> On 22 Nov, 12:31, "John E. Hadstate" <jh113...@hotmail.com> wrote:
>> "Rune Allnor" <all...@tele.ntnu.no> wrote in message
>>
>> news:d2c59763-5c39-4450-8528-b1b316ea50b2@v4g2000hsf.googlegroups.com...
>>
>>
>>
>>> If you think I am wrong when I say "one never samples the
>>> spectrum"
>>> could you please come up with an example on where one
>>> samples
>>> the spectrum directly? Without deriving it by sampling
>>> time domain
>>> data first and then compute the spectrum?
>> Look up filter design by the frequency-sampling technique.
>
> OK, my fault because I let myself slip too far towards Dale's
> position where "Human language conforms to practice not
> deduction". I wasn't precise enough as I -- as is a not
> totally unheard of habit when starting out in one particular
> context -- was staying with the established context of the
> current thread, namely the analyzis of measured data, not
> the design of filters.
>
> By the way, I am still waiting for your enlightenement on
> interpolation schemes which do not intersect the given
> points.
I pointed out one which discards half of the original samples. Did you
miss it? We can discard a greater proportion by changing the numbers.
Jerry
--
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������
Reply by Rune Allnor●November 22, 20072007-11-22
On 22 Nov, 13:24, "John E. Hadstate" <jh113...@hotmail.com> wrote:
> "Rune Allnor" <all...@tele.ntnu.no> wrote in message
>
> news:dd932d49-6727-49f9-9d4a-9e9ee5fa77ec@v4g2000hsf.googlegroups.com...
>
> >> So far as I'm
> >> concerned, you are about a half-sample short of a
> >> discrete
> >> trip to my killfile.
>
> > Please, put me there. If that can prevent me from
> > engaging in battles with someone who knows "several
> > methods" which contradict my position but can't name
> > a single one, then all the better.
>
> Not "can't," moron, "won't".
And *you* are accusing *me* for insulting other people's
intelligence...!?
Rune
Reply by John E. Hadstate●November 22, 20072007-11-22
"Rune Allnor" <allnor@tele.ntnu.no> wrote in message
news:dd932d49-6727-49f9-9d4a-9e9ee5fa77ec@v4g2000hsf.googlegroups.com...
>> So far as I'm
>> concerned, you are about a half-sample short of a
>> discrete
>> trip to my killfile.
>
> Please, put me there. If that can prevent me from
> engaging in battles with someone who knows "several
> methods" which contradict my position but can't name
> a single one, then all the better.
Not "can't," moron, "won't".
Anyway, your wish is my command: *plonk*
>
> Rune
Reply by Rune Allnor●November 22, 20072007-11-22
On 22 Nov, 12:50, "John E. Hadstate" <jh113...@hotmail.com> wrote:
> "Rune Allnor" <all...@tele.ntnu.no> wrote in message
>
> news:e9cc2ff2-2bde-4f9d-bec3-882661f0466c@w40g2000hsb.googlegroups.com...
>
> > If you have read - and contemplated - my other posts
> > in this thread you will find that I do not think any
> > discrete
> > sequence, strictly speaking, can be defined as an
> > interpolation.
>
> You're pretty far out on a limb there, Chief. Everyone else
> in the world understands "interpolation" to mean "finding
> points on an interval between given endpoints subject to the
> constraints of a particular model (often linear)."
In my book (and in my maths books as well) "interpolation"
means "finding continuous functions which intersect a given
set of points."
> Everything about this screams, "Discrete sequence."
One of the most popular methods fro interpolation is based on
splines, which are continuous functions which intersect the given
points. How does spline interpolation fit into this?
> What
> part of "finding points" don't you understand?
How they fit in with a concept which aims to find
continuous functions.
> > My personal opinion is that I can go along with the term
> > as
> > a day-to-day sloppiness, provided the original spectrum
> > coefficients are preserved. Once those are gone, the term
> > "interpolation" has lost what little meaning was left.
>
> It looks to me like you long ago backed yourself into an
> untenable position and are now engaged in a desperate battle
I am just trying to point out the holes in your and others
arguments. Dale B Dalrymple turned personal on me,
I made him aware of this fact and requested him to address
my arguments, not my person. Just read the posts.
> to save face by abusing and insulting the intelligence of
> anyone who will pay attention to you.
I have done nothing but to point out that there is a fundamental
difference between discrete points and continuous functions.
I have based all my arguments on this simple fact. If you and
Dale have a problem with this, and me pointing it out (no pun
intended), then that's your problem.
It is up to you to draw any conclusions about the intelligence
of the participants. Which you, apparently, already have done.
> So far as I'm
> concerned, you are about a half-sample short of a discrete
> trip to my killfile.
Please, put me there. If that can prevent me from
engaging in battles with someone who knows "several
methods" which contradict my position but can't name
a single one, then all the better.
