I'm not going to overpersuade those whose vocabulary is limited by the word anecdoteBet you he isn't and hope you have the patience to find out
And also those who think that it is enough to check THD in steady state, at best, to remove the graph of THD versus frequency
Attachments
Last edited:
Все о качественном звуковедении. Часть 1. Объективные характеристики аудиодевайсов
Objective characteristics of audio devices Alexander Yenin about high-end equipment designed with a focus on classically measured parameters without understanding what the sound quality actually depends on.
Quote:
“And such a system seems to convey a good timbre, and does not get tired when listening. But the sound is as if there are cotton cushions in front of the speakers. Fuzzy and illegible. And many of the details that are clearly audible on speakers with good impulse response are missing in principle. What is the problem? The only thing is that the system does not transmit the signal increase well. The fact is that it is at this moment of the signal growth that we perceive about eighty percent of the information, which is then converted into a phonogram in our head. That is why the sound is so “clouded” - too much detail is missing. Changes in signal dynamics are due to the fact that the equipment cannot transmit the signal at a sufficient speed. "
Note.
It is precisely such a high-speed parameter as the GD (in the parameters on the op-amp - Time Propagation Delay) that is responsible for the speed distortions according to Jiri Dostal. This type of distortion is absent on sinusoidal signals in a steady state and occurs at the moments when the signal deviates from the sinusoid (at the moments when the amplitude or frequency of the signal changes). And since the music signal is not sinusoidal, speed distortions are always present and the lower the tPD, the less they are!
Objective characteristics of audio devices Alexander Yenin about high-end equipment designed with a focus on classically measured parameters without understanding what the sound quality actually depends on.
Quote:
“And such a system seems to convey a good timbre, and does not get tired when listening. But the sound is as if there are cotton cushions in front of the speakers. Fuzzy and illegible. And many of the details that are clearly audible on speakers with good impulse response are missing in principle. What is the problem? The only thing is that the system does not transmit the signal increase well. The fact is that it is at this moment of the signal growth that we perceive about eighty percent of the information, which is then converted into a phonogram in our head. That is why the sound is so “clouded” - too much detail is missing. Changes in signal dynamics are due to the fact that the equipment cannot transmit the signal at a sufficient speed. "
Note.
It is precisely such a high-speed parameter as the GD (in the parameters on the op-amp - Time Propagation Delay) that is responsible for the speed distortions according to Jiri Dostal. This type of distortion is absent on sinusoidal signals in a steady state and occurs at the moments when the signal deviates from the sinusoid (at the moments when the amplitude or frequency of the signal changes). And since the music signal is not sinusoidal, speed distortions are always present and the lower the tPD, the less they are!
This type of distortion is absent on sinusoidal signals in a steady state and occurs at the moments when the signal deviates from the sinusoid (at the moments when the amplitude or frequency of the signal changes)
I think we have a language issue here. I read this to mean, that if a sinusoid signal shape is changed due to some cause, that can be considered distortion.
I think it does not mean that a signal that is not sinusoid by itself, by definition is distorted. It is not.
Another question, if I may: what is 'transient distortion'? Can someone give a plain answer, without a long exposé about what causes it, why it is bad, etc, etc. Just what the f*ck is it??
Is it slew rate limiting - if so, lets call it that.
Is it transient intermodulation - if so, lets call it that.
Is it a change in shape due to a filter?
Jan
I think we have a language issue here. I read this to mean, that if a sinusoid signal shape is changed due to some cause, that can be considered distortion.
I think it does not mean that a signal that is not sinusoid by itself, by definition is distorted. It is not.
Another question, if I may: what is 'transient distortion'? Can someone give a plain answer, without a long exposé about what causes it, why it is bad, etc, etc. Just what the f*ck is it??
Is it slew rate limiting - if so, lets call it that.
Is it transient intermodulation - if so, lets call it that.
Is it a change in shape due to a filter?
Jan
Last edited:
I'd say transient distortion is a one that's not continuous/periodic and
is only present right at the beginning (or at the end) of the change.
Attachments: I know these are extreme cases but besides that they represent that there
is some "special" error at every "new" signal change which cease after a few periods.
About the name: I guess even if it can be categorized to something more fundamental one
we can use it to emphasize its importance. Like we have different word for sparking and lightning.
is only present right at the beginning (or at the end) of the change.
Attachments: I know these are extreme cases but besides that they represent that there
is some "special" error at every "new" signal change which cease after a few periods.
About the name: I guess even if it can be categorized to something more fundamental one
we can use it to emphasize its importance. Like we have different word for sparking and lightning.
Attachments
So you mean it is settling time? Why not then call it settling time instead of all the verbal diarrhea?
Jan
Sorry if I was not politically correct ...
Let me try again:
Dear Cortez,
What you describe is known as 'settling time'. Nothing is missed, because the part you think is 'missing'
is outside the audio bandwidth. The signal you use is illegal in the audio bandwidth.
There probably is some Wikipedia about it, should you desire to educate yourself.
Jan
Let me try again:
Dear Cortez,
What you describe is known as 'settling time'. Nothing is missed, because the part you think is 'missing'
is outside the audio bandwidth. The signal you use is illegal in the audio bandwidth.
There probably is some Wikipedia about it, should you desire to educate yourself.
Jan
Last edited:
The technical phenomena is a fact the audiability can be a belief.
And why not to call it transient distortion..?
Is it a copyrighted expression of yours?
I guess it's too general, just two simple and very general words, like "signal error".
It's good as it is present in the transient "domain" (in the realm of: DC, AC, Transient analyzation).
And it's a distortion as during this "settling" time it doesn't copy the original signal well and not even with a simple delay.
You don't have to be sarcastic that's never a good way to meet. (If that's your primary goal.)
Illegal..!? :-o Is there a law to this?
I can accept if you see it unnecessrarily fast but that's again the belief layer not worth to discuss.
Again, please don't be dirisive nor condescending as there no absolute definition to that term.
Actually I googled some definitions, here you are:
Transient Distortion - Video 44 - YouTube
Transient distortion in loudspeakers - IEEE Journals & Magazine
Transient distortion occurs when the acoustic output level does not change as suddenly as the input signal.
Transient distortion | Article about transient distortion by The Free Dictionary
Distortion due to inability to amplify transients linearly.
Audio system measurements - Wikipedia
Transient response
A system may have low distortion for a steady-state signal, but not on sudden transients.
...
Most loudspeakers generate significant amounts of transient distortion
But this is the best of all definitions:
Transient Distortion Full - YouTube
Peace!
Could you please post the schematic of the amplifier discussed and put a mark on the points that you "measure"?
I meant if there is a small measurable difference one can say:
(Let's say some different "signal error" like 10mV DC @ output)
"Oh, common, that doesn't matter at all..."
And someone else:
"Maybe it does matter..."
You can not prove easily that it does matter but that's true for the opposite as well.
That's why I called it a belief (like a religion...)
(My "can not proove belief" is that very small and transient signal errors matter a lot.
Mainly in the time domain but also in amplitude as well.)
Here you are. Comments:
- the test signal is 50kHz sine wave (I know starting and stopping too fast...)
- input RC = 1k + 1n
- amp delay = 300ns
- in the middle: THD = "0" and the steady sine already has just a phase shift
- I know it's a "stress test" but it's good to visualize the effect
But if you have a better test signal (which is not a continous sine wave) I'm open.
Attachments
I mixed two signals here just not to start from zero and stick to this thing.
The main thing is that if the signal is changing there is not just a delay.
The "simple delay" is only there after a settling time.
So what a real music signal is rather like?
It has almost continuous transients (I dont mean step signals under that) or continuous regular sine waves?
The main thing is that if the signal is changing there is not just a delay.
The "simple delay" is only there after a settling time.
So what a real music signal is rather like?
It has almost continuous transients (I dont mean step signals under that) or continuous regular sine waves?
Attachments
