Prev: Our customer websites are modern, fresh and accessible.
Next: $$jordan coach bag ed tshirt with amazing price
From: Steve Pope on 31 May 2010 02:48
Zhi.Shen <zhi.m.shen(a)gmail.com> writes,

>Hi, Steve

>In some DSP books, it always be described to be used in Nonlinear phase FIR
>filter,
>why few of them are about linear phase FIR filter?

>"Steve Pope" <spope33(a)speedymail.org>

>> Zhi.Shen <zhi.m.shen(a)gmail.com> wrote:

>>>I just read some materials about Lattice architecture.
>>>What its benefit is,compared to direct form?

>> It doubles the number of multiply/adds but often the coefficients
>> can be lower precision.
>>
>> Also, unrelated to your question, it can lead to a more stable
>> IIR filter.

>> It may not be appropriate for all filters, but if you're
>> having problems with coefficient precision it may be worth
>> checking into.

Excellent question.

I would tend to say this is because the types of real-world
systems for which an all-pole lattice filter naturally
falls out as an appropriate model (such as, the human
vocal tract) do not have linear-phase inverse filter functions.

That's about all the insight I can add here.


Steve
From: Zhi.Shen on 31 May 2010 03:41
Thank you, Steve
I have just found a paper about linear lattice FIR filter:

Title: Linear phase FIR-filter in lattice structure
Author: Schwarz, K.
Source: Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International
Symposium on, 1993, 347-350

But I have no idea why so few papers about this issue.


"Steve Pope" <spope33(a)speedymail.org> д����Ϣ����:htvm4k$vi3$1(a)blue.rahul.net...
> Zhi.Shen <zhi.m.shen(a)gmail.com> writes,
>
>>Hi, Steve
>
>>In some DSP books, it always be described to be used in Nonlinear phase
>>FIR
>>filter,
>>why few of them are about linear phase FIR filter?
>
>>"Steve Pope" <spope33(a)speedymail.org>
>
>>> Zhi.Shen <zhi.m.shen(a)gmail.com> wrote:
>
>>>>I just read some materials about Lattice architecture.
>>>>What its benefit is,compared to direct form?
>
>>> It doubles the number of multiply/adds but often the coefficients
>>> can be lower precision.
>>>
>>> Also, unrelated to your question, it can lead to a more stable
>>> IIR filter.
>
>>> It may not be appropriate for all filters, but if you're
>>> having problems with coefficient precision it may be worth
>>> checking into.
>
> Excellent question.
>
> I would tend to say this is because the types of real-world
> systems for which an all-pole lattice filter naturally
> falls out as an appropriate model (such as, the human
> vocal tract) do not have linear-phase inverse filter functions.
>
> That's about all the insight I can add here.
>
>
> Steve


From: dbd on 31 May 2010 14:24
On May 31, 12:41 am, "Zhi.Shen" <zhi.m.s...(a)gmail.com> wrote:
> ...
>   I have just found a paper about linear lattice FIR filter:
>
> Title: Linear phase FIR-filter in lattice structure
> Author: Schwarz, K.
> Source: Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International
> Symposium on, 1993, 347-350
>
>    But I have no idea why so few papers about this issue.
> ...

Google on:
fir coefficient sensitivity
gives 594,000 hits

Google on:
linear phase fir filter lattice
gives 19,000 hits

Even if there is only 1 paper per 1000 Google hits, there are plenty
of papers. The problem is to specify a narrower view of what you are
seeking,and to be willing to scan the first hundred hits rather than
just the first page.

Dale B. Dalrymple
From: Robert Orban on 16 Jun 2010 22:30
In article <d512d56c-f38d-40b0-948b-b95fd7cce808(a)32g2000prq.googlegroups.com>,
allnor(a)tele.ntnu.no says...
>
>
>On 30 Mai, 10:47, "Shen Zhi" <markk...(a)hotmail.com> wrote:
>> Hi, friends!
>>
>> � Is there any good method for "Polynomial Decomposition" being used in
>> Filter Design?
>> For example, if I use Park-McClellan algorithm designing a 10-order FIR
>> filter,then want to decompose the polynomial into five 2-order filters, and
>> keep the overall magnitude response.
>> � Does anyone has good suggestion or known some papers about this issue,
>> please tell me.
>
>Any reason, other than numerical accuracy issues [*], why you
>can't try polynomial rooting?

There are ways to get the roots of the very high-order polynomials (orders
higher than 100) applicable to this specific problem. This might help:

Stathaki, T. Fotinopoulos, I., "Equiripple minimum phase FIR filter design
from linear phase systems using root moments"

http://www.ee.bilkent.edu.tr/~signal/Nsip99/papers/20.pdf

See also

http://cnx.org/content/m15573/latest/

for the Lindsey-Fox algorithm for factoring polynomials

First  |  Prev  | 
Pages: 1 2 3 4 5
Prev: Our customer websites are modern, fresh and accessible.
Next: $$jordan coach bag ed tshirt with amazing price