MQA disappointing

I agree, Chris. Any views from me on MQA are relatively worthless until I get a full-MQA capable DAC. By no means have I written MQA off, and in fact I still hope the hype is warranted. I certainly do not accept the pejorative judgment of ulterior motives on the part of those who rate it highly. It kind of makes me wonder if it was a smart business decision by MQA to allow “first unfold” in streaming, since perhaps many people like me were disappointed, and many of them may have come to premature conclusions about it, rather than waiting to hear full implementation of it. And sure, more new, less-affected older music is totally welcome.
Jim Heckman

This is the MQA is dissapointing thread why would he need to move on?

4 Likes

Actually it is the case. Every method of converting a digital signal to analogue has issues. No one method is absolutely without compromise. And we often disagree about the best way to do things because what might suit acoustic jazz might be wrong for hard house. There is nothing wrong with pointing out the shortcomings of MQA but other methodologies have their shortcomings too, just different ones who’s impact on the end result varies.

Not sure that inaudible high frequency unfiltered ghost images are really that bad. It depends on the amplifier and how wide of a bandwidth it has and how stable it is. May be it could damage tweeters in some speaker designs?

The mathematical trick with upsampling is that if you simply add zeros in between the actual data points then you keep pushing the ghost image higher up in frequency. 8 x oversampling pushes the ghost right up to extreme high frequencies that are easily filtered out even with a smooth filter…

Aliased images from A to D are another matter - these fold directly into the audible band and are terrible.

I prefer Linear Phase filters just like all the audio industry has done for 60+ years. I am not worried about pre-ringing as it really isn’t an issue in consumer audio playback at all. Preringing is something you get from very high Q or sharp filters used in the studio or mastering. Bob Stuart and his manufacturing supporters (all with vested commercial interests) have managed to fool consumers in to buying something they don’t need to fix a problem that doesn’t exist. In the process, they have changed the phase relationship of the music and therefore changed the timbre. It is actually a process that is probably detrimental although since it is a black box we don’t really know for sure. That some people like the sound of MQA isn’t surprising at all - there are countless different sounding amps and speakers and nearly all are liked by some people - many folks like compression in rock and pop and yet this definitely can ruin the sound quality…so if some folks like MQA it is meaningless and guarantees nothing about the quality.

MQA does de-blurring filters both after ADC and another one at DAC; it is end to end process. Especially PCM 24/44.1/48kHz masters when converted to MQA, aliasing becomes more apparent below 20kHz audio band.

Obviously the higher the sample the better is it. As it sample higher such 88.2/96k above, the aliasing image will move away from audio band. My argument is those PCM 24/44.1/48kHz masters when playback on a linear phase filter, the effects of aliasing distortion is virtually non existent. This is not the case with MQA.

1 Like

So if I understand correctly you are saying that MQA has significant problems with audible aliasing noise as a result of their process. I can only imagine what deblurring is. Perhaps they are running all the original master files through a deconvolution filter? Perhaps they are just dithering again after truncation of the last 6 bits that are subsequently used to store the higher sample rate low bit depth unfold information.

Whatever MQA does it is lossy and technically cannot possibly be as good as a complete original full 24/192KHz file. If folks like the sound then it is just another form of audio manipulation such as THX or Dolby processing available on many DSP processors.

I know Mastering engineers that have had much of their work turned into MQA files and they never spoke to anyone about the process. Nobody asked them what equipment was used on the original mastering. There is no way to place information on a PCM file that indicates the equipment used (unlike photography files which record camera model and lens and software versions).

So claims by MQA that they correct the original masters for the effects of the Mastering A to D quality (deblurring) are patently false!

3 Likes

if MQA only looses bits that have zero musical information and if record companies keep good records of the equipment used in the a to d processing and mastering then they can correct for the equipment used and technically it can sound just as good or even better without any reference to the original mastering engineer.

MQA also has techniques for dedecting what kind of equipment was used even if good records were not kept by the company.

What are the albums your friends mastered that have gone through the MQA process? Let us know what they are so we can listen and hear if the sound better or not.

1 Like

How should this work? Sounds like magic

1 Like

Let’s keep the role of mastering engineers in context here. They don’t own their work. They are employees, albeit highly regarded ones in some circumstances.

