Home General Asylum

General audio topics that don't fit into specific categories.

[ Follow Ups ] Thread: [ Display  All  Email ] [ General Asylum ]
[ Alert Moderator ]

Re: Call me a Realist

Dear John,

Let me start with the issue of what you so tenuously term "poor engineering", which would neatly cover the entire digital recording system if one were to be really strict in one's interpretation of the fundamentals, so the issue in my interpretation should be whether DACs with digital filters and over- or upsampling constitute "better" engineering in terms of their "reconstruction" of the signal on the disc when compared to "doing nothing" as we do in the 1xoversampled DACs.

In our view the oversampled and in particular the upsampled version is a much worse interpretation of the original encoding that the one that results from "doing nothing" as it were, here is why,

The sinx/x function is the impulse response of a perfect brick wall filter, derived from the inverse Fourier transform of such a filter response. Fourier theory, as great a tool as it is, however relies upon infinities which exist only as concepts in the mind. In mathematics it is possible to have negative frequency and time, in Nature they do not exist. Fourier theory also requires the infinite resolution of natural numbers, this is not possible in DSP.

If A is a set of data resulting from the A/D conversion of a music signal, and set B as the set whose elements are those of A oversampled and convolved with the sinx/x function of a practical digital filter, with associated truncation errors and such like.

Is A a subset of B? In other words are the elements of A present in B?

Ideally this may be the case, but in practice A is not a subset of B.

Therefore, using logic:

When a set of samples is passed through a digital filter, what you get out won’t be an interpolated superset of the input samples, which is the fundamental premise of the whole technology, they will be an entirely new set of samples.

The ADC samples a music signal and because it has finite resolution in time and frequency an imperfect set of data results. The lost information is not retrievable, it is gone forever.

A digital filter is just that, a filter, the interpolation comes about by virtue of that filtering action. The lost information is not retrieved, in fact, as outlined above, further information must be lost in the filtering process.

Given that the 1xoversampling DACs generally have the ability to uncover far greater differences between recordings, at least in our experience, it goes without saying that the nonos DAC is closer to the "event" on the disc than anything else, imperfect as that may be.

The various analogue filters which comprise the amplifiers and speakers further down the chain are there anyway, regardless of whether a digital filter is used or not, and one would have to question the design competence in amplifiers whose stability is so marginal that the out of band noise from a nonos DAC causes problems.

Surely that is really what should be termed poor engineering?

On the issue of worse sound, may I say first of all that since you have never heard one of our DACs under what you would call controlled circumstances, the comments about the sound cannot be taken for more than conjecture and coming from a leading figure in the audio press it hardly improves industry credibility.

I can easily help you correct this, by getting you any Audio Note DAC of your choice, you may still not like it after hearing it, but that is OK too, all I ask is that you let at least one other Stereophile reviewer listen to it as well.

On the subject of predictability, all I can say is that in the 35 odd years I have had to do with audio, either as a consumer, dealer, distributor or manufacturer I have not come across many products, if any, that had predictable behaviour when placed in different systems and circumstances, so I completely fail to see why this should be held against non-oversampling in particular since it applies to pretty much every product on the market depending on which "corner" one chooses to support.

John, your prejudice is at best misplaced and at worst foolish, as it gets in the way of the magazine you head being in the forefront of audio technology in all its splendour and thus music in all its variety and beauty.

Is your decision to directly condemn, based only on how it measures, what may well be one of the great revolutions digital audio needs to provide greater listener satisfaction, wise or helpful in creating an environment in which we as manufacturers can explore and market the non-conventional solutions that occassionally are the future?

Would it not better to investigate and discuss?

It behoves people in your position to keep an open mind at all times.

Credit goes to my research engineer and friend Andy Grove for the technical part of this post!

Sincerely,
Peter Qvortrup





This post is made possible by the generous support of people like you and our sponsors:
  Sonic Craft  

Follow Ups Full Thread
Follow Ups
You can not post to an archived thread.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2026, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]