Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
In Reply to: RE: MQA is a 'lossy' codec... posted by SteveJewels on January 23, 2020 at 04:19:13
you do not understand how the lossy signal is processed and later "padded" to appear 24 bit.
Would you explain please?
It's very clear how the signal immediately becomes lossy.
I started reading the post in the link.
"There is no question that MQA degrades the quality of the audio for users who do not have an MQA decoder."
Not a particularly scientific point. Like saying your car will not run well without wheels.
It sets a poor tone for the article. I hope it gets better although the author seems to have an axe to grind rather than presenting a scientific case for his position.
Before we get too far into this, please allow me to state my position.
1. I have a turntable and and rather nice cartridge and phono pre. My preference is for analog over ANY digital source I am aware of so this discussion is for what will be the second tier for my system.
2. I work in electronics/electromagnetic compatibility but I do not have a complete understanding of digitization.
3. About 1 person out of 10,000 or so actually understands the Nyquist Theorem.
4. I have Tidal, a Mytek Brooklyn DAC +, a Meridian Explorer 2 and an Onkyo DP-X1a all of which support MQA.
5. If I get to a point where I can discern that MQA adversely affects sound quality more than other digitization/digital storage/streaming schemes adversely affect sound quality, I would abandon MQA.
About 1 person out of 10,000 or so actually understands the Nyquist Theorem.
What does that have to do with the established fact that MQA is a lossy format?
Pick you own poison. I choose lossless in this modern age of inexpensive storage. Why throw away anything from the original recording?
I mention Nyquest in an attempt to bound the discussion to reality. Inevitably someone will come along and say "Well Nyquest says...." and proceed to incorrectly cite Nyquist.
ALL digitization schemes lose some of the original audio content and the overwhelming majority of discussions of digital music are blind to this fact. The terms "lossless" and "lossy" are misleading. NO digitizing method records 100% of the original music.
In my work I use a 10 Gs/sec digitizing O'scope to look at 1 MHz waveforms. That is 4 orders of magnitude greater than the original waveform. I find it to be sufficient for my work even though some portions of the original waveforms are lost. To compare that to audio, the harmonics of some instruments can extend high 10's of kilohertz. Ignoring the mixing that will occur, to sample 4 orders of magnitude higher than 100 kHz would require a 1 Gs/sec sampling rate. I am not aware of any digitization method used for audio files that high.
Am I asserting that MQA is "good", or "sufficient" or excellent"? Nope. I am asking questions and participating in a discussion and appreciate your input.
The terms "lossless" and "lossy" are misleading.
Perhaps to you. It simply means you cannot return the recording to the originally captured signal.
I am asking questions and participating in a discussion and appreciate your input.
Lossy formats were useful ten years ago. I stream 192/24 losslessly via Qobuz and store similar formats on my NAS. Further, all the signal processing in the MQA chain changes the tonal balance.
I share mastering engineer Brian Lucey's opinion :
"Let's just sell the 24 bit files at the mastering session sample rate, not higher and not lower, and call it a day?"
Yes, the original recording cannot be completely recreated from a digital file, be it "lossy" or "lossless".
using bit perfect codecs.
MQA cannot even approach that dithering 24 bit content to 17. Why choose to arbitrarily throw away anything ?
To each his own I guess. My approach is the diametric opposite - my system building steps have been to retain as much information found in recordings as possible.
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: