MQA from an NDS

MQA is licence locked so you can’t transcode it. It’ll play as a slightly hobbled 16bit track unless you have an MQA DAC.

I still have Tidal (for now) but use Qobuz Studio so I get proper Hires streaming.

I am pretty much with Hmack here. I trust my ears. Afterall its my pocketbook. This is why I purchased a Mytek Manhattan II exclusively to try MQA. If it does not work out, I can always fall back on my ND555. I don’t exclude progress and that there is a better way to reproduce music. I look back at an audiophile system of the 70’s: Marantz 8B, Marantz 7C, Thorens TT, Shure V15, and Klipshorns. When I started my journey, that’s what everyone pined for. Today, we have come a long way. Whenther MQA will be another step up from the current SOTA remains to be seen. I hope it survives or dies on its sonic merits, and not the opinions of those who have not heard or tried it. My hats off to Bob Stuart for this brilliant work. Hope it takes me a step closer. The Mytek is scheduled for delivery on 4/22.:grinning:

I will let you know.

1 Like

Would a subscription to Qobuz for true 24bit material not be cheaper?
I could be wrong but the studio master is more likely to be 24bit rather than MQA??

Would a subscription to Qobuz for true 24bit material not be cheaper?
I could be wrong but the studio master is more likely to be 24bit rather than MQA??

Does the ND555 do Qobuz? The NDS doesn’t. I can resort to the old fashioned way: HD Tracks to a memory stick then plug into my NDS.

Ive also ordered a Mytek Brooklyn Bridge and I’m going to try bailyhills method and also use it as a separate streamer, exclusively for MQA. Im interested to compare it to the NDS for non masters Tidal. I am also having a super-dooper power supply built by Sean at CHC to maximise the SQ. You never know, it might be a NDS beater!

The new generation streamers support Chromecast so you should be able to play any streaming service.
For first generation streamers you can use Bubble Server to act as a proxy for Tidal and Qobuz. Better SQ than streaming direct too; all transcoded to WAV for another boost in SQ.
:blush:

Hello Guinnless

Its not a matter of the maximum bits in this case. Hi Res Audio has chosen to markedt the bit rate, as that’s what is easy to understand. For me, its a little bit deeper than that. MQA corrects for the A/D converters and other issues with the Record Process. Something that comes thru in normal Hi Rez. It also does some things to the D/A converter in the decode process to eliminate some phase errors. It seems that most critical listening reviews that I have read say that MQA is better is many cases and at least no worst that Hi Rez. I don’t know if this is all true, but I do understand a little bit about digital systems, having done graduate school work in digital torque loops for high performance gyroscopes back in the day, when the conventional wisdom said that it could not be done. My Thesis was on that subject and resulted in 4 patent-able ideas about what needed to be done to optimize these digital loops. We have been flying defense systems with these designs for 20 years. A different problem, but I did may classes in digital theory. So I could never have done what Bob Stuart has done, I can follow his process and its consistent with everything I know about digital theory.

As for Qobuz being cheaper, well I am well past reasonable cost in my hi fi system, (ND555/PSU555, Naim Core, Townshend DC 47 Autopreamplifier, Apogee Scintillas, Krell KSA 80B, Inakustics P3500 Conditioner, Townshend Podiums and Townshend f1 Fractal interconnects. Working on a 40th anniversary update of the Apogees with all SOTA components.

Hello Russ
Interested to hear how your Mytek Brooklyn Bridge for MQA works out. Keep me posted.

1 Like

Clearly some have bought into the MQA snake oil. Why anyone would wish to support a lossy, DRM nobbled, license-bound format is beyond me…

1 Like

How about SQ?

There are more important issues like sheer enjoyment of music. Haters gonna hate i guess…

Instead of that how about unlossless for MQA, that way we could have doubleplus-unlossless for MP3

Mike

According to a Bob Stuart video on Mytek’s website MQA is lossless.


2:30 seconds in.

