Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.
|
|
|
|
Re: Iterative Match Filterings
Posted:
Apr 11, 2012 11:37 AM
|
|
On Wed, 11 Apr 2012 07:53:43 -0700 (PDT), Bret Cahill
<BretCahill@peoplepc.com> wrote:
>> >>> > Depends on what you want from the matched filter
>>
>> >>> I want to do the classical thing.
>>
>> >>But supposing matched filtering could be adapted for another use where
>> >>it was better than another popular filter, say, a Wiener filter?
>>
>> > Supposing you can drive a screw into a piece of wood faster with a
>> > hammer than a screwdriver, =A0does that mean it's a better method?
>>
>> Eric, have you ever considered philosophy as a career? :)
>
>He dodged the issue so he might work out as a GOP political
>"philosopher."
>
>Match filtering, or, if the additional deconvolution step is new,
>"Matched Filtering/Magnitude Recovery" utilizes all of the phase angle
>info as well as the PSD info and is, therefore, better than any other
>filter as far as recovering the magnitude of the original signal.
"Matched" filtering can use all of the phase info, too, and very often
does. You still seem to have a lot of misconceptions about what
matched filtering is, what it does, and how it does it.
What you've been describing is not matched filtering.
>Matched filtering/signal recovery also seems to be a better way to
>recover square waves out of noise than screwing around with triangle
>shaped convolutions.
The triangle-shaped convolution output *IS* the result of matched
filtering against a square-wave input. If you get anything else out,
it wasn't a matched filtering process. Your whole notion of
"recovering the original waveform" is NOT a matched filtering process
or objective in any usual sense. This has been pointed out to you
multiple times.
You're the one that cited Charan's tutorial page on matched filtering.
You should take the time to actually understand it.
Eric Jacobsen
Anchor Hill Communications
www.anchorhill.com
|
|
|
|