Rune
Reply by John E. Hadstate●November 22, 20072007-11-22
"Rune Allnor" <allnor@tele.ntnu.no> wrote in message
news:8bb57c27-76bb-42f2-853a-b8793491c3d1@s8g2000prg.googlegroups.com...
>
> By the way, I am still waiting for your enlightenement on
> interpolation schemes which do not intersect the given
> points.
>
By not replying to your previous trolling, I was trying to
help you retain some vestige of dignity.
Look, Rune, I know you're not stupid so what in Hell is the
matter with you? It can't be too much Thanksgiving because
that's an American phenomenon, and it's too early to start
celebrating Christmas or New Year's Eve.
Reply by John E. Hadstate●November 22, 20072007-11-22
"Rune Allnor" <allnor@tele.ntnu.no> wrote in message
news:e9cc2ff2-2bde-4f9d-bec3-882661f0466c@w40g2000hsb.googlegroups.com...
> If you have read - and contemplated - my other posts
> in this thread you will find that I do not think any
> discrete
> sequence, strictly speaking, can be defined as an
> interpolation.
You're pretty far out on a limb there, Chief. Everyone else
in the world understands "interpolation" to mean "finding
points on an interval between given endpoints subject to the
constraints of a particular model (often linear)."
Everything about this screams, "Discrete sequence." What
part of "finding points" don't you understand?
> My personal opinion is that I can go along with the term
> as
> a day-to-day sloppiness, provided the original spectrum
> coefficients are preserved. Once those are gone, the term
> "interpolation" has lost what little meaning was left.
It looks to me like you long ago backed yourself into an
untenable position and are now engaged in a desperate battle
to save face by abusing and insulting the intelligence of
anyone who will pay attention to you. So far as I'm
concerned, you are about a half-sample short of a discrete
trip to my killfile.
>
> Rune
Reply by Rune Allnor●November 22, 20072007-11-22
On 22 Nov, 12:31, "John E. Hadstate" <jh113...@hotmail.com> wrote:
> "Rune Allnor" <all...@tele.ntnu.no> wrote in message
>
> news:d2c59763-5c39-4450-8528-b1b316ea50b2@v4g2000hsf.googlegroups.com...
>
>
>
> > If you think I am wrong when I say "one never samples the
> > spectrum"
> > could you please come up with an example on where one
> > samples
> > the spectrum directly? Without deriving it by sampling
> > time domain
> > data first and then compute the spectrum?
>
> Look up filter design by the frequency-sampling technique.
OK, my fault because I let myself slip too far towards Dale's
position where "Human language conforms to practice not
deduction". I wasn't precise enough as I -- as is a not
totally unheard of habit when starting out in one particular
context -- was staying with the established context of the
current thread, namely the analyzis of measured data, not
the design of filters.
By the way, I am still waiting for your enlightenement on
interpolation schemes which do not intersect the given
points.
Rune
Reply by John E. Hadstate●November 22, 20072007-11-22
"Rune Allnor" <allnor@tele.ntnu.no> wrote in message
news:d2c59763-5c39-4450-8528-b1b316ea50b2@v4g2000hsf.googlegroups.com...
>
> If you think I am wrong when I say "one never samples the
> spectrum"
> could you please come up with an example on where one
> samples
> the spectrum directly? Without deriving it by sampling
> time domain
> data first and then compute the spectrum?
>
Look up filter design by the frequency-sampling technique.
It has a close relative in filter design by the windowing
technique. In both cases, one starts out with a
specification in the frequency domain, samples it, and
produces an impulse response in the time domain that, when
used as a filter, produces an approximation to the frequency
domain specification.
Reply by Rune Allnor●November 22, 20072007-11-22
On 22 Nov, 09:49, dbd <d...@ieee.org> wrote:
> On Nov 21, 9:24 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
>
> dbd: > To respond to this in a manner consistent with your tone in
> this> > > > discussion, I would have to point out that we never 'actually sample'
> > > > > a signal in the time domain. Sampling is a process that modulates a
> > > > > continuous signal with a Dirac comb. Since we can't actually generate
> > > > > a Dirac comb we can never actually sample
>
> > If you think I am wrong when I say "one never samples the spectrum"
> > could you please come up with an example on where one samples
> > the spectrum directly? Without deriving it by sampling time domain
> > data first and then compute the spectrum?
>
> I never suggested your statement that you quote was wrong. I did point
> out that if we apply strict formal definitions universally, we can't
> honestly call what is done to actual signals in the time domain as
> meeting a formal definition of sampling. I then suggest that formality
> has little to do with usability. Please don't make up disagreements
> for me.
So why are people spending so much time with formal defintitions?
> > Maths and engineering is not and has never been a democracy.
>
> We agree on that.
> But choices in the use of language and vocabulary are democratic.
Nope. What counts are explicit, unambiguous definitions. That
requirement is the main reason why most people find texts on
maths and engineering all but unreadable. Every word counts.