Mytek also state
“MQA is the only technology today which provides a practical way of streaming high resolution audio.”
Utter :ox::wc:

Hmm I can happily stream 24/192 from Qobuz…:sunglasses:

‘its a little bit deeper than that. MQA corrects for the A/D converters and other issues with the Record Process. Something that comes thru in normal Hi Rez. It also does some things to the D/A converter in the decode process to eliminate some phase errors. ‘

That just is not the case, unless the very original encoding low pass filter exactly matches the output DAC reconstruction filter.
In music these days, many instruments also have their own ADC converters or are natively digital so for this ADC and DAC matching won’t benefit here.
In the recording industry I understand there are a handful of ADC types used and effectively replay DACs match to these in their own differing ways. End to end chain matching is a nice idea for the lab, but in practice means nothing.

All reconstruction low pass filters (or any filter) will produce a phase shift, the exact phase response will vary with filter type. The aim is to have a linear phase response in the pass band, and finite impulse response filter types generally have better phase characteristics compared to infinite impulse response types. Many DACs use FIR reconstruction types (such as Chord Electronics DACs with a finite filter kernel of a given number of samples or ‘taps’), but Naim have decided to use IIR filters, as they believe in paractice with their architecture that produces the best sound.
So in short the approach of phase handling at reconstruction can vary by different DAC implementations and manufacturers so as to produce the best subjective SQ performance of a given product.

Interestingly the MQA reconstruction correction filter is relatively quite complicated so as to reduce the audibility of MQA encoding errors introduced into the original audio… Here is an example, and will vary on how the MQA encoding has been optimised with respect to the original media content so as to produce the most subjectively pleasing modification of the original sound. [the red line is the MQA encoding downsampling filter, and the black is the MQA composite reconstruction filter response]

http://www.audiomisc.co.uk/MQA/origami/Fig6.png

Depends on your definition of sound quality…

One thing is very clear, and that is that people’s preferences in sound differ. Some prefer the music replay system to be as neutral/transparent as possible, conveying the recording as released by the recording artist (which of course whoever is responsible for the mastering). Others prefer the sound modified in some way, whether that is emphasised rhythmic elements, tailored frequency response (e.g.curtailed or emphasised low bass), ‘colouration’ etc etc. In practice no real life system, which inevitably includes an encoding process (analogue or digital), various electronic components, speakers and a listening room, will have absolute fidelity to the recording. Choices of equipment sensibly aim to get the closest to whatever the individual perceives as the right sound, wherever on the spectrum of fidelity that lies.

MQA loses some of the information in the original recording, while in the full process it reconstructs according to a certain set of parameters. That overall process will have a sound signature - and, in the context of the foregoing paragraph, the result might be something that some people prefer compared to the sound of the original recording. My perception is that this is the reason why some people do apparently like the sound of MQA, whereas those who prefer to hear the original recording as was may prefer to play the original unadulterated recording.

For those preferring to hear the original recording, presented only with the character imposed by their existing choice of system (Naim or not), MQA may be an anachronism. Of course no-one actually knows whether they like it unless they hear, and the outcome on hearing could go either way - some people may love it, others hate it. Before spending good money on enabling the full MQA process when you already have a system playing the way you like but which is non-compatible, I would suggest that it would be wise to do a comparative audition (of course in the same system).

To my mind MQA as promoted by Tidal is simply a means of them increasing their profits by reducing their online data bandwidth, for which they pay, compared to simply offering the original unadulterated files for online streaming.

1 Like

Nice to see an articulate and perfectly reasonable and rational argument against MQA rather than the unfortunately rather frequent rants or attempts to ridicule those who don’t conform to the anti-MQA orthodoxy.

I pretty much agree with virtually all of your post, and in particular with your first paragraph. Most of us make subjective choices in respect of our individual preferences when we purchase hi-fi components, rather than rely purely on specifications. Very often we end up with components or systems that sound very different from those of others who have gone through the very same process. We can’t all be ‘right’, and those of us who approach music and enjoyment of music reproduction logically without prejudice or dogma have to accept that there is no absolute ‘right’.

There are some people who just cannot accept this, and who adhere rigidly to their narrow perspectives and prejudices. Perhaps to them it feels good to belong to a cause, whatever that might be. I have no axe to grind with those who argue that MQA is not technically perfect - they are absolutely correct. I have no axe to grind with those who believe that MQA may be an attempt to stealthily introduce DRM and take over the music download world. I don’t agree with this point of view but who am I to say that their position is any less valid than mine. However, I do have an axe to grind with those who criticise MQA and those who enjoy listening to MQA Masters simply because ‘Bob Stuart is evil’ and ‘MQA is the work of the devil’ or those who criticise the sound of MQA files without ever having listened to them on appropriate equipment.

