[music-dsp] applying an FIR filter
hplus at mindcontrol.org
Sun Jan 27 12:53:14 EST 2002
It's the last sample that got filtered, because any other interpretation
would mean that you knew about samples happening "in the future". This
means that the latency of your filter is equal to your number of taps
(minus one, perhaps, depending on your view of it). Note that the center
of mass of your FIR is likely to sit in the middle, so it's useful to
think of the "signal delay" as half of the FIR size (which is different
from the "processing lantency").
The corollary is that you have to precede your input signal with
(NTaps-1) zeros if you want to be able to output anything from the first
sample on. You also need to trail by (NTaps-1) samples to make sure that
all ringing is emitted at the end of the signal.
> -----Original Message-----
> From: owner-music-dsp at shoko.calarts.edu
> [mailto:owner-music-dsp at shoko.calarts.edu]On Behalf Of Marc Poirier
> Sent: Saturday, January 26, 2002 2:37 PM
> To: Music DSP list
> Subject: [music-dsp] applying an FIR filter
> I have a question about applying an FIR filter to a digital signal.
> Okay, I have an FIR filter set up with an odd number of taps (let's
> say 31) & the coefficients are symmetrical. Now I apply that filter.
> I take 31 consecutive samples from my signal & sum them, scaling them
> with their corresponding coefficients & all that. Now once I've done
> that, which sample is considered to be the one that got filtered? Is
> it the middle sample? Or the last one? Or the first one?
> I hope this question makes sense, thanks in advance for any help...
> [ Destroy FX - http://www.smartelectronix.com/~destroyfx ]
> dupswapdrop -- the music-dsp mailing list and website: subscription info,
> FAQ, source code archive, list archive, book reviews, dsp links
dupswapdrop -- the music-dsp mailing list and website: subscription info,
FAQ, source code archive, list archive, book reviews, dsp links
More information about the music-dsp