvolt.com
>

Geez,  FFTW3 seems to cream most of the other implementations in almost all
tests on every architecture tested on the site.  Pretty impressive.

Jacob

Clay <clay@claysturner.com> wrote:
(snip)
 
> A simple diffraction grating will give you a dft for a single
> frequency bin for optical frequencies. Crystal lattices will do this
> for x-ray and gamma frequecies.

It does seem that gamma diffraction can work.  There is even
a report of a gamma lens based on rings of crystals with the
appropriate spacing and angle to diffract gamma rays to a
small focal spot.

-- glen

> > What is the highest frequency that could be handled with discrete
> > fourier transform software?
>
> http://www.fftw.org/benchfft/

Phew!  What a relief!  For awhile there I was concerned about 10^-36
errors but no more!

> Vladimir Vassilevsky
> DSP and Mixed Signal Design Consultanthttp://www.abvolt.com

Bret Cahill wrote:

>>>What is the highest frequency that could be handled with discrete
>>>fourier transform software?
>>
>>http://www.fftw.org/benchfft/
> 
> 
> Phew!  What a relief!  For awhile there I was concerned about 10^-36
> errors but no more!

There is no need to stick out; everybody already knew what you are.

VLV

> >>>What is the highest frequency that could be handled with discrete
> >>>fourier transform software?
>
> >>http://www.fftw.org/benchfft/
>
> > Phew! �What a relief! �For awhile there I was concerned about 10^-36
> > errors but no more!
>
> There is no need to stick out; everybody already knew what you are.

Thanks for the interesting link.


Bret Cahill

On Sep 29, 2:23&#4294967295;pm, John <sampson...@gmail.com> wrote:
> On Sep 29, 1:16&#4294967295;pm, Bret Cahill <BretCah...@peoplepc.com> wrote:
>
> > What is the highest frequency that could be handled with discrete
> > fourier transform software?
...
>
> Fs/2

or maybe a millismidgen less than that.

r b-j