I agree that MQA is promoted by Tidal primarily as a means of increasing their profits or minimising costs in an attempt to compete in an increasingly competitive market. I have a lot of respect for Qobuz (I purchase a lot of my downloads from them) and I very much hope that they too succeed in this increasingly competitive and difficult environment. I would happily use Qobuz for streaming myself if their music catalogue was more in line with my musical tastes.

Your argument that the MQA process introduces a specific sound signature that those of us who tend to enjoy MQA masters find ‘pleasant’ is one that I can happily accept and with which I have no argument. My own position is that I generally (but certainly not always) find that I prefer Tidal MQA Masters to their 16 bit equivalents on Tidal. I have never made any claims (or at least I don’t think I have) that Tidal MQA Masters sound ‘better’ than their ‘standard’ hi-res equivalents.

I think that your argument could also be said to apply to brands of hi-fi equipment such as for example Naim Audio or Linn (from the LP12 era of the 70s and 80s). These were and are brands that attract many enthusiastic followers, arguably largely because of that ‘exciting’ house sound, but are disliked by others specifically because that ‘house sound’ is perceived by them not to be as ‘accurate’ as that of some others.

We all have different subjective tastes in music and music reproduction, and the hi-fi industry (for the time being at least) provides a wealth of different possibilities. Were this not the case then our hobby would be a little easier to pursue but much more boring.

2 Likes

So Roon Core -> SonoreUPnP Bridge -> NDS
This get you MQA playback of locally stored MQA or Tidal Masters to 24/96 with 1st unfold or Qobuz streaming upto 24/192
No need for new DAC or converting to S/PDIF.
Once you have the Bridge working fine, SQ is identical to any UPnP based source, plus you have the Roon interface.

This forum is quite tolerant of different results with different speakers, rooms, interconnects and ears. It seems normal for one thing to work for one and not for another person. In fact, even some Naim kit works for some but not for all. The recent dilemma over upgrading with a 552 or a 555 is an example. This is very professional and shows collegiality. It seems that everyone lets their ears decide.

What surprises me is how the topic of MQA does not get the same love. I can understand that the technical details push the envelope, and it is almost unbelievable that such a process can sound good. But we see positive reviews and positive responses about the SQ from well respected forum members. I would be perfectly happy if folks posted that they auditioned it for a week, and the sound of standard kit was preferred. Its clear that most folks have not auditioned it–particularly those opposed.

I am curious, so I am going to do just that. The technology excites me–having a background in digital system design. But I have been sold a bill of goods before, so I am going to audition and experience what there is to hear.

1 Like

1st unfold doesn’t give you the real experience of streaming mqa. So i completely understand Bailhill on wanting to add a dac with mqa decoding. If he has the budget to do it, why not.
I find it a bit disappointing from naim, which produces ones of the best streamers on the market, to not be able to stream qobuz, nor tidal mqa, nor true dsd. Specially with a streamer at more than 21k.
And to end this nice story, when the nd555 will be replaced in some years, the owners will have to throw away around 7k in selling it.

The first oversample decode (unfold) does provide effective decoding of the MQA initial higher frequency spectrum elements (approx between 20kHz and 40kHz) and stacks them onto the low frequency element of the encoded single. The next oversample decodes (unfold) adds and stacks the next spectrum block of encoded frequencies onto the precious stack. (Approx above 40 kHz)
The encoding filtering to reduce the audibility of encoding noise in the baseband signal is applied.
Therefore the difference between the first and second decode (unfold) is the degree of temporal and frequency information, similar to the difference between 96kHz and 192 kHz sampled material.

I wouldn’t say the experience is different, just the potential decoded resolution.
In other words the first unfold of MQA is fully realising the MQA experience, what it is not doing is not extracting the full potential information encoded into the signal… this info might be effectively absent or be rich, depending on the MQA encoding parameters and the original source PCM.

It might have got more love if it was open source and there wasn’t incorrect or misleading information on websites of vendors of MQA DACs.