Deczky all-pass filter design Matlab

On 22 Mar 2009 10:44:28 GMT, Erik de Castro Lopo
<nospam@mega-nerd.com> wrote:

>dbd wrote:
>
>> On Feb 6, 4:42 pm, Erik de Castro Lopo <nos...@mega-nerd.com> wrote:
>>> Rune Allnor wrote:
>>> > I don't know the filter other than what I found here:
>>>
>>> >http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel2/640/5939/002305...
>>>
>>> Unfortunately, thats locked behind the IEEE knowledge firewall.
>>>
>>> Erik
>>
>> The IEEE knowledge firewall is a lot like the gates at Disneyland. A
>> lot of people want to go there. Some of what's there may be personally
>> amusing to you. It's easy to find out what's there to see if it is
>> interesting to you. You need to pay for your ticket to get in. Annual
>> passes are available.
>
>I have voiced my complaints about IEEE and the pricing structure of
>its papers before:
>
>http://groups.google.com/group/comp.dsp/browse_thread/thread/cdae9b1fb99e3e0a/6cdb6db5f834e8c8
>
>I would be happy  to pay a fair price, but as far as I  am concerned
>that is not what the IEEE is offering. Quite frankly I see the IEEE as
>being a fine example of an orgnisation without any market place
>competition  and little accountability.
>
>Here are a couple of my objections to the IEEE. I would appeciate it if
>people could point out any errors I make.
>
>  - As I understand it authors who have their papers accepted into IEEE
>    journals pay money to help with the publishing.
>
> -  Paper reviewers and editors provide their services for free.

That's correct for reviewers, I don't know about editors.   Likewise
session chairs at IEEE conferences are unpaid.

> - Univesitiy  libraries around the world pay extortionate fees to carry
>   IEEE journals.
>
>Now I ask, where does all  this money go and wouldn't it be relatively
>easy to provide the same (or hopefully even better) quality  of
>information for a fraction of the cost?

I suspect so.   

>Given  the above, why does a IEEE paper downloaded over the internet
>cost US$13 for IEEE members and US$25 for non-members?

I don't think it's inappropriate to incentivize membership, so having
a lower price for members seems reasonable to me.   The amounts,
naturally, are aguable.

Understand also that even if you're an IEEE member, AND a member of
the society that published the paper, you still have to pay the same
member price to download a paper.   In other words, there's not much
incentive to join a society unless you want their magazine. Sometimes
(as in the case of the Signal Processing Society, IMHO), that may be
worth it, but you don't get access to the transactions without a
subscription specifically to the transactions, even if you're a member
of the society.

And you can't get a subscription to the transactions without being a
member of the society, so access to the transactions winds up costing
a bunch.

In addition to your points it seems to me (and others here who've
voiced similar complaints) that the paper screening/review process is
broken as well.  It's an academic circle jerk as far as I can tell.
Many good papers do make it through, although IMHO they're not the
majority, and a lot of good papers get screened out for what seem to
me to be really questionable reasons. 

Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.ericjacobsen.org

Blog: http://www.dsprelated.com/blogs-1/hf/Eric_Jacobsen.php

On 22 Mar, 18:01, dbd <d...@ieee.org> wrote:
> On Mar 22, 3:44 am, Erik de Castro Lopo <nos...@mega-nerd.com> wrote:> dbd wrote:

> .> Here are a couple of my objections to the IEEE. I would appeciate
> it if
> .> people could point out any errors I make.
>
> > &#4294967295; - As I understand it authors who have their papers accepted into IEEE
> > &#4294967295; &#4294967295; journals pay money to help with the publishing.
>
> > &#4294967295;- &#4294967295;Paper reviewers and editors provide their services for free.
>
> .> &#4294967295;- Univesitiy &#4294967295;libraries around the world pay extortionate fees to
> carry
> .> &#4294967295; &#4294967295;IEEE journals.
> Where is you data?

I didn't have access to the actual numbers, but I when
thsese questions were up where I worked ten years ago,
the lunch-table talk mentioned $100,000 per year for
an organization. I don't know if that was unlimited
access to all of IEEE (it might have been) or if it
was a subset.

The problem occurs if you pay these amounts of $$$
for maybe 5 - 10 people to get access to what they
need. It's very easy to shut down such a subscription.

> What do they pay and what should it cost to get
> unlimited access for large numbers of people?

This is the hard part, and to answer this, one first
needs to establish why IEEE is there in the first place"

- To serve some purpose?
  - If so what purpose?
  - Whos is to benefit?
- To make money?
  - If so, who are to benefit?
- Something else?
  - If so, what?

Naively, I assume that IEEE is exists to serve as a
forum for professionals (mainly members, but not
exclusively) to keep up with the latest developments,
and to exchange ideas.

For this to work, the objective ought to be that one
facilitates journal archive services (like IEEExplore)
to get access to the journals. There is no point in
doing things for free, so *some* fee would have to be
paid. However, infrastructure and operation costs are
more or less constant once the system is up and running,
so the objective is not to maximise profit per
transaction, but to price the service such that people
actually pay to download the articles.

With present transaction volumes, one might have to
charge $10 to cover operations costs. Price the same
transactions to, say $2 per article, and you might
just find that you get 50 times the number of
transactions.

> > Now I ask, where does all &#4294967295;this money go and wouldn't it be relatively
> > easy to provide the same (or hopefully even better) quality &#4294967295;of
> > information for a fraction of the cost?
>
> What do you think it would cost to acquire several hundred thousand
> articles and maintain search function and adequate secure access
> bandwidth to them in a reliable (in the multi-decade sense)? What
> would the size of the staff be?