1 Like

Yes technology can seem like magic to many :slight_smile:

You can get some idea of how this can be done from this, though in this case this did have access to the original ADC.

And also in the answer to questions 18 and 27 here.

Though no doubt this will all just be fairy tales to you :slight_smile:

1 Like

MQA lossy compression limits the bandwidth to around 1.5Mbps. It does by doing band split from actual 24 bit PCM into 16 bit and use to code from 0-22.05k while remaining 8 bit is used to code from 22.05-44.1k bandwidth, assuming 88.2kHz sampling. In another word, MQA only provides 8 bit of resolution above 20kHz while a 24 bit PCM provides 24 bit resolution up to 44.1k bandwidth assuming 88.2k sampling.

This is only true if the 24 bit PCM master is 88.2/96k or higher. If the 24 bit PCM master is 44.1/48k, then MQA can provides 24 bit resolution (lossless) up to 22.05k bandwidth assuming 44.1k sampling.

The maximum MQA core sample output is always at 88.2/96k irregardless whether the PCM masters are 44.1/48k, 88.2/96k, 176.4/192k or 352.8/384k.

2 Likes

Are uou sure that “converting methot” recognizes music mood.
Don’t you thing that your conclusion is a little bit too radical?

Agreed. My point was not that Mastering engineers are special but they might know more about the origin and processes that generated the final file. So to ignore them and batch process MQA and then claim to create some deblurring and authenticated files is a bunch of pure baloney.

The mathematical reality is that the inversion process to create the original music prior to A to D and everything else (mix and mastering) is non-unique. Without knowing exactly what was done to create the original file there is no way to re-create it. Even knowing exactly what was done it may still be impossible to create the original - for example if the mastering engineer notch filtered certain frequencies then there is no way to recreate them with any degree of precision. So the whole master authentication process is just make believe.

MQA is just another form of Dolby - a lossy process.

3 Likes

This Q&A is not science but marketing BS. There is nothing that is properly explained. This is a bunch of hand waving. The graphs are meaningless.

The premise that we hear analog waveforms (like you can draw them or plot them out) is completely wrong. The entire foundation of MQA is based on this totally false assumption. Since the whole foundation of MQA is false or wrong then the whole effort applied is worthless. We categorically do NOT hear analog waveforms. We hear probably less than 1/10 of the audio in most music as has been proven by MPEG. Furthermore we hear harmonics more than anything else. Our ears brain actually behaves in a rather digital fashion. Our hearing is limited in bandwidth to about 20 KHz and therefore we can’t begin to possibly hear a analog waveform with subtleties up to 96 KHZ - we just pick up the low frequency content. Our hearing is limited by loud low frequency sounds that can totally mask higher frequency sounds for relatively little long periods - we cant even begin to hear more than 1/10 of most musical audio waveforms.

2 Likes

No. The concept of inverting errors when you know the process is not controversial at all. It is routinely done in Photoshop and Lightroom, for example, compensating for limitations in cameras and lenses at the source end, and for ink and paper at the destination end. I do it every day. And it is increasingly done right in the cameras, up to and including Leica.

The fact that it may sometimes be challenging, due to missing documentation or problematic process, is well recognized.

But that doesn’t invalidate the concept or turn it into make believe.

What percentage of recordings is amenable to correction, and what percentage is not, would be an interesting thing to know. But it’s easier to make blanket statements.

2 Likes

Filters have different characteristics that can align themselves with musical presentation. That might be vitally important for an acoustic piece where you are preserving timbre and timing but less so for electronic synthesised music with a beat.

I fixed your quote for you.

AJ

Agreed about photography, the software camera file even has the information about what lens etc. was used. A PCM file tells you nothing about the equipment used to make the recording…so how can you correct for the unknown?

It is the deluxe box set remaster of LZ from 2014 that you like so much. MQA is not as important as what was done in the remastering 4 years ago.

It is reasonable that some studios would have that information archived. It is also reasonable that some would not. So it is a practical problem, and the real world statistics would be interesting.

But it is certainly not inconceivable.

And nobody has said with authority and conviction that virtually no studios have the info.