Believe it or not, while sighted tests are susceptible to all sorts of bias, occasionally they are actually correct... or, to be more precise, occasionally what you think you hear is actually closely related to what you DO hear.
I've also heard rumors that some art experts were actually able to tell a real Rembrandt from a cheap forgery, and be correct at least some of the time, before colorimeters and mass spectrometers were invented.
In the case of ringing, I wold agree that it is not "THD".... I would say it technically falls under the category of "noise".... and, further, "data correlated noise".
(Which is just a fancy way of saying that it is noise that is correlated in some fashion to the signal.... )
However, my point was not what to call it...
My point is that it is something which is clearly audible, at least under some circumstances, but will not show up on traditional frequency response, THD, IMD, or S/N measurements.
It is only detectable, or measurable, with certain specific types of test signals and test equipment... however, those test signals are quite characteristic of music, which pure sine waves are not.
(And anyone versed in science knows that, when devising a test, it must be designed to approximate the actual usage conditions as closely as possible, and not be chosen "because it's easy to measure".)
I'll give you a precise analogy.
Let's assume that, in an otherwise very quiet room, I set off a firecracker once an hour...
If I measure the SPL of the explosion instantaneously I will get a very high reading...
But, if I average the reading over an hour, the firecrackers will make almost no difference at all in the reading...
And NEITHER of those measurements would accurately reflect "how audible the firecracker really is".
(Neither is wrong; but neither accurately represents the entire reality of the situation.)
Ringing is exactly that sort of situation...
If you measure the contribution it makes to the average levels of noise or distortion it will be small...
However, if you measure it over very short periods, at certain instants, there will be a signal which is pure noise (or whatever you care to call it)...
And we're right back to the question of whether a burst of that particular type of noise, of that particular amplitude and duration, is audible or not...
And, incidentally, being electro/magnetic and electro/mechanical, even the "linear transfer response" of speakers isn't really linear...
So the amount and type of ringing will be different if you make the measurement at different SPL levels.
So, in this specific case, if you send a continuous 440 Hz sine wave through a Sabre 9018 and an AD 1955, and measure the steady state THD, it will be very low for both.
And, if you measure the frequency response of both, it will be quite flat - and nearly identical.
HOWEVER, if you send a tone burst at some particular frequency through both, then measure the output for a few seconds after the signal stops, you will see significant
DIFFERENCES in the output.
The actual differences will themselves be very different depending on the frequency, duration, and envelope characteristics of the test signal you choose.
And we're right back to determining whether those particular differences are or are not audible.
And, as far as i know, nobody has actually conducted a comprehensive test to determine whether the various filter settings on specific DACs are or are not audible.
All of the tests I've seen, when the details were documented at all, tested individual specific hardware.
For example, when whoever it was did that test between CDs and high-res files, what filter did they use on the 44.1k "CD quality sample"?
How steep was the filter?
How much ringing, and at what frequencies, did it exhibit?
And, did the particular music they used as a test signal include transients with very little ringing to begin with, so as not to mask any ringing caused by the filter?
In fact, what was the overall provenance of the music signal they used?
And how much ringing was present in the master recording because of the particular microphones, preamps, and ADCs they used?
And what filter was used on the high-res audio signal?
Etc...
This is where I see a very solid distinction between "scientific testing" and "consumer reporting".
Are we trying to determine whether "there is a clearly audible difference between most DACs when tested with typical consumer music files"?
Or are we trying to determine, scientifically, if there is ANY audible difference between different DACs, even if it can only be detected with certain test signals, or on certain audio tracks.
Many folks may only be interested in the former...
Whereas, my interest is more scientific, so I'm looking at the latter...
As far as I know, MANY people subjectively say that they often hear significant differences between Sabre DACs and other DACs (in specific).
The manufacturers claim that there are audible differences - and claim to have "done the tests".
Are you aware of anyone else who has actually tested for an audible difference between a variety of products with Sabre DACs, and a variety that use other brands, with a variety of music and test signals, and a variety of other equipment?
I have personally never seen or heard of such a test.
And, since there ARE in fact measurable differences, I prefer not to claim to know, beyond my subjective personal experience, whether the differences are audible or not.
I would find it interesting to see actual PROPERLY CONDUCTED tests - with actual results.
However, lacking that, I prefer to avoid reaching conclusions based on vaguely similar data, derived from incomplete test results.
You could spend your entire life banging blocks of metal together - and eventually conclude that "banging blocks of metal together never produces an explosion"...
I would even say that "it's a good ballpark estimate that all you get when you bang metal blocks together is a bit of a clang".
However, I would suggest you rethink your conclusions BEFORE someone believes you, and decides to try it with Plutonium...
(Luckily, or unluckily, some scientists figured out that particular exception to "the rule", and tried it out.)
It has been my personal (subjective) experience that MOST DACs do sound very much alike... except some of them don't.
It is my THEORY that the difference I hear is related to the filter differences I can measure... but I could be entirely wrong there.
(Perhaps it's the color they painted the plastic case that really does it...)
I would rather think that the reputation of the Sabre DACs would account for the 1dB of subjectively perceived high boost (where 1dB shouldn't be audible anyway)--unless you specified anywhere that the tests are blind?
That matches no definition of THD I'm aware of... 1. it would be at the same frequency you fed it with, hence not a harmonic 2. it can be accounted for by the linear transfer function of the speaker which does explain ringing without the need for talking about distortion at all.
Again, there is no technical sense in which this is distortion--as for controlled research, I believe that controlled studies relating to CD vs high res would fit the bill, since CD reproduction would have included the brickwall ringing while the high res version wouldn't (not to mention including extra ultrasonic info).