When I post here after doing many more things during the day, it looks like I take out jj scientifically, you as a Taliban Chicago barbarian and BTox as an ignorant New Jersey cheerleader who is a product of affirmative action without merit.WildCat said:Looks like jj took out the Eurotrash.
It's ironic that in lon's attempt to establish his bigoted stereotyped view of Americans, he reinforced the French stereotype of being arrogant contemptuous pricks.
Technical and Technical IP questions for Ion
Ion said:
Your theft of credit from Cooley and Tukey has been demonstrated.
Your inane, simple mistake regarding "unstable" FIR filters has been dismembered. That mistake alone should be embarrassing enough to completely demonstrate your ignorance.
Your claim that Goertzel was an FFT was shown to be false. You knew that from the start, and were given a host of times to retract that statement, but attempted, instead, to try to deflect your illicit FFT claim to a DFT claim.
Your claim that stability of FIR filters was determined by Laplace was false. (Well, they can't be unstable. Your claim that anything could determine instability of something that can't be unstable has an obvious logical hole, but you won't admit that, and try to sneak off into claiming that by using a different, IIR implimentation, you could contrive instability. Guess what? We all know that IIR's can be unstable. You said FIR, not IIR. You've even attempted to defend that laughable absurdity.
Your claim that you have "dismembered me scientifically" is professional libel.
You shall now retract the claims above, filly and without qualification.
You shall apologize for your attempt to cause harm to my reputation.
You have spread what appears to be deliberate misinformation about academic credit, mathematical content, and practical application of common algorithms in your field.
Apologize now.
There is nothing more technical to say to you. You stand fully refuted on every point in dispute. The only points you hold are irrelevant to the discussion, introduced by you as intentional deception, and were not in dispute from the beginning.
Now come clean. It will hurt less here than it might elsewhere.
You've lost entirely to a fat old white boy from the hills. According to you, that's the worst of america. If you're right, then you amount to NOTHING, because the worst of america has destroyed you in technical debate, although I admit it wasn't much challenge after you ◊◊◊◊ out that whopper about unstable FIR filters.
I repeat, it will hurt less if you admit it here, and now.
Regarding:
Including the fact that an algorithm for computing D.F.T. is Goertzel.
With less than N log N.
No matter your ignorant semantics of "...partial...:, etc.
Mind you, considering BTox and WildCat, you show promise.
But not good enough...
I advise you jj to go back to school -preferably in Europe- and learn it.Ion said:
Including the fact that an algorithm for computing D.F.T. is Goertzel.
With less than N log N.
No matter your ignorant semantics of "...partial...:, etc.
Mind you, considering BTox and WildCat, you show promise.
But not good enough...
Ion said:Regarding:
I advise you jj to go back to school -preferably in Europe- and learn it.
I require your admission that I haven't said anything incorrect, and that you stalled, stalled, and stalled some more, and have yet to fail to admit that Goertzel is not an FFT.
You have repeatedly misstated what the Goertzel algorithm does. I have cited a helpful document to you, but you have yet to acknowledge your utterly wrong positions.. As you continue to repeat them, they become deceptions merely because you have failed to do due dilligence, if nothing else.
Including the fact that an algorithm for computing D.F.T. is Goertzel.
One instance of Goertzel calculates one line of a DFT or FFT. We agree on that. One instance of a Goertzel algorithm does no more or less than that.
Unless we use it as an N^2 algorithm it will not calculate a whole DFT or FFT.
in http://www.ece.utexas.edu/~mason/codesign/pass/embedded.html
it says very clearly "Goertzel is not considered a Fast Fourier Transform (FFT) because it is order n2, not order nlog2. "
With less than N log N.
Your statement is false. The quote above clearly shows that. Goertzel is, as the web reference, or basic understanding shows so clearly, less efficient unless you only want a FEW output lines.
You have admitted this, yet you continue to deceive and say "less than N log N". You know, I know, we all know that it can't do anything close to N log N for anything more than a few lines, let alone an actual transform.
Why do you continue to make this lie, sir? You are fully found wanting by every reference available.
No matter your ignorant semantics of "...partial...:, etc.
Blah, blah, blah.
You have repeatedly attempted to pass off an algorithm that calculates one line of a transform as the whole transform.
You are going to have a whole lot of trouble figuring out a spectrum from one line, you know. Need we go into Parsival's Theorem, etc, to point out the misuse you are making of the word "transform"? Please explain to me how your 8 lines are 1:1 and onto. Oops, they aren't. Ergo, they aren't a transform. You knew that all the way from the beginning, too.
You know your statement is a lie, sir, and yet you continue. Again, let us cite a handy internet reference:
http://www.cambr.uidaho.edu/papers/fft_on_rdpp_slides.pdf
Where, in the process of explaining a case where Goertzel is USEFUL, clearly states that Goertzel is O (n^2) and the FFT is O(n long n). Goertzel, while it has its applications, is, as this and every other reference we've cited at you shows, is O(n^2) for an actual transform. This paper is even, due to a particular processor design, espousing the use of Goertzel, and it says clearly that it is inefficient in terms of calculation.
You use it in DTMF as a set of filters in a filterbank. Fine, but don't claim it's any kind of DFT, you've just taken some DFT basis vectors and used them as a matched filter. You aren't even doing a transform, just running a filter bank. It's a good filter bank, but that's all it is.
As it clearly points out in http://www.mathworks.com/access/helpdesk/help/toolbox/signal/goertzel.shtml again you are wrong in asserting that it's faster for an actual transform. Quoting:
---
Note fft computes all DFT values at all indices, while goertzel computes DFT values at a specified subset of indices (i.e., a portion of the signal's frequency range). If less than log2(N) points are required, goertzel is more efficient than the Fast Fourier Transform (fft).
----
Once again, it states clear that if you need a transform, not just some bits and pieces (useful though they may be), Goertzel is not faster for any realistic size of transform. It may be useful on a painfully specialized processor, but again that's not the claim you just made, you just again fallaciously said
With less than N log N.
Which can be seen to be wrong by inspection.
It can be seen wrong via any of a plethora of net references, or references in books, papers, etc.
You are, once again, shown to be completely wrong. Your claim is false.
You are attempting to cover your theft of academic credit from Cooley and Tukey with this absurd smokescreen.
Mind you, considering BTox and WildCat, you show promise.
But not good enough...
It is quite unwise of you to continue down this path. I suggest that you now admit that you are utterly inept, uninformed, and ignorant compared to a fat old white boy from the Applachians.
ADMIT IT NOW. A fat old white boy from the hills has utterly whupped your butt.
Now, admit it, retract all of your accusations, and admit that your level of knowledge and ability to debate are both shown to be inferior to the americans you demonstrate such bigotry against.
----
Some more references showing your claims false:
http://www.hsdal.ufl.edu/Projects/IntroDSP/notes/Lesson_15_Other_DFTs.doc
saying: "Goertzel's Algorithm
Goertzel algorithm is build using 2nd order IIR resonantors in a filter bank configuration. The chirp-z transform DFT (CZT), uses convolution to perform a DFT. Both have limited use in some applications and with specific technologies. The Hartley transform produces data in the transform space that can be algebraically mapped into the DFT if required. Their applicability to fast high-order DFT computation is minimal and, as a result, will not be developed at this time."
and http://www.hsdal.ufl.edu/Projects/IntroDSP/notes/Lesson_12_DFT2.doc
that says:
"Goertzel's Algorithm
Goertzel's algorithm pre-dates the FFT. It has been used for many decades to selectively analyze a signal in the frequency-domain. Goertzel's algorithm is based on an interpretation of the DFT Equation 1 in terms of a convolution. "
And that, in a nutshell, is what you do in DTMF decoding. It analyzes SELECTIVELY. It is not a transform.
There are 100's of references. Each one shows you to be wrong.
Now, let's get back to some of your other lies. You asserted, a while ago, that I didn't know what Fermat's Theorem was. Please detail your evidence. Please explain this in the presense of the counterexample presented.
Show me the unstable FIR (not some IIR implimentation of one).
Show me where Laplace reads on the stability of FIR's, or IIR's in the digital domain, for that matter.
You can't do any of those. You are shown to be false, false, and false.
You accuse others of ignorance, yet you fail nearly every time you make an assertion.
Apologize, admit your false nature, and the false nature of your claims, and admit that you have been bested by fat old white american hillbilly.
Ion said:
As I have shown, and as the various references I've quoted have also shown, Goertzel is a subset of a DFT. It can be a full set if you put up with substantial inefficiency compared to an FFT.
Get your own story straight first.
You may now offer your abject apologies. You are a liar, a libeller, and your claims about Goertzel being faster than an FFT for a full transform are shown to be fraudulent.
WildCat
NWO Master Conspirator
- Joined
- Mar 23, 2003
- Messages
- 59,856
A genius in your own mind. Keep tilting at those windmills, lon. Anything to avoid responding to jj's challenge, eh?Ion said:
Why don't you just hop your frog a$$ back to France where you can all pat each other on the back, while telling yourselves that despite France's steady decline towards irrelevancy, you're really superior to all those stupid, fat, ignorant Americans?
And yet, you French had to pass laws to try to keep American influences from encroaching upon your language and "culture"! If that's not an admission of inferiority, and of desperation, I don't know what is.
And why don't you make showering one of the "many things you do every day"?
Why don't you just respond to jj's criticisms? So far,my score has jj up about 12 to 0. Approaching the slaughter rule...
Cat,WildCat said:
you are the backup chorus to someone who fails in his education to this:
Nevermind Laplace yet, but just based on the example of my quote above, how does it feel to be the backup chorus to an ignorant?Ion said:
You are more ignorant than an ignorant like jj...
No I don't know when to give up, since they bring me in because jj, Cat and toxy can't handle D.T.M.F. Detection by Goertzel and I do.BTox said:
'Goertzal' -as it is persistently misspelled in this thread even after I correct it- that's still a pearl of senility.
(As for your favorite topic, toxy, the consumer confidence went down in February from a high in January, meaning like I said watch for the consumer confidence being high in January only as a mini-thread but not as an economic full recovery.)
Ion said:
Thanks for the confirmation. And as to DTMF, of course I cannot handle it. My education is in biochemistry - surely yet another subject you haven't a clue about.
Ion said:
Ah, so the 40 point rise in consumer confidence since last year is just a blip but a 10 point fall in one month is the real trend (preliminary data of course, the real data comes out next week - and btw the conference board data is most trusted). Yet more ion inanity. Do you know what a recession is yet? Check out your beloved Germany and see if they are still in recession...
BTox said:
Just for the record, it's been a solved problem for years, BTox. He has what sounds like a good implimentation, but of course that doesn't address his fraudulent claims about FFT's vs. Goertzel.
His basic point of deception is that he'd like to snow us into thinking that one use of a Goertzel filter, giving us one line, is somehow equal to an entire FFT. Of course, it's 1/N of a full transform, leading to what all of the references agree is a process that is order( N*N), as opposed to the FFT that is order (N log2 N).
He's also attempted some other deceptions, by trying to slip back to DFT's, which are an inefficient way to do anything beyond a 4 or 8 point transform except in the most unusual of circumstances, instead of acknowledging what an FFT does. Of course, he did say "DFT" and "N log N" in the same breath the other day, which of course isn't true except for FFT-type algorithms, but so it goes, so it goes.
