MQA

The growth in market share is about 30%. Jumps from 4% to 6%, 6 to 8.5% share.

Darko needs to be friends with everyone, including Schiit and Meridian. Otherwise no more DACs to review.

P.S. I might write a SBAF position on MQA.

 

That is true, but I happened to be looking at the nature of the wavesforms, and how the bits don't change as much and get smooshed at the crests of the peaks. Compression tends to squish stuff up to the top and bottom leaving very little in the middle.

 

As far as I can tell, MQA is bullshit.

It's just another random lossy format that comes at a premium.

I do not see the reason for it's existence, from the customer's perspective.

 

It comes down to this. What is the problem MQA is trying to solve?

1. Given what humans can perceive, I can see a plausible argument that more bits than 16 is necessary. Maybe 18. Let's say 20 to make absolutely sure.
2. Given what humans cannot perceive (assuming proper sampling and reconstruction), I cannot see a argument that we need anything past a sampling rate of 48kHz.
3. Therefore the proper solution to solve the quality problem would be 18-20/48 PCM or for the sake of simplicity 24/48 PCM.

This brings one important question:

1. Why does MQA unnecessarily encode 24/96 (the stuff past 24kHz that we cannot hear) into 24/48 when lossless 24/48 PCM would have been absolutely fine (and actually superior because the least significant bits are not used for encoding stuff we cannot hear) as a end-user playback format.

 

Because money.

U need Meridian Quandary Audio ~does Obi-wan hand waving~

 

The quantisation error, the way MQA is describing it, seems more than 0.5 LSBs. Like way more depending on the frequency band and whatever other random stuff they are doing.

Vinyl is another story. The noise in there maybe of a different nature overall than the cluster f**k the Meridian guys are describing.

As a matter of fact, I'm not sure I fully understand what they are doing. First they mention u can put this in a lossless container, and throw the "lossless" key word here and there. Only to later find that they are changing quantization levels, resampling ... there is no way in hell that's lossless. Then they fold over the high frequencies into the low frequencies using some envelope thingy, followed by some residual deal. Add more "lossless" and "audiophile" keywords. Now put that together, shake it, add water and done. Now pay us for it. WTF!

I'm not sure what we are getting at with correlation. I can have a tone which may correlate poorly with the input audio signal.

 

How many molecules between peaks and troughs of a groove? 65536? All discrete steps?

 

AFAIK quantization noise can be correlated or uncorrelated. Noise is basically stuff added to the intended signal. Quantization means you have a finite precision to represent something. Quantization noise comes from reducing precision.

 

^ Where I was getting to, but explained better than I could. Analog's source of noise is inherently different.

Also, vinyl deforms to a metal stamper. Metal atoms and molecules are probably orders and orders smaller than long polymer chains.

 

And this is why MQA will fail. BD (BluRay Disc) has essentially failed even though is backed by all of the major studios. 4K isn't anything people have asked for (it is ridiculous in terms of storage size requirements - you can even argue 2K uncompressed is better than 4K compressed.) In fact, all the movie studios, major and minor are failing big time in the post theatrical (or TV) release home entertainment market. Right now, its easier to steal content than it is to buy it. In some cases, if you want 720p or 1080p content, you have to steal it.

 

Again, that doesn't take into account that vinyl is shaped by a stamper. It's not like bits of the plastic molecules only come out of the stamper in discrete chunks one by one resulting in consistent discrete steps. The polymers get stretched and compressed.

Back on topic, I think SBAF should create new high quality playback standard, free of charge, and fully compatible with most existing playback gear and even playback containers / lossless compression format.

I propose this: NWA - No Wanker Audio

Specification: 24/48 PCM

 

That's my opinion too. People are hearing differences on how the DACs internally handle higher sampling rates. I took a DSD128 file (encoded with one of those Korg ADCs, the ones that are now discontinued) and downsampled to various formats using Adobe Audition, which is supposed to have a very good resampling algorithms. I heard no difference between 44, 88, 174 sampling rates. I did hear a difference between 16 bits and 18 bits, but not 20. I used the Vega DAC for this comparison. I didn't use the PWD2 I had back then because I knew the PWD2 treated material with different sampling rates differently.

As far as "HD" terminology, how does one explain why "HD radio" sounds significantly worse than regular FM radio when there is good FM reception?