Originally Posted by paradoxper
Right. He indicated differences between DACs through subjective measures. Makes a ton of sense in this venue, er, or not.
Esledude touched on your question 2 posts up.
Umm yes they do indicate differences were perceived. Consistently so. Do things not count as fact unless level-matched? When you change songs do you need to level match to confirm that you have, in fact, done so? I don't think you would, so there IS some information we can gain from listening without level matching, correct? I wasn't present at the time, but i doubt they were doing their listening at wildly different levels and at these levels they found differences. Not if the perceived differences only came about through things being louder, the loudest dac in the test surely would have done. I do not recall that it did. (Will check.)
As for the statement "analog design doesn't matter" really? Where's the proof of this? Isn't amplifier/buffer design all about reducing distortion? What if it was poorly designed? We've seen competent companies stumble before - meridian explorer with it's high output impedance, akg 812 with its funky square wave response (source: inner fidelity) so some quite blatant oddities that would have showed up in the design stage do reach the market. Who's to say there are none in dacs?
There's so much absolutism around here. It is so decisive when people say things "sound the same" or "don't matter" or "inaudible". But without any data it's just the same as others saying "unicorns". The people who say otherwise have to drag up all sorts to meet with blunt statements with no actual factual data or even any reasoning to back it up.
Dac chips do the same things therefore all dacs sound the same. Great. Now that's pretty much the same as saying all computers with the same cpu and clock speed perform tasks at the same speed regardless of ram, hdd, operating system other software. It may be measurable with a stopwatch but surely a few seconds are unnoticeable by humans therefore they're the same. What?
In the final part of his post eslude mentioned that (and i paraphrase) it is possible for dacs to sound different and also that it is "fairly uncommon" for them to do so. Now this is, i think, quite a bit a way from "all dacs sound the same".
I am not advocating that *all* things that can be measured can be detected by the human ear. But there is a trend here of:
1.) stating that non difference can be found through measurements, then
2.) well alright some can be, but it's not audible
And both are without any justification other than saying pretty much that dacs are dacs. I'm pretty sure we all understand there's more than a chip in that box and everything has to be designed and the design may or may not influence the outcome. It is clear that in some cases they do. There's an article and i think a youtube clip from ps audio (iirc) about the role the reference voltage plays in their new dac. Basically it says that their fpga works against a reference voltage and if this voltages shifts the conversion goes awry. Now in a lab with a controlled simulated mains supply this may work perfectly, but a good designer will design in a good tolerance to real home supplies while a bad designer may not thus things will possibly sound odd with the latter. Just an example of how other parts in a dac can make a difference. Not saying it will, just that it can. It is for me to find information that it does (in any case not just this) AND for people who think otherwise to produce their information that it does not. Not just keep banging on about things are "inaudible" and "no difference".
And even if we ignore all results/comments however unscientific, the reasons i mentioned earlier (3 of them) remain as possible causes of differences in the eventual output. And if there are possible causes we should look into information on what is and is not audible.
Finally, no company wants to increase cost so if they really do make no difference, why bother with designing these things? It cannot all be marketing/trends/fashion because if it is, dacs will look a $hit ton better than they do now cos that will be all that matter. Benchmark and i think Mytek are also in the pro industry so they must be doing something right and reasonable. People made a song and dance about how many companies use Sony CD transports but the industry as a whole accepted it, so why not do the same with DACs? It would make a load of sense, surely to just buy a dac board the same way they did transports because not so long ago a lot of the more established manufacturers really had little knowledge about digital audio and that would save them a load of money trying to design their own. A lot of facts that have nothing to do with placebos or bias also point to the possibility of there being something more to dacs than "they're all the same" and "nope, inaudible".
Lastly, just because it has the word "science" in the title of this forum it does not exclude discussions about observations. Purrin's test is just as valid here as the statement saying "all dacs sound the same". How many dacs have you listened to? Is it a representative sample? Level-matched? What are the control conditions and variables? Was it a double blind test? If this venue is open to one, it must open to another.