The content I reviewed in the first installment of this series was provided by Ethan Winer, a very familiar name to many audiophiles and enthusiasts. Ethan wrote a reference book called “The Audio Expert”. I have copy and value the information provided in his book. I wrote my own book to focus more on the production processes involved in making a great recording and methods that audio enthusiasts can use to improve the quality of their listening experiences. It turns out that Mr. Winer is a staunchly anti high-resolution audio. His stance is based on “if listeners can’t tell the difference, then the whole high-res business is a fake”. He has written repeatedly that CD 44.1 kHz/16-bit audio is perfectly adequate for “most normal music in most normal listening situations”. I agree. But audiophiles are not dealing with normal music in normal systems. We strive for the very best sounding music replication possible. And I believe that high-resolution audio does make a difference!
Readers of this blog might think that I would agree with Mr. Winer. I strongly object to marketing of standard-resolution audio as “hi-res music” and have tried for more than a decade to clarify the confusing offerings by “high-resolution” digital retailers — either as downloads or as MQA streams. They’re NOT high-resolution. The CEA, NARAS, equipment manufacturers, labels, DEG, and others have consistently pushed misinformation about high-res audio. But Mr. Winer and I disagree on the merits of producing and releasing recordings that were recorded using with sample rates and word lengths greater than CD specs — 44.1 kHz/16-bits. I acknowledge that with the exception of a few surveys, most investigations have failed to prove that humans can reliably distinguish between standard-resolution and high-resolution recordings. Some of the most well-known studies have been flawed. The well-know Meyer and Moran that equated the quality of CD vs. SACDs and DVD-Audio discs used only standard-resolution sources!
So I was surprised when the files offered for comparison on Mr. Winer’s site barely qualified as high-resolution. In Part I of this series (click here to link to it), discussed the first classical example. The spectra barely extends beyond a standard-resolution file and makes this file a poor example for comparing resolutions. I’ve heard from Ethan about this issue and he remains critical of my analysis. So I went back to the “pop” tune I acquired from his site. The illustrations below show the parameters of that file.
Illustration 1 [Click to enlarge]
This is not a good example of high-resolution audio. The ultrasonics in the “high-resolution” version actually increase in amplitude above 20 kHz — that’s not natural. And the “standard-resolution” or down converted version has “phantom” ultrasonics. Ethan told me that he added these to confuse people like me that would use tools to identify which file was which. But the added ultrasonics just confuse things further. This file also doesn’t contribute to a fair test. The dynamic range of 24-26 dB can be reproduced by less than 8-bits!
So I decided to analyze one of my own “pop” tunes to see what a real high-resolution files looks like. I used a track from Ernst Ranglin’s “Order of Distinction” Blu-ray disc. Alana Davis sings “My Boy Lollipop”, a classic track from 1962 (I remember the Millie Small original all those years ago). Take a look at a bona file high-resolution track.
Illustration 2 [Click to enlarge]
This is what I had hoped to see when I downloaded Ethan’s tracks. Notice the natural diminution of the amplitude of frequencies as the get higher. They shouldn’t look blotchy and discontinuous as in the previous illustration. The sharp division between standard and high-resolution is evident at 24 kHz. The files are otherwise identical except for the ultrasonics. And a critical component that is part of “My Boy Lollipop” is the extreme dynamic range — above that of a CD (which tops out at about 93 dB when dither is applied). The dynamic range represents that uncompressed signals coming from the band and measures 96 of more! Would you be able to perceive any difference between these two versions? Probably not. But I would maintain that a rigorous study with bona fide high-resolution has not been done. So we don’t really know.
Anyone attempting to perceive differences between Ethan’s “pop” tune at standard and high-resolution specs wouldn’t have a chance. That was my point in analyzing the tracks I received from him. I think Ethan and I agree on the “normal” high-resolution music in the marketplace. But audiophile deserve more and I think that my files deliver it.
I’m in the earliest stages of working with a prominent mastering engineer on another test. We’ll see where it goes.