I should also add in point #2 above that the perception will be strongly dependent on the signal type as well as the level. Pure tones are notably more audible, but somewhat unrealistic. I would expect a steady tone to be far more audible than a temporary tone.
I also know that different musical sources have drastically different signal characteristics and this will clearly play a significant role in detectability. For example, lets say that we have a highly non-symmetric nonlinearity and a signal source that is also highly non-symmetric (Tracy Chapman has highly non-symmetric signals.) Switching the phase of the signal source will have two different levels of audibility - and that's just the tip of the iceberg.
I also know that different musical sources have drastically different signal characteristics and this will clearly play a significant role in detectability. For example, lets say that we have a highly non-symmetric nonlinearity and a signal source that is also highly non-symmetric (Tracy Chapman has highly non-symmetric signals.) Switching the phase of the signal source will have two different levels of audibility - and that's just the tip of the iceberg.
i would like to know what you considered a jab and it does concern me.
sorry that for you that statement isn't palatable or acceptable it's clear and simple enough and as far as i can tell is not ambiguous.
as the conditional statement that it is.
one can also have a signal in a loudspeaker that contributes no audible distortion.which would also be true.
no but gradients of blue can exist.
looks to me like your creating a gradient scale of "blues" here.
this is obfuscation of the original point.
conditional statement that really perturbs me.
i'll gladly expand on this later.
the "we" in this is not clearly defined could you clarify that please.
i think some of this was hastily composed, so i don't know if i'm sure of what your saying here but i think the gist comes through
aye there's the rub!
if the gradient or condition of valuation is fixed all other possible permutations are rendered moot.
or more simply stated why does he get to, or be allowed to fix/set conditions by which the topic is discussed or analyzed?
so that i can ensure i am understanding this correctly:i take it that you mean pure tones make audible distortion much easier to perceive.
transients(which i consider akin to an impulse) like like a kick drum or snare tell me more about a horn whether low frequency or high than a steady state or long duration decaying signal like say a cymbal
amen
Each distortion has a best way to detect it. May I suggest you list known distortions and rate them so as to familiarise yourself with their importance..
Each distortion is inaudible below some level. Don't upset the balance if improving one will compromise another, because then one becomes audible.
The most common failure in speaker designs is probably one of balance.
..as this outdated way of approaching the issue won't lead you to a good speaker. Many a DIYer has had a headache over this one. All systems produce distortion, and it doesn't really matter.
Each distortion is inaudible below some level. Don't upset the balance if improving one will compromise another, because then one becomes audible.
The most common failure in speaker designs is probably one of balance.
pursuant to the discussion of distortion
The difference between linear and nonlinear distortion
Types of Audio Distortion
The difference between linear and nonlinear distortion
Types of Audio Distortion
IME 1/3 octave noise is the signal with which you will get highest audibility of nonlinear distortions, even low order like 2nd or 3rd harmonic.
http://www.diyaudio.com/forums/ever...tortion-new-listening-test-2.html#post4535595
I think it is quite difficult to have a rigorous definition of "a music signal", if we want to make something close to a scientific test. Beethoven's no.9 will probably give a different result on audibility than Tracy Chapman 😉.
Noise is not a steady state signal, I would guess, though it has a statistically rigorous definition.
Noise is not a steady state signal, I would guess, though it has a statistically rigorous definition.
It is impossible to find a single musical signal that represents all music.
Steady-state just means that the signal is "stationary" (stable statistics) and does not vary over time. Your brain can lock in on the sound and detect things that a signal like music would not allow.
One can certainly do "scientific" tests with music signals. Its done all the time.
Steady-state just means that the signal is "stationary" (stable statistics) and does not vary over time. Your brain can lock in on the sound and detect things that a signal like music would not allow.
One can certainly do "scientific" tests with music signals. Its done all the time.
Now we agree, noise is a stationary signal. Still I do not think it would be a steady-state signal.
Re "scientific tests" with music signals - IMO they are questionable and reflect just one and only test condition (HW used etc.) and are valid just for this specific test condition. If you make a test with a group of people in your listening room, the results of such test are not transferable.
Re "scientific tests" with music signals - IMO they are questionable and reflect just one and only test condition (HW used etc.) and are valid just for this specific test condition. If you make a test with a group of people in your listening room, the results of such test are not transferable.
Last edited:
What you say is true of all scientific tests.
If you used bandlimited noise it would also only apply to that condition. Which is more real: some piece of music or noise?
According to your opinion no one could ever make a valid test of anything. So you don't buy any of the results from Toole? They all use a very unique room - not applicable to any other room?
We have a different definition of steady state I guess. It does not mean the signal is deterministic, just not time dependent.
To me it just sounds like you may want to throw out any test whose results you don't like as invalid by nit-picking them to death. Just like the global warming debates.
Good scientists take all available data and weigh it by its applicability and significance and come to the most rational conclusion that fits the available data. This process can never lead to "proof", but it does get you somewhere.
If you used bandlimited noise it would also only apply to that condition. Which is more real: some piece of music or noise?
According to your opinion no one could ever make a valid test of anything. So you don't buy any of the results from Toole? They all use a very unique room - not applicable to any other room?
We have a different definition of steady state I guess. It does not mean the signal is deterministic, just not time dependent.
To me it just sounds like you may want to throw out any test whose results you don't like as invalid by nit-picking them to death. Just like the global warming debates.
Good scientists take all available data and weigh it by its applicability and significance and come to the most rational conclusion that fits the available data. This process can never lead to "proof", but it does get you somewhere.
Last edited:
If some (any) music piece is performed by Motörhead, the spectrum is nearly identical to band-limited noise ;^).
At least the noise is a defined signal and the experiment is repeatable, contrary to pop music samples 🙂
In case we want to find thresholds of audibility, defined signals have to be used and hardware with lowest distortion has to be used. It is easy to speak about nonlinear distortion as non-issue if the speakers are distorting and music samples are undefined signals. With different speakers, different music and different listening level the result may be quite different. With sine tones, noises and good headphones we have at least some control above the experiment.
In case we want to find thresholds of audibility, defined signals have to be used and hardware with lowest distortion has to be used. It is easy to speak about nonlinear distortion as non-issue if the speakers are distorting and music samples are undefined signals. With different speakers, different music and different listening level the result may be quite different. With sine tones, noises and good headphones we have at least some control above the experiment.
A music source is a defined signal and the experiments are repeatable.
It sounds to me like you need to do your own "scientific" studies since you find fault with anything that anyone else does. Let me know how that goes for you.
It sounds to me like you need to do your own "scientific" studies since you find fault with anything that anyone else does. Let me know how that goes for you.
- Status
- Not open for further replies.