Not really. If a full scale 20khz sine wave is continuously "exciting" the amp to a point that no sane musical signal will, however dynamic and full of transients it might be, how can it make sense to call that a steady state measurement according to the definition you gave, which is "after applying the excitation signal, one waits until the system has settled to take the numbers" ? Why would the amp react differently to an isolated transient compared to repeated impulses ? It might be a naive question ...
And indeed, as scottjoplin said, my question applies only to amps (or preamps, headphones amps, etc).
Here's another food for thought......The Manufacture and Sale of Saint Einstein - The Propaganda of Supremacy
I'm still trying to understand audio, well certain bits like BQP and Goop and a bunch of related topics......Electro Sensitivity and 5G are upcoming problems, I am also hearing of adverse health effects of electric cars 😱.
Dan.
When I saw the prototype of the Harley-Davidson LiveWire electric motorcycle
a few years ago, it had a horribly loud, high pitched (~5kHz) whine from the
power converter. Hopefully that has been improved by now for the release.
Last edited:
Actually violates Nyquist something done continually with artificially generated test signals.
If the logic rigorousness of page 107 in your previous link satisfies you, that’s fine for you.
If you think there is intellectual merit in “anti-Einstein” trend that’s also fine for you.
I would suggest you refrain from further promoting here such liking's of yours.
Animosity surfacing under any sentiment cloth and excuse doesn’t get past the radar. Forum Rules 1 & 2
diyAudio Rules
George
Isn't this particular strand in the discussion about measurements of transients (& transient conditions) as opposed to measurements when the system has settled into a steady state?
The point being that measuring with steady state sine wave signals may well miss transient conditions due to settling time
Even ScottW's friend Mallinson talks about settling time between states in his presentation on DACs
Last edited:
It seems to me that he means, by switching, generating frequencies above 20kHz, if that is what he meant I'm questioning the relevance
Except maybe when the transfer function is changing dynamically, such as due to temperature.
In addition, there are probably more types of nonlinearities in mixed signal devices such as dac as verses what we tend to see in amplifiers. Noise floor modulation may be triggered by transients not represented by 20kHz or other fixed single-frequency signals.
FWIW, the Fourier transform of a single cycle sine resembles a sinc(x) function, having the exact analytic expression
F(w)=-(1/8PI)*[exp(-2PI*i*w/wo)/(wo-w)+exp(-2PI*i*w/wo)/(wo+w)-1/(wo^2-w^2)]
where wo=2PI*fo is the single cycle sine frequency (fo=1/To). Visually, this transform in the frequency domain will look like a set of amortised spectral components at wo, 2wo, 3wo,...
For a 20KHz single cycle sine (T=50uS) and 44KHz sampling this signal is violating Nyquist, so the reconstruction of the original 20KHz single cycle sine, is compromised to the point of nonsense. Or the front end anti-aliasing filter will transform the single cycle sine into an unrecognizable (but possible to reconstruct) mush.
This behavior is often used by FUD spreaders that want to prove the CD format "inferiority" compared to higher sample rate formats. Set aside testing a system with a signal that violates the original design constraints doesn't make any sense, such single cycle signals are actually synthetic constructs that are never encountered in the music reproduction. Show me an instrument that emits a single 50uS cycle sine.
Last edited:
@syn08,
The purpose was providing a smokescreen to hide the fact that it is quite difficult to find a quote from the relevant literature that corrobates your point of view.
We both know the reason for this litte difficulty, don't we?
@mmerrill99,
Of course I was; it's one of the misterys of forum discussions that people are quoting a short paragraph but are able to dismiss relevant parts of the just quoted.
In this particular case it is surprising, as PMA perfectly knows that I know about the ability of a 16/44.1 system like Red Book CD to reproduce a 20 kHz FS sine wave.
So, even when missing the "switching on" part, his post is a surprising one....for several reasons.
The purpose was providing a smokescreen to hide the fact that it is quite difficult to find a quote from the relevant literature that corrobates your point of view.
We both know the reason for this litte difficulty, don't we?
@mmerrill99,
Of course I was; it's one of the misterys of forum discussions that people are quoting a short paragraph but are able to dismiss relevant parts of the just quoted.
In this particular case it is surprising, as PMA perfectly knows that I know about the ability of a 16/44.1 system like Red Book CD to reproduce a 20 kHz FS sine wave.
So, even when missing the "switching on" part, his post is a surprising one....for several reasons.
His links did give me a good laugh though.
Conspiracy Theories 'r' us.....
Perhaps Max should treat his posts with some of his own goop? 🙂
I do; it's because you don't want to admit things that come against your agenda. You are conveniently snubbing inconvenient truths; you were told about you confusing (or not understanding, or playing dumb) the difference between a steady state system and a steady state system response; you just got a couple of posts above the reasoning why testing a 44KHz digital system with 20KHz single cycle sines is, mathematically, pure BS. Therefore, somebody has to rub your nose in facts dumbified to your understanding level, although even so you continue to play your foul game.
Of course, you will keep FUDing around and conveniently ignoring arguments, distort, obfuscate, insinuate, create straw men, ad hominem, etc... That's your mission, keep soldiering. If you were not around, soemebody would have to invent you for entertainment purposes.
Sudden, instant turn on of anything (no matter is it sine, square, DC, you name it) spectrally exceeds 22.05kHz as every little kid knows. It is just a relative measure if you speak about 20kHz or 10MHz BW. The limit is always there. Digital audio only works for <=Fs/2 and everyone knows this limitation. However, 20kHz sine is no problem for 44.1/16. Not speaking about switching transients (with > Fs/2 content). Naturally there is an initial envelope when a tone starts, nothing like sudden or step-like turn-on.
Word games, again and again. Nothing new.
Last edited:
I believe it's specific to this thread where people either expend a lot of effort missing the meaning of a post & instead spend their energy looking for a word or phrase or anything that they can disagree with.
i encountered it with my colloquial use of "technicolor sound" from ToS & then with some use of words in my talking about auditory perception.
It seems to me to be a well trodden way of trying to shut down any useful discussion by deflecting & redirecting down a rabbit hole \bout something nonsensical.
Yes, pretty much everyone knows that so I too wonder what he is trying to achieve with his post except to make it appear that you don't know 16/44.1 can reproduce 20KHz sinewaves?
Last edited:
As it was about the "usual measurements using sine waves" it was; I even mentioned "pure tone" iirc, therefore the 18 + 19 Khz or something similar are already a different class of signals used for measurements.
First of all, it is just the way it is meant; "steady state " means just waiting until the system has settled.
Further it is used to get frequency domain information about the system, like frequency response (level over frequency) or distortion figures (THD).
What happens before the system has settled is temporal information and although we know that in a digital record no frequency higher as the Nyquist frequency can occur (and in a sane record there will be a safety margin so the maximum frequency will be lower), we still don't know from our steady state measurment about the systems behaviour before it has settled.
And while the FS signal provides the worst case excitation as large scale parameter, there still is the small signal settling behaviour of the system that we have to examine.
Presumably DACs are included too, so sure. 🙂
No reason to be touchy.
Settling time to a step change, every little kid knows. For signals slower enough than that there are no issues, the system just follows. Waving hands, confusing the issue.
Settling time to a step change, every little kid knows. For signals slower enough than that there are no issues, the system just follows. Waving hands, confusing the issue.
- Status
- Not open for further replies.