Originally Posted by Mickey Mouse
I suggest: stop sarcasm and learn about the stuff I suggested previously, please.
When I try to learn/examine the propaganda about this stuff such as Bob Stuart's AES paper on "typical" CD filter audibility I learned:
The transition band filter he used wasn't taken from an actual CD player, at all
, it was his own concoction forced to aggressively filter in a very narrow 500Hz region rather than a comfortably wide 2kHz band as is found in many/most modern CD players.
He didn't use TPDF dithering as is the industry norm for the best sound and which he even admits to in one of his product's owner's manual* [proving he is aware of this].
I was actually the very first person to bring this up, 11/06/14, in another forum: "Did he give an explanation why rectangular dither was used instead of triangular, considering it is know to introduce noise modulation which triangular is immune from? "
From the AES comments
Arnold Krueger Comment posted February 23, 2015 @ 15:20:18 UTC
"I have a problem with this paper's description of the ABX test, which seems to be based on the classic but irrelevant 1950 Munson and Gardiner JASA paper rather than the more recent and relevant 1982 Clark JAES paper.
I agree with Stefan Heinzmann's comments above about the use of either no dither or RPDF dither rather than the industry standard TPDF dither.
It appears that the dither used was spectrally unshaped, whlie it has long been known (for example as expounded upon in the JAES by Vanderkooy and Lip****z, etc.) that for critical applications perceptually shaped dither should be used.
My studies of modern 44.1 KHz DACs suggest that transition bands on the order of 2 KHz are common and that the ca. 500 Hz transition bands used in the simulations are atypical.
The sample rate of the simulated digital filters was apparently 192 Khz, but in fact typical digital filters used in modern DACs run at 8x or higher or 352.8 Khz.
In my mind the above points don't exactly support the phrase "Typical Digital Audio Filters in a High-Fidelity Playback System" used in the title."
I also learned that the release of this paper was timed just months before the release of MQA, which it indirectly supports, probably since AES has a strict policy that papers must not be promotional propaganda for products or services so he released this under the guise of "Here's an impartial examination of typical CD player filters " first,
conveniently before the world [and AES editors] knew of the release of MQA.
"It is well established that applying dither with triangular probability
distribution [TPDF] of ± 1 LSB to a quantisation process results in perfect
linearity, contributes no noise modulation and permits resolution of signal
elements below the LSB. It is therefore the ideal tool to achieve
[bold text emphasis mine.]
Quote source: Meridian 518 owner's manual, bottom of page 37, here.
So in other words he knowingly used a methodology to avoid
transparent processing, so as to make the "typical" CD filter stand out in audibility testing.