Is MQA good enough to get Analog Guys to enjoy it? Or still cause Digital Fatigue?

[EuroDriver]

So it is not only upsampling and filtering - HQ has another type of processing. Do you have details on the operation I quoted in bold?

Hi microstrip,

Thank you for asking the questions I can not answer
Jussi told me that this is what he does in his software. He used to say 18 months ago, the artifacts I can find, now he says he finds all of them !

An interesting anecdote from the video field which is not comparable or similar to audio, but just an example of what it possible with clever algorithms and brute force computing can do. A problem that Jussi and his work colleagues worked on was to remove the reflections of the sun from a video of the sun being reflected off the waves and ripples on the sea surface. A lot of people said it could not be done, but with persistence and effort, Jussi and his co-workers succeeded.

If I am sounding like a cheer leader for Jussi Laako, I am unashamedly so. He has put thousands of work hours and 19 years of effort, and changes just Euro 110 for a licence. How many of us can claim such a effort / price ratio ?

 

[RayDunzl]

I am saying there is a full bench behind them so there will be no die off. The bar for them is lower in forums like this that are so dominated by over the top systems and people.

Oh no, I'll be in the last bucket next year...

 

[rbbert]

I am saying there is a full bench behind them so there will be no die off. The bar for them is lower in forums like this that are so dominated by over the top systems and people.

Although you present no data to support this assertion.

 

[microstrip]

Hi microstrip,

Thank you for asking the questions I can not answer
Jussi told me that this is what he does in his software. He used to say 18 months ago, the artifacts I can find, now he says he finds all of them !

An interesting anecdote from the video field which is not comparable or similar to audio, but just an example of what it possible with clever algorithms and brute force computing can do. A problem that Jussi and his work colleagues worked on was to remove the reflections of the sun from a video of the sun being reflected off the waves and ripples on the sea surface. A lot of people said it could not be done, but with persistence and effort, Jussi and his co-workers succeeded.

If I am sounding like a cheer leader for Jussi Laako, I am unashamedly so. He has put thousands of work hours and 19 years of effort, and changes just Euro 110 for a licence. How many of us can claim such a effort / price ratio ?

Many thanks for addressing my tricky questions. The Euro 110 is really a bargain, as you say, but the time and effort to set a system to test HQ player and test it in adequate conditions is significant, otherwise I would try it tomorrow! But it is high on my list, although due particular conditions I do not have the time to seriously evaluate anything more now.

BTW, my test will be challenging IMHO. Until now I have found that with the Vivaldi system, a CD or SACD spinning in the matching dedicate transport sounds significantly better the same file going through any type of server. So any server will be compared against the Vivaldi transport ... 95% of the music I really appreciate only exists on those formats.

 

[PeterA]

The graph shows 75% at age 40 and older which is what I said.

Yes, the furnished data supports your statement. Your guess was correct, and your concern is not unique.

 

[amirm]

Although you present no data to support this assertion.

No, I was the one with the data. It was the other assertion which had none. Mine showed 1/3 of the visitors to ASR Forum are at the age of 44 and lower. It is a successive filling of the pool:

Now if you have data that says those people will get out of the hobby as they get older, I like to see it.

 

[microstrip]

(...) It seems to me more and more that digital theory, in particular the Nyquist theorem about sampling rate, was right all along -- there was so much potential in that lowly CD from the start, and we are now just figuring out how much incredible detail information, sheer musical resolution, is buried in those pits.

I think that no one with a solid technical formation questioned the digital theory, that is built on perfect ADCs and perfect DACs, or the Nyquist Theorem - but they simply considered the limitations imposed by the actual implementations and accepted its consequences, evidenced by listening.

I never doubted that the information was there since many years ago - a considerable number of exceptional listening sessions with redbook digital along more than fifteen years have proved it to me. But to extract and expose it systematically and consistently in an adequate way with many types of recordings and systems has proven a difficult task.

We have no scientific basis in the influence of the so called "small differences" in sound quality in the high-end. As pointed before by other members, in fundamental aspects the high-end is still mainly empirical. IMHO WBF has proven to be a great place to debate digital.

 

[amirm]

Yes, the furnished data supports your statement. Your guess was correct, and your concern is not unique.

Nope. He is wrong. If he were right then there would be no bucket of young players in these high-end forums. I go to tons of audio shows and there is hardly any such sentiment of the market collapsing after we go away. To be sure, it is a popular cry to say the marketing is dying but it simply is not. Here is JA in stereophile: http://www.stereophile.com/content/how-revive-high-end-audio#eDJHJ4Q8HDPqVx3t.97

"Postscript

That open letter was sent eight years ago and, perhaps to no-one's surprise, it had no impact or effect. Many observers feel the situation is even worse in 2013 than it was in 2005, with the high-end audio industry even further alienated from customers younger than the baby-boom generation. But with the resurgence of the LP, especially among young music lovers, the advent of computer- and mobile-based audio that is no longer limited in quality by the unmusical noise of lossy codecs like MP3, and the explosion of headphone-based listening, which allows audiophiles of limited means to buy and enjoy Class A audio components without having to spend more than four figures, I believe the future of high-end audio is brighter than it used to be. You might say that it now has a future!—John Atkinson


Read more at http://www.stereophile.com/content/how-revive-high-end-audio#pdz3qiCSFTHbZedb.99

 

[PeterA]

Nope. He is wrong. If he were right then there would be no bucket of young players in these high-end forums. I go to tons of audio shows and there is hardly any such sentiment of the market collapsing after we go away. To be sure, it is a popular cry to say the marketing is dying but it simply is not. Here is JA in stereophile: http://www.stereophile.com/content/how-revive-high-end-audio#eDJHJ4Q8HDPqVx3t.97

"Postscript

That open letter was sent eight years ago and, perhaps to no-one's surprise, it had no impact or effect. Many observers feel the situation is even worse in 2013 than it was in 2005, with the high-end audio industry even further alienated from customers younger than the baby-boom generation. But with the resurgence of the LP, especially among young music lovers, the advent of computer- and mobile-based audio that is no longer limited in quality by the unmusical noise of lossy codecs like MP3, and the explosion of headphone-based listening, which allows audiophiles of limited means to buy and enjoy Class A audio components without having to spend more than four figures, I believe the future of high-end audio is brighter than it used to be. You might say that it now has a future!—John Atkinson


Read more at http://www.stereophile.com/content/how-revive-high-end-audio#pdz3qiCSFTHbZedb.99

All he did was take a guess at the percentage of people over 40 who are buying the media. Your graph proved him right about his guess. So he was right, not wrong. He then expressed his concern that the demographics don't bode well for the future of the high end. Your graph seems to support his claim, not prove that he is wrong. He expressed a concern based on the numbers and made no claims which he could not prove. You, however, "believe" that because there is data showing a "bucket" of 25% of these two forum members are aged below 40 that the high end is not in trouble. This data is not an indication of the future state of the industry. I'm glad that Atkin believes it too. If you have data showing that the young bucket will stay in the hobby as they get older, then we can discuss this interesting topic in a different thread.

 

[rbbert]

About 14% under 35 doesn't seem to me to be much filling the pool in the future. If they all stay in the hobby, and the percentage of young people entering remains at this level, it will soon be gone

 

[Ken Newton]

This is an interesting interpretation that I haven't read elsewhere and which seems rather unorthodox, but I may be mistaken. Do you have references for that?

If you Google the sampling theorem, you'll typically find the theorem stated with a summation running from negative to positive infinity stipulated. This indicates that the associated expression of sampling is to be evaluated over that range, from negative to positive infinity.

I find it much easier to understand how transient signal spreading/smearing occurs by utilizing the following hypothetical example:

Suppose you have a 255 stage FIR interpolation filter. Now, send 100 cycles of an Fs/4 signal (which is one-half the Nyquist frequency) through that filter to represent an transient signal. These 100 cycles will, on average, be comprised of 4 samples per cycle. So, 100 cycles x 4 samples per cycle = 400 samples. Therefore, the 100 cycle transient signal input to the interpolation filter will have a duration of 400 sample periods.

Our hypothetical interpolation filter consists of 255 stages, which, because of how FIR filter convolution functions, smears/speads all transient signals by an additional 255 sample periods. Therefore, our 100 cycle hypothetical transient becomes smeared from 400 sample periods to; 400 + 255 = 655 sample periods.

The more samples comprising an input transient, the smaller becomes the relative spreading of that transient's duration. Conversely, the shorter the signal duration the more progressive the relative spreading.

 

[amirm]

About 14% under 35 doesn't seem to me to be much filling the pool in the future.

It is actually. Their disposable incoming in that age bracket limits how many can afford high-end gear. As they get older, their income increases and with it, the population expands.

If they all stay in the hobby, and the percentage of young people entering remains at this level, it will soon be gone

Nope. There is constant flow despite the strong limiting factor of high cost of high-end gear.

MP3 revolution has been with us for nearly two decades. If it was going to kill the high-end, by now there would have been no young people in the pool. Yet there is.

Again, spend some time in the industry, talk to people and you see that there is no gloom and doom. There were when we had the recession. But with the economy getting stronger and strong, and cost of excellent music reproduction getting lower and lower, I see no danger of collapse.

 

[amirm]

And oh, note the development of lossless streaming by spotify: http://audiosciencereview.com/forum/index.php?threads/spotify-offering-lossless-confirmed.1416/

That should have never happened if the market was going backward.

 

[Folsom]

Am I the only one that fears MQA being a standard among recording could be a problem? What if we don't like the results all the time?

Still, I don't care for the monopoly affect of it.

Frankly while MQA might bring a benefit of some kind, I don't believe the potential of what we have has been addressed nearly well enough.

 

[Ken Newton]

It seems to me that that is also what higher resolution digital formats in general, whether PCM or DSD, are trying to minimize?

Not necessarily. It depends on how the available ultrasonic bandwidth is utilized. As I recall, most of the hype orginally surrounding high-rez PCM was that it sounded better due to it's ability to capture and reproduce the ultrasonic band flat up to the Nyquist frequency. In other words, the ultrasonic bandwidth was frequency-domain optimized. What then ensued was long debate about whether or not ultrasonic content could somehow be perceived.

It seems apparent, from reading publicly available information regarding MQA, that the available ultrasonic bandwidth is utilized for signal time-domain optimization rather than to maintain a flat frequency response of the channel up to Nyquist. MQA's specified anti-alias and anti-image filters softly roll off the channel's frequency response across the one or two ultrasonic octaves provided above the audio band. This soft roll off is a necessary consequence of obtaining channel time-domain optimization. High-rez PCM is also capable of such time-domian optimization, but there just is no requirement for it. As I understand it, MQA includes such a requirement.

I'm not familiar with the all of the channel coding details of DSD, but I suspect that it, too, may be time-domain optimized.

 

[rbbert]

It is actually. Their disposable incoming in that age bracket limits how many can afford high-end gear. As they get older, their income increases and with it, the population expands.


Nope. There is constant flow despite the strong limiting factor of high cost of high-end gear.

MP3 revolution has been with us for nearly two decades. If it was going to kill the high-end, by now there would have been no young people in the pool. Yet there is.

Again, spend some time in the industry, talk to people and you see that there is no gloom and doom. There were when we had the recession. But with the economy getting stronger and strong, and cost of excellent music reproduction getting lower and lower, I see no danger of collapse.

I'm not saying anything about the future of our hobby, just that your graph doesn't suggest a rosy future. I assumed (since there is nothing in the image to suggest otherwise) that the percentages refer to the number of members in those age groups, not how much they spend. If it's just numbers of individual audio/videophiles, in 20 years we'll be lucky to have half the number of participants we have today; if in fact the relative percentages of young members stays the same, in another 10 years after that the numbers will be horrible. Longer life spans and rapid population growth in the 15-25 yr age group could change that conclusion, but again, that data is not in the graph.

Yes there is other data to suggest that the future is not that bleak, but it's not in the graph you posted.

 

[rbbert]

Not necessarily. It depends on how the available ultrasonic bandwidth is utilized. As I recall, most of the hype orginally surrounding high-rez PCM was that it sounded better due to it's ability to capture and reproduce the ultrasonic band flat up to the Nyquist frequency. In other words, the ultrasonic bandwidth was frequency-domain optimized. What then ensued was long debate about whether or not ultrasonic content could somehow be perceived...

I'm not an expert on digital audio theory or even its current application, but I have certainly read that a supposed benefit of higher sampling frequency and greater bit depth could be a reduction in quantization distortion. Yes that is supposed to be inconsequential in the amplitude domain, but as you point out, not in the time domain?

 

[rbbert]

Another thing to consider about the future of high-end audio; when I started the San Diego Audio Society in 1977, about 60% of the members were under 40. I suspect others' experience from that time period is similar, perhaps even demonstrating a more significant preponderance of young audiophiles.

 

[Mike Lavigne]

Another thing to consider about the future of high-end audio; when I started the San Diego Audio Society in 1977, about 60% of the members were under 40. I suspect others' experience from that time period is similar, perhaps even demonstrating a more significant preponderance of young audiophiles.

I think the generational period of Baby Boomers in their late 20's to late 30's (I was 26 with my second child arriving in 1977) is not comparable to any other when you consider the cultural connection of this numerically huge generation to the music of the 60's and 70's. you almost cannot separate the music from the generation. it was an industrious generation that also went to work and lot's of them had disposable income, and this was before the recreational choices were so diverse. having a stereo was pretty main stream right then. and the older guys over 40 were relatively far fewer numerically then.

so what do we learn from the baby boomer love for hifi/stereo/music? i'm not sure. probably it was a cosmic accident. hundreds (thousands?) of books have been written about things related to it.

 

[Ken Newton]

I'm not an expert on digital audio theory or even its current application, but I have certainly read that a supposed benefit of higher sampling frequency and greater bit depth could be a reduction in quantization distortion. Yes that is supposed to be inconsequential in the amplitude domain, but as you point out, not in the time domain?

Quantization noise is something apart from the sample rate. The sampe rate determines the information bandwidth that can be captured without aliasing, while the quantization noise has to do with the bits of resolution of each sample. One doesn't determine the other.

That said, to complicate matters, sigma-delta modulation techniques utilize oversampling to open ultrasonic bandwidth to relocate audio band quantization noise, where it's then called, out-of-band noise.