Craig M. said:
It has been possible for a long time* to insert an ADC > DAC into an analogue loop and it make no difference to the sound. There are documented blind listening tests where this has been done.
You miss the point entirely. Just because it is technically possible to prove this in a double-blind test does not mean that every DAC above a certain level sounds the same, especially when it is connected to an abnormally high-resolution amp and speakers.
Look, I am wiling to concede that there are many things I value in hifi that if you isolated them and put me in a double-blind test, I probably would fail. But beware of the myth of the fairness of the double-blind test. There are differences that such a test does not reveal - for example, there may be minor distortion that is hidden by the greater distortion present in many power amps. But hooking it up to an amp with a lower noise floor may reveal differences. Or perhaps both A and B sound identical to every listener in the test, but keep both systems for a week and you might discover that one is quite fatiguing to listen to after an hour or so, but the other you can listen to all night. In both of these examples, the amp used for the test, and the duration of the test, respectively, have a far greater influence on the outcome of the test that what you think we're testing for does.
Speaker builders have a similar argument about capacitor quality in crossovers. Some people think you can't hear the difference, some people know that you can. If someone asked me what I thought, I would say that it depends on the quality of your source and amplification.
But more to the point: all of these quality elements might be inaudible if you isolate them for an A v B test. The speaker cables, the DAC's output stage, the brand of tubes, the capacitors in the speakers, impendence correction in the amp, power supply isolation, etc. - I'd be hard pressed to tell the difference on an A/B test with any of them alone.
But I am not hard-pressed at all to tell you that the key to great sound is paying attention to ALL of it. Maybe one of these elements doesn't make a huge difference, but attention to all of them, collectively, is the difference between high-end and average. You can pay almost any price for a 100-watt amp - the difference in price and quality is all in the details.
I think the restaurant analogy is valid - cooking is a similar fusion of science and human sensation. Could Michelin reviewers tell the difference between two identical main courses, in which one used a few tablespoons of fine wine as a sauce ingredient, and another used plonk? If there were no other differences I am thinking probably not. Or maybe the 'golden tongue' Michelin reviewers would say they could taste it, and some would scoff and them and that whole snobby culture. Well, maybe they taste it, or maybe they don't, but I do know that a restaurant that uses fine quality ingredients across the board, and pays attention to every detail, cooks one level of food, and other restaurants, who maybe start with the same cut of meat but are not as obsessive about the details (because they just know it's a waste, in a double-blind who could tell?) - they cook at another, less star-annointed level. Because at that point, the end result, EVERYONE can tell the difference.
It is certainly a valid debate as to how much of a difference each element makes relative to the others. I happen to think that the source is very important, and in an all-digital system such as mine, the DAC is the source and whatever it is outputting is going straight into my amplifier. But that is open to debate - would I benefit more from upgrading the amp or upgrading the DAC? But anyone who tells you that a certain element doesn't matter at all - careful of them. If there is an analog signal running though it, it matters. If you lavish the finest attention on your entire signal path, your system will sound better, just as when you cook with the finest ingredients, the dish will taste better. It is really quite obvious.