Golden Ears Fail Again

[theaudiohobby]

Now, I know these books are not recent (I took these subject a while ago), nor are they as authorative as your Google searches; but humor me and check out your textbooks. Please provide a reference to the key concepts of time and frequency "domains". Such books are a cure for ignorance. Any explanation as to why these books don't mention this assumedly basic concept?

First music signal is a time-variant signal, that it characteristics vary with time, however in electronics (especially in digital audio but not exclusive to it) musical signals can be represented in the as time invariant signals, here the signal is defined in terms of it frequency domain performance. This is possible because the frequency domain performance of a musical signal is defined in terms of discrete time samples. An analog signal can be successfully defined as an infinite set of successive discrete time signals. The frequency domain performance of time domain signal is effective a summation or (an integral) of the time domain representation of the given signal i.e. time parameter is a fixed quantity in the frequency domain representation of a music signal. In the time domain, the relationship between the input and output of a given signal is defined by the impulse response, however as can be readily appreciate this impulse response of a signal is frequency dependent, the frequency domain removes the complexity by assuming discrete time samples.

As can be readily seen from the brief introduction above, the frequency domain representation is simple performance parameter than time domain representation since it assumes time invariance i.e. fixed impulse response. This approach was popular in the early days of audio but the technology of the time, i.e. low sampling rates e.g 44.1KHz soon uncovered an unforeseen problem, i.e. an FIR filter in the name of brickwall filters have detrimental effects on audio quality at low sampling rates. Generally, filters in general are not subjectively benign at audio frequencies, due to their large departure from what can be regarded as an impulse response. Therefore the departure from the ideal impulse response has to be accounted for, an issue that is largely acknowledged in loudspeaker design and digital audio and more recently in amplifier design because of it measurable as well as audible consequences. The time domain behaviour is an issue in amplifier design as the published measurements of many amplifiers readily demonstrates that not an inconsiderable number of them show frequency dependent distortion and some of this frequency dependent distortion has its origin in the time domain.

So in answer to your questions RobotCzar here goes,

Why is it you think a book on "audio circuits" (whatever those are) would be different that one on electronic circuits?

Audio circuits deal with time variant signals, not all signals are time variant anyway.

Aren't audio circuits electronic?

Sure they are.

Aren't electrical waves carrying an audio signal following the same laws of physics as those carrying other signals?

Yes, the behaviour of a 300MHz signal and a 20KHz are different in so far as the relevance of various performance parameters, In certain fields, it sufficient to define a signal in it frequency domain representation.

Do differences in the "frequency domain" and "time domain" apply only to audio signals?

Most certainly, there is whole body of work on this, I have already referenced some in the past.

Does Mr. Horowitz back up your claims about the significance of the "domains"?

I am fairly confident that he does since he mentions the concepts alongside a RC circuit, moreover frequency domain and time domain resolution have become fairly standard parts of signal representation theory.

What do I have to do to get you to explain how errors in the "time domain" are resulting in bad audio?

I have already done so, it is by no means definitive but it points you in the right way

And please explain why audio circuit designers keep ignoring this "problem"?

The simple answer is that they do not and this can be readily seen in the material that you have poo pooed, when you amplifier references. Remember this, frequency response measurements does not capture phase shift nor does it capture the any departure from the ideal impulse response. For example, Time alignment is time domain effects compensation facility.

There is sufficient reputable work out there, some of the previous material were lectures notes from Stanford!, to show that you that your posts were simplistic and totally wrong wrt to their conclusion on time domain and frequency domain behaviour. I hope I have succeeded in humouring you and that you will eat humble pie and accept that you were 'out of your depth' when you ridiculed my objections to your overly simplistic view of audio quality measurement and perception.

Finally, what kind of references on time and frequency domain do you want to see?