IEEE already does this, as part of day-to-day operations.
Every new article is available in the electronic archive.
Substantial amounts of the historical archives are already
scanned and available in electronic form.

> If it's cheap, here is a business
> opportunity for you and everyone else. Where are the takers?

Wrong. The 'business opportunity' is only there for
IEEE, since they run and administer the archives.
The question is how to make those archives available
to the user community.

> Disneyland costs $59 for a day and a few hundred dollars for an annual
> pass. For a few hundred dollars you can get annual unlimited personal
> access to a significant portion of the IEEE content.

Let's say 'a few hundred $$$' is $750 and one article
costs $15 to download. That's 50 articles per year.
How many people download that many articles? In my
experience, not very many. Those who do are usually
grad students, who have access through univeristies.
How many of those 50 articles turn out to be useful?
In my experience, maybe 1 in 10 turn out to be useful,
the rest are peripherical at best. The problem is that
one never knows what any given article turns out to
be beforehand.

Anyway, with this pricing policy one can expect to
pay $150 between each useful article.

Why would anybody do that? I have no idea, and it seems
few other do either. The corollary is obvious: What is
the long-term purpose of IEEE if no one sees any point
in getting the articles they publish? Why would
anyone want to become members? Get involved?

> If you think there is a point here,

I'm pretty sure I know what Erik's and Eric's points
are. I'd be interested in hearing your arguments that
everything is fine as is - if that is indeed what you
mean? I can't find any clear opinions stated in
your post.

Rune

On Mar 22, 12:57 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
...
> I didn't have access to the actual numbers, but I when
> thsese questions were up where I worked ten years ago,
> the lunch-table talk mentioned $100,000 per year for
> an organization. I don't know if that was unlimited
> access to all of IEEE (it might have been) or if it
> was a subset.
>
> The problem occurs if you pay these amounts of $$$
> for maybe 5 - 10 people to get access to what they
> need. It's very easy to shut down such a subscription.
>
Under the current IEEE cost of memberships there are alternative
routes to provide coverage for 5-10 much cheaper.

Is $100,000 out of line for thousands of undergrads, hundreds of grad
students, and hundreds of faculty? Eric's claims were 'extortionate'
for 'Univesitiy  libraries'.

> ...
> Let's say 'a few hundred $$$' is $750 and one article
> costs $15 to download. That's 50 articles per year.
> How many people download that many articles? In my
> experience, not very many. Those who do are usually
> grad students, who have access through univeristies.
> How many of those 50 articles turn out to be useful?
> In my experience, maybe 1 in 10 turn out to be useful,
> the rest are peripherical at best. The problem is that
> one never knows what any given article turns out to
> be beforehand.
>
> Anyway, with this pricing policy one can expect to
> pay $150 between each useful article.
>
Real numbers:
IEEE Membership $170
Society memberships $20-50 each
Publication costs after society membership $5-50 per publication

Each publication subscription includes unlimited access to the
archives of that publication. no per article fee.

Your $750 can buy unlimited access to 8 or more journals.

The IEEE member per article fee is only for unsubscribed publications.

There is also a fixed cost per month alternative subscription option
for a set number of articles per month selected from any of the
publications. Some societies with a large number of publications also
have a group publication subscription rate.
>
> I'm pretty sure I know what Erik's and Eric's points
> are. I'd be interested in hearing your arguments that
> everything is fine as is - if that is indeed what you
> mean? I can't find any clear opinions stated in
> your post.

I'd like information to be cheaper.

At the moment, I think that the IEEE non-member article rates are
comparable with the rates of commercial journals for non-subscribes in
academic topics.

IEEE member rates for large scale access for frequent users are priced
much lower than the numbers presented by those who have complained
without checking the facts. Since unlimited access is available, what
the price per article is depends on your frequency of usage. IEEE
pricing is certainly in favor of frequent users. That is acceptable
practice in most markets.

>
> Rune

This isn't an issue limited to comp.dsp and the IEEE. The development
of the web has altered operations of the ACM, JASA, and commercial
engineering publishers. (And the expectations of internet users who
have had a taste of some things for free for a while.) Engineering is
only one of the myriad subject areas facing these changes. I think
that complainers who don't expect to be dismissed as whiners should
get their facts straight and if they want a change, should suggest
what they think is a feasible alternative, if only as an attempt at an
existence proof that there is a feasible alternative. Don't we expect
everyone to do their homework at comp.dsp?

Dale B. Dalrymple

dbd wrote:

> What do you think it would actually cost? If you think there is a
> point here, what are -your- numbers?

I can give a  lower bound.

dbd wrote:

> What do you think it would actually cost? If you think there is a
> point here, what are -your- numbers?

I'm  an engineer, not an accountant, but if my accountant told me I had 
to charge US$13 to deliver a PDF over the internet I'd be finding a
new accountant.

Erik
-- 
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/erikd/Blog/

dbd wrote:

> What do you think it would actually cost? If you think there is a
> point here, what are -your- numbers?

I'm  an engineer, not an accountant, but if my accountant told me I had 
to charge US$13 to deliver a PDF over the internet I'd be finding a
new accountant.

Erik
-- 
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/erikd/Blog/

dbd wrote:

> What do you think it would actually cost? If you think there is a
> point here, what are -your- numbers?

I'm  an engineer, not an accountant, but if my accountant told me I had 
to charge US$13 to deliver a PDF over the internet I'd be finding a
new accountant.

Erik
-- 
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/erikd/Blog/