So here's the thing. If the circuitry can change the sound of a DAC, how is it possible that all DACs, with all kinds of circuitry, could sound the same, when this DAC alone can sound 2 different ways?
As it has an “Upsample” switch, I assume that when it’s switched off it is not up/over sampling, IE. It’s a NOS DAC. Since digital audio was first proposed nearly a century ago, a fundamental requirement is to bandlimit the signal to half the sampling rate, IE. To have an anti-alias filter when converting to digital and an anti-image filter when converting back to analogue. NOS DACs made in the last 20 years or so for the audiophile market often break this fundamental requirement. This results in certain issues, including a gradual high frequency roll-off starting around 2kHz and extending to 22.05kHz (in the case of a 44.1k sample rate), plus relatively high levels of ultrasonic artefacts (“images”), which are likely to cause IMD downstream. Both of these are within the audible spectrum and therefore many NOS DACs can be differentiated (by human hearing). When some say “all DACs” they obviously don’t include broken DACs, and NOS DACs can be described as “broken”, although deliberately “broken by design” rather than because some component has failed.
Are they perhaps related to the fact that the above measurements are related to specific aspects of the electrical signal while what we hear is a complex or a sum of aspects?
That’s certainly a major part of it, although there are also other factors. When we measure signals (analogue, digital or acoustic) we are obviously measuring the properties of those signals using absolute scales (decibels, Hertz, etc.) but human hearing typically does not perceive these absolute scales, it perceives different, relative scales. For example, the difference between say 100Hz and 200Hz is obviously 100Hz and is perceived as 1 octave, the difference between say 600Hz and 1,200Hz is obviously 600Hz, six times more than the previous 100Hz difference but it’s perceived exactly the same, a difference of 1 octave! Other perceptions are not directly related to an actual audio signal property but to a “complex sum of aspects” which also are not absolute. Our perception of “loudness” is a good example. Loudness is not a property of audio signals, it’s a perception which is a combination of properties (frequency and amplitude) but it’s not a straight “sum”, it’s more complex than that (involving an average of time and a transfer function). It seems that many audiophiles are commonly not aware of this basic fact. When we measure audio equipment/signals we obviously cannot measure human hearing perception because there isn’t any, audio equipment/signals do not have any human hearing perception.
But it’s easy to hear differences (for me and whisky agnostic) between two speakers, even the same speakers in two different rooms.
It ALL depends on the magnitude of the differences and where in the frequency spectrum they are. The same speakers in different rooms are going to produce numerous variations throughout the audible frequency spectrum and typically with swings of 20dB or more, due to room acoustics. Even cheap DACs are virtually ruler flat, any variations are in the least audible (or inaudible) frequency spectrum and typically the swings are less than 0.4dB.
It’s also very easy to hear differences between two masters which is where I’m spending my time (to find them) and money now.
Of course, there should always be audible “differences between two masters”. Why would a record label spend the time and money for a mastering engineer to make another master that sounds exactly the same as the master they’ve already paid for?
As I’ve seen other members mentioning it before, it is important to match levels between two devices under test, and it’s not always easy, as it should be below 0.2dB.
A match of 0.1dB or less is required scientifically. The problem is that a small difference in level is not perceived as a difference in level/loudness, it’s perceived as difference in sound quality, the louder one is usually perceived as “sounding better”. This is why there has been a “loudness war” for the last 60 years or so.
Hey CofA, I'm rusty on this, but if there are measurable differences between 2 DACs, what happened to Bit Perfect?
There will always be measurable differences between 2 DACs. In fact there will always be a measurable difference between exactly the same DAC, set identically and converting the exact same recording! The conversion process involves the application of dither (which results in white noise) and all analogue circuits introduce thermal noise (which is also a type of white noise). White noise is random amounts of all the frequencies (with certain statistical distributions) and because it has this random element it will always be different, even the same DAC set identically. This difference is minuscule and usually too low to even be resolved into sound (let alone be audible) but we can measure it.
G
