the importance of analog sound in prog

 

The views I hold and the positions I take are not based upon a belief system, though any statement made carries a degree of implied bias that is the inevitable consequence of answering or addressing someone else's opinion. If I state that someone's view is not accurate it does not follow that I accept that the opposite view is (by inference) not inaccurate. If a subjective statement contains an objective inaccuracy then I am at liberty to address that inaccuracy from an objective viewpoint without making a judgement on the accuracy of any counter statement from an objective or subjective viewpoint. Correcting the objective inaccuracy does not change the subjective opinion, if someone prefers something and makes an inaccurate claim as a result of that then showing the claim is inaccurate does not alter their preference or pronounce that they are wrong, it just means their resulting deduction was not valid in that instance. This also holds true for all analogies made in support of any opinion - showing that the analogy is invalid does not defeat the original opinion, it merely shows the analogy wasn't a particularly good one.

 

Science, technology and engineering are not belief systems, they are not immutable truth, they are subject to change and modification as new evidence is collated and that is the basis of all science. General relativity was not a paradigm shift in science, it did not replace Newtonian physics but was a refinement on something that worked (as opposed to something that didn't work, such as Aristotle's ideas on gravity), the sea change occurred in the change to Newtonian gravity from Aristotelian gravity 200 years earlier. Showing that science can refine and improve on one idea does not show that science's apparent inability to support a completely unrelated idea is a failing of science. When all or any of the known evidence fails to support the hypothesis then the hypothesis is wrong, not the evidence - if an amplifier shows a varying frequency response then it shows that the response is not linear and it is not perfect, therefore the amplifier is neither linear nor is it perfect. If that amplifier is used in a system then that system cannot be described as a perfect system, regardless of how much better that sounds, or whether that sound is preferred over a system using a more linear amplifier. Not saying which I prefer is not a prevarication on my part, I merely make the observation that a non-linear response cannot be called perfect or superior. My determination not to make a subjective statement over a system I have never heard is not an unreasonable stance to take because age and experience has taught me that I cannot make subjective claims based upon objective evidence. I, of course, can make assumptions and speculate on what a system could sound like, and from that I could predict whether I would like it or not, but I cannot make an unequivocal statement about it or support or deny anyone else's opinion without first hearing it.

 

I do not use technical specifications to show how good a system is, I use technical specifications to show that my explanation of why a system subjectively sounds like a good system is supported empirically. If a system with 0.1% distortion sounds better to the ear than one with 0.01% distortion then the specification shows that the preferred system is more non-linear and that is all. The science and understanding of what that distortion is provides the explanation of why that would sound better. My assertion in that instance would be that the nature of the distortion is a contributory factor in the sound characteristic of the system, just because the values taken without understanding what they mean appears not to support that subjective assessment (that it sounds good, great, superior) does not discredit the system or the science. But if someone makes the claim that the system is more linear because it sounds better then all I can do is point to the specification and say that it is not [more linear]. 

 

 

 

I will confess that my hearing is not what it was when I was 19, I am now 55 and have spent a lifetime not looking after my ears: I listen to loud music continuously; I have worn ear buds plumbed into Walkmans and iPlod for long periods; I have spent hours recording and mixing music while wearing enclosed headphones; and I have attended more live rock and metal gigs than I care to count (including being in the mosh-pits at Slayer, Metallica and Motorhead concerts) - I have also worked in industrial environments before the days of compulsory ear-defenders and enforced health and safety regulations. I have had my hearing tested and like most people my age, I know I cannot hear anything over 16KHz. Yet I can hear the distortion present when one byte of a waveform sampled at 192KHz is corrupted - and that is counter intuitive because I know that corruption equates to an impulse transition of 5.2uS, or a Nyquist frequency of 96KHz, so how is this possible? Does this not disprove everything I've said about what we can and cannot hear in whatever system you prefer? Does this not show to be a lie everything I have said about the frequency response of one format over another, of amplifiers and speaker systems and moreover, what we has humans can possibly hear? How is it that I can hear this impulse "click" when my hearing has been scientifically measured to be incapable of hearing anything above 16KHz????

 

Of course science comes to the rescue and provides the answer that requires an understanding of mathematics to fully appreciate: that impulse function in the time domain does not equate to an impulse in the frequency domain. What that means is that the click is not a single frequency or band of frequencies that are above what we can hear, it contains frequencies that we can hear (even with aged hearing) which is why we can hear it. (For anyone who does understand the mathematics) the proof of that is relatively trivial - if we apply Laplace transforms that convolves the time domain impulse into a frequency domain response we can show that the impulse function in the time domain is the unity function in the frequency domain, this means that the ideal impulse contains every known frequency from 0Hz to infinity, and obviously some of those fall into the 20Hz to 20KHz bandwidth of our audio systems. Now we cannot create an ideal impulse function in the real world (an ideal impulse has zero width), what we present to the amplifier is a pulse of finite width (in this case 5.2uS) and the Laplace transform of a pulse is represented by the sinc function in the frequency domain [this is the sin(x)/(x) function mentioned several pages back] which means that the pulse contains a lot more low frequencies (that we can hear) than high frequencies (that we cannot). What that looks like (on a spectrum analyser or after Fourier analysis) is this:

So a fast pulse that appears to be much faster than we can possibly hear, or what can possibly pass through any audio system, is in reality a simultaneous spread of frequencies that we can hear that extends way beyond the bandwidth of the audio system, our ears and (perhaps surprisingly) the 192KHz (1/T) of the sampling frequency. Therefore the science fully supports what our ears tell us is true. (This also explains why we can hear clicks and scratches on a vinyl surface btw).

 

So have I now shot myself in the foot regarding Sean Beavan's comment that the human ear can detect differences as small as 6us? (or 6ms or 6ns or whatever actual number he was thinking of was)? [is this an example of me attacking those with greater knowledge?] ... No, I have not. The key phrase in the statement is "detect differences" and nothing I have said here confirms or denies that, a 5.2us click is not a time difference, it is a single event.

 

It could be argued that if the science supports what our ears tells us is true in this example then it should support everything our ears tell us is true in every example. That would be a bold claim that needs to be supported by fact and obviously that requires every unsupported claim to be looked at from a scientific standpoint, which is all I have ever attempted to do here. A claim made without the support of fact is irrelevant and meaningless. I am not here to support a belief system that proves one thing or another, I am trying to determine whether claims made have a valid and supportable explanation within known science. When someone claims that thermionic valves (tubes) sound warmer I have demonstrated the science behind that claim and shown that the subjective assessment can be married to an objective analysis - the science of this has been known since the middle of the 20th century, the engineers who designed and built those valve amplifiers back then knew that just as modern engineers know this today.