Home Computer Audio Asylum

Music servers and other computer based digital audio technologies.

RE: Can you pl

Unfortunately for you, I had read the paper and I agree with fmak regarding this work.

I come from the perspective that the cost of distribution and storage is now so cheap that there is no purpose to be gained by reducing the quality of recordings for distribution to listeners who have high quality playback systems. When purchasing recordings for download I always select the original quality of the recording if known, or barring that the highest available quality. In particular, I never purchase 44.1 kHz recordings made from higher sampling rate originals if there is a choice. (In this regard, analog recordings, even cassette recordings, effectively have a higher sampling rate, as can be determined by listening, or for those who want to be wowed with curves and plots by arguments such as those in the first paper.)

The psychoacoustic problems discussed in the MQA papers are all caused by the inadequate sampling rate chosen by Sony and Philips for the CD. These problems are fixed by using a higher sampling rate, and the closer to the original sampling rate the better. All that MQA has done is to use clever encoding tricks to steal some noise performance (in the band up to 22 kHz) and use the bits that are saved to reproduce some of the higher frequency information that is needed to avoid filtering artifacts. As such, it can do nothing to improve the quality of an original high resolution master that is transmitted in its original form to the user. All it can do is to reduce the bits that have to be transmitted and stored. However, these bits don't cost much. Most downloads fit in 1 to 2 GB per album, even in hi-res formats and these amount to a bandwidth cost of $0.10 to $0.20 USD. This is less than the credit card processing costs involved in purchasing a download. Thus there is no real economic benefit to be had from using this psychoacoustic trickery. However, there are real risks involved in pschoacoustic trickery, namely that a CODEC that was deemed to be transparent one day is later found to have audible artificts. Once noticed and pointed out, it can become obvious to many listeners that the "perfect sound" wasn't perfect after all.

I have mentioned this before (maybe not in this thread) but I studied the issues of PCM formats and transparency extensively before doing digital remastering of analog recordings. I spent some time looking into some of the previous Meridian work on DvD audio and looking at the various tradeoffs and encoding tricks used therein. I also spent over 100 hours listening to various A-B encodings comparing various conversion techniques, involving all manner of bit depths, filtering algorithms, etc., eventually concluding three things: 1. the 44.1 kHz rate was not transparent and could never be and 2. it would be possible to work within or very close to the CD bit rate and get most of the benefits of higher resolution digital formats and 3. there would be no point in doing this, because it would be easier to just supply the higher resolution format in the first place to any critical listeners.


I also investigated the range of tradeoffs involved in filter design, before rejecting the 44.1 sample rate. I concluded that it was simply impossible to achieve transparency (for most recordings) at this sample rate, with tradeoffs involving some combination of impared tonal accuracy, high frequency "air", imaging specificity and sound stage depth. These subjective qualities can be directly related to filter parameters, e.g. particular configurations of the iZtope 64 bit SRC filters used to downsample or upsample out of 44.1 kHz. Like Stewart, I concluded theoretically, measurably and audibly, that the most effective upsampling filter used to play a 44.1 kHz filter depended on the particular downsamling filter used to create the 44.1 original. However, the choice wasn't particularly critical, and can be readily made by users of DACs or computer audio systems such as HQPlayer on a recording by recording basis. Also, even if the best choice was made, the result was still unsatisfactory for most program material that was well recorded, even if it came from a high quality cassette tape, let alone a 15 or 30 IPS master tape or live microphone feed to a high rate ADC.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar


This post is made possible by the generous support of people like you and our sponsors:
  Atma-Sphere Music Systems, Inc.  


Follow Ups Full Thread
Follow Ups
  • RE: Can you pl - Tony Lauck 01/24/1613:10:32 01/24/16 (0)

FAQ

Post a Message!

Forgot Password?
Moniker (Username):
Password (Optional):
  Remember my Moniker & Password  (What's this?)    Eat Me
E-Mail (Optional):
Subject:
Message:   (Posts are subject to Content Rules)
Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:  Automagically notify you when someone responds.