Stereophile Tests of MQA

[amirm]

This rose up in the Tidal thread but merits its own thread:

In this month's Stereophile, in the back pages (p127-129), Jim Austin is doing a little more digging into MQA.

This month's shovelful looks at the impulse response, and promises more to come.

---

I see it is posted now:

https://www.stereophile.com/content/mqa-tested-part-1

 

[amirm]

Can anybody with more knowledge than me comment on what Austin writes on impulses etc? Is it correct? I struggle to understand it, and my experience is that lack of clarity in writing reflects lack of clarity in thinking.

There are a few things in there. Let's start with the basics.

They requested an impulse response in digital form for the MQA system. An impulse is a pulse that goes from zero to "one" in zero time and returns back to zero. In frequency domain, it contains all frequencies. As such, how a system interacts with it, defines how it really works.

The thing to remember about an impulse response is that it is theoretical. In practice it can never be generated because it requires infinite energy and infinite bandwidth. Such a thing does not exist in nature.

An impulse response is frequently used to characterize a filter and that is what is done in that article as the introduction. The file itself is artificially generated using a computer program. This makes it easy to go from zero to maximum value since these are just numbers we can create and put in the file. When we do this and apply it to classical reconstruction filters in DACs, we get a graph like he is showing in the article:

So instead of our one large peak, we also have smaller ones before and after it. This is mathematical and cannot be avoided when using this type of filter. It is what the filter does.

Immediate concern is raised that this should be creating audible artifacts. The "ringing" after the impulse is considered benign since it is temporally (time based) masked by the large peak and hence not heard. The ones before it -- called pre-echo -- are considered harmful since they come before the main pulse so nothing masks them. This is at the heart of the argument for "apodizing" filters and those used in MQA.

There is a problem here though. Such an impulse created in digital domain has far more energy than it would if we were capturing analog signals. That is by definition so since we limit the bandwidth of any sound we capture to half the sampling rate. In the case of 44.1 Khz sampling for example, anything above 22.05 Khz would be "illegal" and won't be in the file. An impulse then captured from analog would not have anywhere the same amount of energy/high frequency content and with it, the pre-ringing shrinks substantially. This is NOT covered in the article and should have.

The article digs into other filter types that don't have pre-ringing which is common in many DACs such as slow roll off filters:

The problem with these filters is that they don't do what they are supposed to do, i.e get rid of all the frequencies above half the sampling rate. By not doing this, they allow those extra, unwanted signals to wrap back into the audible band. Such is also the case with MQA but not addressed yet in the article.

So this is the intro. I will dig into the MQA part next.

 

[amirm]

Next they asked MQA folks to create a digital impulse in MQA format for them. This is odd as they kind of talk about in the article in that MQA's claim to fame is that it corrects for some ills in analog to digital conversion. A file that starts in digital form, i.e. an impulse, doesn't go through that.

Anyway, the run that impulse through a Mytek DAC and get this:

So pre-ringing is eliminated and post-ringing is pretty short. In that sense the mission they set out to take part in, has been accomplished.

The "surprise" bit is that they ran the non-MQA impulse with the Mytek and got this:

In other words, the same output can be generated in the decoder whether the input was MQA or not!!!

In later text and comments from Bob Stuart it says that the Mytek DAC that was used here must be subjecting all input to MQA treatment once that mode is on, whether the input is MQA or not. I can attest to this as in my review of Mytek it clipped the signals and I was told that was because of MQA pipeline being on even with non-MQA signal.

 

[amirm]

Here is a quick summary of all of this: the MQA filter in the DAC does produce what they say, that is, no pre-ringing and short post-ringing. In that sense, the claim of "better time resolution" by MQA folks is demonstrated.

I originally said this article is incomplete. That is because in the next breath they should have measured what happened to the cut off frequency and how much aliasing was let through. This is mentioned in words but not in measurements. It is like they stated the positive but left out the negative. I find this very odd. The purpose of an impulse response is to fully quantify a digital filter and very important part of that is its frequency response. We can't leave that out this way.

