You are once again using "steady state measurement" in a proprietary and incorrect way.
Your previous "definition:
Also doesn't make any sense. Linear time invariant systems (LTI) is a class of systems and it does NOT by definition exclude any "transient processes" (whatever one can infer you mean by that). Time invariant systems are systems that have the actual output independent on the past system behavior. "Steady state" is nowhere used in defining a LTI.
To add insult to injury, the definition of LTI has nothing to do with "steady state measurements", so your "therefore" logic implication is simply a blow in the wind. You were clearly told what a "steady state measurement" is understood by, however you continue using your own definition to promote yet another pile of bull chips.
Last edited:
If you had ever produced any cogent argument why it wasn't a model (instead of just insisting it was & that's, that) I might change my mind but so far, nada.
Just incase you forgot Mark it was about the fact that it is not a model that music is made of sine waves.
What are the distortion products produced during a fade of a single frequency sine wave signal.
Dan.
Dan.
I don't care if you change your mind or not, like I said you don't understand the basics, very easy for you to educate yourself, not my job sorry. If you ask nicely I have an example for you but I reckon I'd be wasting my time, your bias is too strong
With all due respect mmerrill99, there are only two possibilities.
a) You have no idea what you are talking about
b) you purposefully insult well over a century of cumulated education of those members you are addressing.
I personally still believe it's a) but others may think otherwise.
Mostly that is conceptually true. However, strictly speaking music is not made out of sine ways extending infinitely forward and backwards in time. That would make them start and end beyond the probable bounds of the existence of the universe.
What we can do is assume the music, say a song, repeats infinitely over time, and find the sine waves that would exist along with it in that case.
As a practical matter, we can ignore the extremes of mathematical theory and use FFTs as a very practical tool. I do that too and it can work quite well in many cases. Always worth doing. However, as a practical matter there are some things it won't show well. There was a patent IIRC to randomize the crossover voltage of a class AB or class B output stage. That makes the crossover distortion appear at a random point in each waveform. If we take a long FFT to get good frequency resolution, that crossover distortion will seem to have disappeared which might make someone miss it or believe it to be inaudible, which it is likely not. If we use a short FFT of one cycle of a test sine wave, then the crossover distortion will not be random for that one cycle, but then noise will also appear to be more deterministic.
Again, I would say music can be said to be composed of sine waves, at least for many practical purposes. However, it can also be considered not to consist of sine waves in reality, if we are concerned about certain practical problems that may be encountered.
Sorry, I know its complicated, but the intent is not to muddy. Its just the way the world and reality is, its quite complicated when we look closely. And, often we argue with different underlying assumptions which if we just spelled them out clearly might resolve the dispute right there. Really, people probably do agree in many cases when it comes down to all the details.
Last edited:
Sorry when I was in China, they said they intended to integrate them together. I guess the performance did not justify it.
Of course, and guess why? Because the Fourier theorem is violated. Such a system would no longer be a time invariant system, the system non linearity is now a function of time, Unless you consider the average of the stochastic process involved in randomizing the crossover distortion, then Fourier would successfully apply. BTW, such a time variant system could also generate non harmonic spectral components.
So your example is hogwash.
Otherwise, it was clearly shown by myself and others that the Fourier transform applies very well to aperiodic signals. If you believe such an approach would bring any new insights, then please apply it and come up with some extraordinary proof for an extraordinary claim.
Last edited:
Yep, as I thought, not a shred of counter argument, just the usual attacks & insults of the poster
As they say when that is all that's left for countering, it's obvious which side is threadbare of thoughts/understanding of their contentions
When are we going to hear the "fourier denier" trotted out? What about nobel prizes? Haven't heard that counter in a long while. "centuries of cumulated knowledge" is a twist on it but I guess there are only a couple of different ways to hurl the same old attempts at insult?
Yes, apparently a sliding bias schema would not be good enough to minimize crossover distortions 😀.
What is being missed in the sniping is the joy to the 5% that there is now another option in reduculously OTT performance coming on the market which, for the DIY'r should be about 1/3 price of the Benchmark AHB2. I think this is cause for celebration even if totally wasted on me
Examples of an infinite sine with various windowing functions to make it physically realizable.
Windowing adds a continuous spectrum to the original sine frequency.
https://www.researchgate.net/profil...ect-of-windowing-functions-on-a-sine-wave.png
No worries, even their external headphone driver IC would not cut it for Richard based on the somewhat arbitrary requirement that a headphone be able to drive planar magnetic headphones to deafening volume levels.
I wonder who designed the IC in the dongle? Apple does a lot of stuff in house these days, but I'm not sure an audio codec chip is important enough for them to acquire a team. IIRC they have used Cirrus and Wolfson parts in the past.
I must have missed it, what is the new amp?
There are things so wrong that arguing about would be useless. You know the saying with teaching <your favorite pet> to sing. It would be too hard and he would be annoyed.
You know very well I was speaking in that context about practicalities of numerical approximations used in measurement, what they show and don't show.
- Status
- Not open for further replies.