drummerman said:
Not at all a stupid question. Measuring well is objective whilst 'sounding good' is purely subjective.
If measuring well would be all-important, quite a few expensive (or otherwise) products wouldn't cut it and probably never make it into production. I will start a thread on this and give examples.
We are all different and thus, a universally acceptable sound does not exist. However, if I'd pay a lot of money for something, be that speakers, an amplifier or whatever then I would expect good engineering practice and certainly would not accept things such as channel imbalance, poor pair matching, high uneven harmonic distortion, a lot of hum/noise and whatever other little ailments you can think of. Unfortunately there is a lot of that about and it simply isn't necessary, actually its undesirable and poor.
There are other examples of 'engineered' sound. For example, many years ago I used John Shearne amplification which was deliberately made to sound more tubey. It measured not so well but it did so because it tried to ape better valve amplifiers. It was deliberate engineering.
With DAC's, they mostly come on boards which manufacturers buy and apparently, they come with clear instructions on what and what not to do as to make the most of the in-built capability of the DAC. - I said it before, some companies manage to make a perfectly good product worse by applying their own ingredients. Audiolab is one such manufacturer which managed to produce a product which measured in some respect rather poorly. Perhaps that was a bad sample supplied, who knows and there are other far worse 'offenders' but personally I'd rather not have to much signal manipulation in a DAC. Especially as it usually only gets worse further down the line/chain.
My example of the Onkyo receiver was simply to show that it can be done at relatively low cost.
regards
Thank you for your thoughts and insights.
For example, Audio Note and Linn take very different approaches:
AN's NOS Dacs remove all signal manipulation such as over sampling, noise shaping, re-clocking or jitter reduction, as they claim these interfere with the critical time domain requirements of the signal. They have also dispensed with all filtering in the analogue domain, to further retain good wide band phase-frequency and dynamically coherent behavior....closer to the original master tape.
They prefer the Analogue Devices AD 1865 18 Bit converter chip, which they say sounds better than the 20 or 24 Bit versions. Nb. The above info was mooched from the manual of their 2.1x signature Dac that I tried for a while.
Linn argue that the Dac plays a small part in the system of converting the stream of audio into the analogue domain. Unlike a lot of other manufacturers, they disable all the internal processing of the Dac (upsampling/filtering etc )which they believe to be limited, and just use its conversion stage.
Upstream of the Dac, they apply their own more sophisticated processing which is beyond the capability of the humble chip.
44.1 kHz is upsampled to 352.8 kHz
48 kHz Hi-Rez is upsampled to 384 kHz
I believe the Dac they use is the Wolfson WM8741 (except the Sneaky - 8740)
The above info may sound like I know what I'm talking about, but in reality I only have a very cursory understanding of these very different approaches...which probably measure very differently.
So in my very convoluted way, I'm trying to say that unless you have a very in depth understanding of the measurements you're looking at, and what implication it will have on the final sound, it's better not to get hung up on them, and go listen.
IMO Both approaches work really well, if properly implemented, and are put in the right system.