 

[RayDunzl]

In other words, the same output can be generated in the decoder whether the input was MQA or not!!!

Uh...

Something doesn't make sense to me.

If an Impulse (in digital) is a single full-scale bit, how would that be different in MQA or PCM.

I can see where the output would be different, but the input?

I'm trying to imagine what the "origami" would look like in a 48kHz file containing a 192kHz Impulse (single full scale bit).

Ok, I fail.

 

[amirm]

Something doesn't make sense to me.

If an Impulse (in digital) is a single full-scale bit, how would that be different in MQA or PCM.

It would have to be different because the MQA encoding embeds signals in the (clear) PCM stream. So the impulse is no longer what it was.

That said, they showed a clear phase change of sorts so more than that is going on.

 

[oivavoi]

Thanks Amir! Now I understood it!

How come you're able to explain this so much better than the stereophile guy?

 

[amirm]

. He was doing OK at first but then became obscure enough that I had to read it a second time to see what he what he was saying.

 

[oivavoi]

. He was doing OK at first but then became obscure enough that I had to read it a second time to see what he what he was saying.

Yeah, same here. I understood nada after the first part. Usually not a good sign.

 

[oivavoi]

And concerning MQA then, it seems that what MQA does in a dac is just to apply a specific type of digital filter? Which can be either good or bad, based on the sum of positives and negatives. And from your explanation, it achieves one good thing at least, but probably does bad things as well, that they've conveniently left out so far.

 

[amirm]

That is not the only thing. The DAC also uses signals that tell it how to recreate the ultrasonics up to twice the bandwidth. And if the source was 4X higher, it forces yet another upsample.

 

[Wombat]

I needed some background to get a basic understanding of MQA.

This article endeavours to relate available information re MQA, in early 2015, to existing audio processing methods.

https://www.xivero.com/de/hypothesi...per-technical-analysis-of-mqa-by-mqa-limited/


An early overview which opines that control of distribution(watermarking) and embedding a DRM protection system may be a driving consideration:

https://archimago.blogspot.com.au/2015/01/musings-miscellanies-on-audio-encoding.html


I look forward to seeing what the 'magic' is or isn't.

 

[watchnerd]

One of the cool things about the RME ADI-2 Pro is that it lets you pick different digital filters. I think the TEAC DACs that use the same AKM chips offer similar choices.

Here are the selections amongst the conventional options:

Hmmm, "Short Delay Slow" sure looks as good as the MQA version.

What do I need MQA for, then?

And then we have this choice:

Even better impulse response than MQA! Nearly ideal!

Bob Stuart needs to be told that if impulse response is the most important thing, MQA sucks compared to NOS.

 

[Purité Audio]

I must experiment with those filter options, have you seen RME are about to release a cheaper ‘domestic’ version of the ADI Pro , no A-D but it is cheaper £1000 and it has a remote!
Keith

 

[watchnerd]

I must experiment with those filter options, have you seen RME are about to release a cheaper ‘domestic’ version of the ADI Pro , no A-D but it is cheaper £1000 and it has a remote!
Keith

Yeah, the remote would be a pre-req for me to use it in my living room, but the lack of analog input kills it for me since I also have a TT.

 

[Cosmik]

Hmmm, "Short Delay Slow" sure looks as good as the MQA version.

I would say "...as bad as the MQA version".

 

[watchnerd]

I would say "...as bad as the MQA version".

Is it the phase issues or the high frequency roll off you dislike?

 

[Cosmik]

Is it the phase issues or the high frequency roll off you dislike?

Both - i.e. the fact that it is not correct.

 

[Purité Audio]

Yeah, the remote would be a pre-req for me to use it in my living room, but the lack of analog input kills it for me since I also have a TT.

I have mentioned ( many times) that a remote version of the ADI PRO-2 woud be superfragalistic ...
Keith

 

[RayDunzl]

What does the Impulse Response of an Analog (parts bin) low pass filter look like?