The settling time of a speaker is not a joke.
I.e. how long it takes for the energy to reach a low level, like 0.1%.
Plus the ADSR on the way to that point.
Plus the electron scattering and phonon acoustics in the voice coil, wiring, cables, et cetera.
©
I.e. how long it takes for the energy to reach a low level, like 0.1%.
Plus the ADSR on the way to that point.
Plus the electron scattering and phonon acoustics in the voice coil, wiring, cables, et cetera.
©
Last edited:
A FR slice within a CSD plot is the FFT magnitude display of a specific time interval of the full IR which has been windowed out (with fade-in and fade-out) -- that's why it's called a slice. From a slice of bread you can't tell the outlines of the original loaf it came from, that's right.Yes, if we have 1000 data slices of FR recorded over time, for example at 50 nanosecond intervals x 1000, then it is identical to IR.
If we have a single data slice of FR, we can not return to the IR.
Are we on the same page now?
I've noticed many people in this forum which believe they can predict CSD with a single data slice of FR, via SPL tracing and inverse FFT or whatever.
None of them could exemplify this.
No one can accurately predict CSD with a single FR slice, are we in concordance?
A CSD is defined as a consecutive display of *short term* and *windowed* FFT magnitude data (phase not plotted nor useful here), stepped through the pulse. Whereas the full FR plot is, well, a FFT display of the full length of the impulse response, Vice versa, the IR waveform is the inverse FFT (iFFT) of the full frequency response (with mag and phase parts, of course).
You definitely got to grasp this true identity and what it means. When you have the full data of a FR slice within a CSD (that is, with phase data), you can transform back to the actual time data display of the IR it was taken from, within the time limits of the slice. So you can and do actually alwas "return" from a FR to the IR with an iFFT, and vice versa with an FFT.
Your ADSR "impulse response" example is bogus, what are you after?
(Note to self: DFTT)
You do not need to provide an example, but you should accept the fact that there is not a single example anywhere on the internet of FR = CSD, showing all the fine differences, as a simple inverse transfer function would.
What are you so scared of?
Last edited:
There are lots of times where there aren't examples of such things on the internet. Engineering happens to be one area where things often aren't disambiguated, primarily because if it's clear once you have learnt the basic theory that there isn't any point in discussing it (at least among engineering circles). I would imagine that the CSD = FR thing is something that is fairly unimportant and simply ends up as such from the core of control theory. This isn't to say that it isn't discussed on the internet anywhere, it is, but it perhaps isn't discussed in a white paper or similar because it's irrelevant from the point of view of the engineer who understands control theory. This isn't to say that I do, but I respect those that do and the fact that what they say actually makes sense. There have been a number of times when I have wanted to learn about something specific, something you'd figure there would be a paper on, but there isn't. To learn about it I needed to go back to the basic principles and learn the fundamentals that described it.
No one is scared of anything at least from my point of view, I don't understand why you posted it in the first place.
Live example:
FR and CSD plots from CLIO, a well respected professional measurement system.
Attachments
Sorry I meant a procedure like this.
1 - You upload a CLIO FR measurement, without any phase information
2 - We predict the exact CSD, via inverse FFT or anything like that
3 - You upload the real CSD, then we compare the results of predicted versus real
That's fine, but I've noticed that some people believe in flat FR like it's the queen of England and are scared of considering that the time parameter may hold a large amount of sound quality within it
In amplifiers as well, it seems like some people are scared of the Otala / Lohstroh TIM concept
Kastor has spent thread after thread demonstrating that he doesn't understand even the most basic aspects of Fourier analysis or signal processing, and steadfastly refusing to spend the time learning the basics. I suggest that folks suspend interaction with him until such a time that he has put in the work necessary to understand the fundamentals so that intelligent discussion can occur.
lol
I think the issue that I have with ignoring people is that people do read what people write and a lot of the time if we were to ignore people who don't know what they are talking about, rather than trying to correct them, then we would simply have a forum of the blind leading the blind. This wouldn't be particularly helpful, but you do sometimes have to be careful about where you do choose to stick your nose in.
In one of the commercial forums someone was having trouble with the sound of their amplifier and someone else was preaching about how changing the caps to some boutique variety, with literally no important specifications listed in the datasheet, except some audiophile hyperbole, would help sort out the sound. I posted saying otherwise (with recommendations towards caps they should try and why technically, if indeed they did need changing) and my reply was deleted. There was also a reply with the fellow who had recommended it telling me to essentially shut it because they'd had enough of people like me. The other fellows completely misleading and money wasting recommendation was not moderated out, nor was his reply to my now deleted post. Sometimes the blind want to be lead by the blind and it's better to be let them lead. 😱
I find it almost impossible not to correct the drivel I see (and recognise as drivel) posted on this Forum.
There are many areas where I don't have the expertise and I have to keep quiet.
There are many areas where I don't have the expertise and I have to keep quiet.
You are right actually, I need to put in at least a few hours of studying Fourier, which I haven't.
Nonetheless, this is not a "clear cut" case, in the thread linked earlier, the defendant made speaker drivers for a living and posted many measurements.
For sure we hear mostly in the frequency response area, but surely we can not equalize a Titanium and paper cone speaker driver to sound identical? Is that the idea that some support here? I really want to know because it sounds like from a different planet to me.
Nonetheless, this is not a "clear cut" case, in the thread linked earlier, the defendant made speaker drivers for a living and posted many measurements.
For sure we hear mostly in the frequency response area, but surely we can not equalize a Titanium and paper cone speaker driver to sound identical? Is that the idea that some support here? I really want to know because it sounds like from a different planet to me.
Oh, one could probably EQ a titanium and paper driver to measure almost the same, but I suspect they'd still not sound identical
If you have a single cone material but change the surround, it will sound different. And it's easy to measure why. Saying, "Material X sounds Y, but material A sounds Z" ignores the interesting complexity in transducer design.
As a closing comment, if you want to test for "cone material sound" without introducing uncontrolled variables (SY is right) then get Tang-Band W4-1337(Titanium) and W4-1320(Bamboo Fibre) as those are the only drivers I'm aware of which are really identical in every minute detail, except for the cone. With a moderate passband (<3kHz) you don't need any EQ to make them measure 100% the same (also off-axis, this is were it really counts btw). Higher up they don't... that's why I chose the bamboo cone for FR because it's smother on the top-end and did not fall apart sonically as much as the metal cone when overdriven (out of xmax).
- Status
- Not open for further replies.