> > it is about making useful deductions.
>
> Human language conforms to practice not deduction.
This thread started from a very explicit question about maths.
My point was that the question was more or less invalid
because it is phrased in terms of a "practical" language
which is at odds with the precise definitions of the maths.
> > > considering the wording of the original post, it seemed an abusive
> > > distinction to lay on the OP who was already struggling to compose a
> > > coherent question in DSP-English.
>
> > Nope. There is nothing at all in the OPs post to indicate he or she
> > is a non-native English speaker.
>
> Nor in my remark. You are seeing things that are not there.
You say you can see that the OP "struggles" to formulate a
question. I can see no such signs.
> The issue is with multiple dialects of DSP-English. The OP asked a
> formal question but used a vocabulary term that you denigrated as
> useless due to lack of formal definition.
Yes.
> > > If Rune had made the same remark to
> > > you or many others who frequent comp.dsp, I wouldn't have considered
> > > giving the advice I gave the OP.
>
> > What advice?
>
> My advice to the OP not use the term interpolation around you. I
> thought it would lead you to lengthy diatribes that would distract the
> thread for any response to the OP's question, but of course, I was
> already too late.
I can't answer for anybody else, but my experience is that technical
language serves a purpose. In other words, terms and definitions
are there because they are useful. Unfortunately, that's at odds
with the same terms having been derived either from every-day
language or being "borrowed" from other contexts. Once that
happens, confusion is imminent.
> As to your points, you suggest a formal position but fail to apply the
> same formality to your own arguments.
If you actually read my arguments instead of arguing against me on
a matter of principle, you might want to modify that opinion. I at
least make the effort to be consistent. I don't claim to be
successful
100% of the time, but I make the effort 100% of the time.
> This inconsistency is one of the
> reasons people find to disagree with you. And we only argue with the
> points we think you fail to make, not with the many that you do make.
Then why not just ask for me to elaborate? I've asked others for
elaborations on at least three occations in this thread alone. On one
occation I got a reply which helped me see a flaw in my own argument.
You could try that, too. You might actually learn something in
the process.
Rune
Reply by dbd●November 22, 20072007-11-22
On Nov 21, 9:24 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
dbd: > To respond to this in a manner consistent with your tone in
this
> > > > discussion, I would have to point out that we never 'actually sample'
> > > > a signal in the time domain. Sampling is a process that modulates a
> > > > continuous signal with a Dirac comb. Since we can't actually generate
> > > > a Dirac comb we can never actually sample
>
> If you think I am wrong when I say "one never samples the spectrum"
> could you please come up with an example on where one samples
> the spectrum directly? Without deriving it by sampling time domain
> data first and then compute the spectrum?
>
I never suggested your statement that you quote was wrong. I did point
out that if we apply strict formal definitions universally, we can't
honestly call what is done to actual signals in the time domain as
meeting a formal definition of sampling. I then suggest that formality
has little to do with usability. Please don't make up disagreements
for me.
> ..
> > > > Perhaps this is a culture or language derived bias. After at least 18
> > > > posts by at least 7 authors you seem to be the only one objecting to
> > > > the use of the term interpolation in the discrete frequency domain.
> > > > That seems to validate my initial advice to the original poster.
>
> Maths and engineering is not and has never been a democracy.
We agree on that.
But choices in the use of language and vocabulary are democratic.
> it is about making useful deductions.
Human language conforms to practice not deduction.
> > considering the wording of the original post, it seemed an abusive
> > distinction to lay on the OP who was already struggling to compose a
> > coherent question in DSP-English.
>
> Nope. There is nothing at all in the OPs post to indicate he or she
> is a non-native English speaker.
Nor in my remark. You are seeing things that are not there.
The issue is with multiple dialects of DSP-English. The OP asked a
formal question but used a vocabulary term that you denigrated as
useless due to lack of formal definition.
> The question is clear: How to
> show by maths that zero-padding in time domain amounts to
> interpolation in frequency domain. That calls by default for a
> rigorous definition for how (or if at all) the term "interpolation"
> can be used to generate a discrete sequence from another
> discrete sequence.
>
> > If Rune had made the same remark to
> > you or many others who frequent comp.dsp, I wouldn't have considered
> > giving the advice I gave the OP.
>
> What advice?
My advice to the OP not use the term interpolation around you. I
thought it would lead you to lengthy diatribes that would distract the
thread for any response to the OP's question, but of course, I was
already too late.
> I am used to the fact that people argue against
> whatever points I make on the sole basis of my signature.
I fail to comprehend any problem with your signature. What do you see
there that I have missed?
As to your points, you suggest a formal position but fail to apply the
same formality to your own arguments. This inconsistency is one of the
reasons people find to disagree with you. And we only argue with the
points we think you fail to make, not with the many that you do make.
Get over it.