From: John Suters (johnds_at_[hidden])
Date: 2001-11-01 14:25:34
> No, my understanding is you choose the behaviour. You can use
> FFTW_MEASURE and it will actually search for the fastest way to
> perform your fft, but you can instead use FFTW_ESTIMATE and it will
> not introduce this overhead.
Fair enough then. I'll have a bit of a closer look.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk