Is Digital Audio A Mature Science?

[Allan F]

7 minutes ago, pkane2001 said:

 

Facts need to be checked. Otherwise they may wind up as 'alternative'.

Only those that are seriously in doubt. As you know, the term "alternative fact" is a creation of the current White House and is an oxymoron.

[pkane2001]

1 minute ago, Allan F said:

Only those that are seriously in doubt. As you know, the term "alternative fact" is oxymoronic.

 

Claim that human ears are capable of detecting audio differences that are not measurable is very seriously in doubt.

 

[Allan F]

8 minutes ago, pkane2001 said:

 

Claim that human ears are capable of detecting audio differences that are not measurable is very seriously in doubt.

 

In your opinion, perhaps. Those claims are not in doubt by the innumerable who have heard such differences repeatedly for decades.

[pkane2001]

3 minutes ago, Allan F said:

 

In your opinion. Those claims are not in doubt by the innumerable who have heard such differences repeatedly for  decades.

 

That's an irrefutable proof. I bow to your faith in unknown and unnamed innumerables.

[The Computer Audiophile]

4 minutes ago, pkane2001 said:

 

Claim that human ears are capable of detecting audio differences that are not measurable is very seriously in doubt.

 

 

1 minute ago, Allan F said:

 

In your opinion. Those claims are not in doubt by those who have heard them repeatedly for many decades.

 

This is a topic that has always interested me and deserves a good discussion without the noise on the extremes.

 

I waver back and forth on this one. Sometimes one certain days it seems like I can hear an incredible amount of stuff I should be hearing. This is without me sitting down to "test" something. Just kicking back and listening. I understand the science completely and I know it flies in the face of this experience. An interesting topic that hopefully the adults around here can discuss.

[Allan F]

8 minutes ago, The Computer Audiophile said:

 

 

This is a topic that has always interested me and deserves a good discussion without the noise on the extremes.

 

I waver back and forth on this one. Sometimes one certain days it seems like I can hear an incredible amount of stuff I should be hearing. This is without me sitting down to "test" something. Just kicking back and listening. I understand the science completely and I know it flies in the face of this experience. An interesting topic that hopefully the adults around here can discuss.

 

The science only "flies in the face of this experience" to the extent that one believes that current knowledge is exhaustive and that everything that counts is measurable according to that knowledge.

[pkane2001]

9 minutes ago, The Computer Audiophile said:

 

 

This is a topic that has always interested me and deserves a good discussion without the noise on the extremes.

 

I waver back and forth on this one. Sometimes one certain days it seems like I can hear an incredible amount of stuff I should be hearing. This is without me sitting down to "test" something. Just kicking back and listening. I understand the science completely and I know it flies in the face of this experience. An interesting topic that hopefully the adults around here can discuss.

 

Chris, check the study that I shared earlier in this thread. At the very least it points to why comparing audio playback without controlling for bias is wrought with potential errors. 

[crenca]

I am only a junior objectivist, so take this with a grain of salt:

 

The usual measurements (in both digital and analog domain) are significant but not in any way a total picture of audio.  Also, there are many open questions (such as the how and what of the "sound" of distortion of all types).

 

Most of those between the "extremes" understand this and besides occasionally making an informed but speculative opinion (e.g. "the spike of distortion at 5k can be heard as a dip in FR response...") don't dwell on it because, this relationship between measurements and sound has these unknowns and nooby seems to know what to do about it.

 

All that said, between the overemphasis and built in tilt of audiophiledom towards subjectivist side of the extreme and all the voodoo from the industry, it is unlikely that any balanced conversation about the the significance of measurements can rise above the culture...

[marce]

There is work and info out there from many sources... But mention someone like Toole or the rest and as stated endless arguments about the person, science just about anything apart from the data start... And if the data flies in the face of audiophile beliefs it is dismissed as are many laws of physics that don't fit with the beliefs.

[The Computer Audiophile]

17 minutes ago, pkane2001 said:

 

Chris, check the study that I shared earlier in this thread. At the very least it points to why comparing audio playback without controlling for bias is wrought with potential errors. 

I agree 100%. Wrought with potential errors. I think everyone should understand this no matter what side of the fence they are on.

[Albrecht]

41 minutes ago, crenca said:

I am only a junior objectivist, so take this with a grain of salt:

 

The usual measurements (in both digital and analog domain) are significant but not in any way a total picture of audio.  Also, there are many open questions (such as the how and what of the "sound" of distortion of all types).

 

Most of those between the "extremes" understand this and besides occasionally making an informed but speculative opinion (e.g. "the spike of distortion at 5k can be heard as a dip in FR response...") don't dwell on it because, this relationship between measurements and sound has these unknowns and nooby seems to know what to do about it.

 

All that said, between the overemphasis and built in tilt of audiophiledom towards subjectivist side of the extreme and all the voodoo from the industry, it is unlikely that any balanced conversation about the the significance of measurements can rise above the culture...

""Most of those between the "extremes" understand this and besides occasionally making an informed but speculative opinion (e.g. "the spike of distortion at 5k can be heard as a dip in FR response...") don't dwell on it because, this relationship between measurements and sound has these unknowns and nooby seems to know what to do about it.""

 

This is a point that makes sense in a certain context of FR & performance and certain levels of gear. And again, that super valuable empirical knowledge solves this issue by selecting amplification that has a slight boost in the areas that the speakers have a slight drop: like some models of Manley amps and Paradigm speakers have such a synergy.

 

Against the arguments from psuedo-scientists or "measurers," - there isn't really any voodoo in the industry, - Thank goodness. Indeed the quote "high-end" audio industry has a firmer foundation in cost per performance than many others. Whenever we have an expensive component, - that component is almost always contributing to a better sounding event. If not, - it, and it's company, - doesn't last long. For example there are very few DACs/CDPs that perform better than the Meitner/EMM Labs. There was a great reason why these were all over CES for many years. They were/are the "gold standard" for DACs/CDPs, and were nearly universally recognized as such. Because it was proven to everyone that they performed far better than their competitors and most all others.....  But of course, - outside reason, - trolls attacked them on the basis of irrelevant and cursory measurements that were in no way representative and reflective of how the SYSTEM that contained that component sounded. (In many ways, these trolls still propagate forums like this making wild speculative claims without the [critical] knowledge of comparative experience). The efficacy of any component cannot, and should not,  be judged on any cursory measurement, but on how well it contributes to the overall subjective enjoyment of the listening experience.

 

 

[Ralf11]

the above post is almost entirely false

[The Computer Audiophile]

11 minutes ago, Ralf11 said:

the above post is almost entirely false

Can you give some additional information? It would help the rest of us.

[Ralf11]

it's chock full of it - too much to do a full rebuttal on

 

one, however, is "there isn't really any voodoo in the industry"

 

then there is the denigration of scientists, and there are at least 2 or 3 on here - but that is a common canard raised by the Slithering Subjectivists

[mansr]

2 minutes ago, Ralf11 said:

it's chock full of it - too much to do a full rebuttal on

 

one, however, is "there isn't really any voodoo in the industry"

I stopped reading there. No point in continuing.

[Ralf11]

you might, however, enjoy my SQ / Measurements thread - it's new as of today

[The Computer Audiophile]

6 minutes ago, Ralf11 said:

too much to do a full rebuttal on

I understand. 

[crenca]

5 minutes ago, mansr said:

I stopped reading there. No point in continuing.

 

I made it all the way to "...Indeed the quote "high-end" audio industry has a firmer foundation in cost per performance than many others"

[fas42]

17 hours ago, 20hertz said:

Also, as a response to all the other nonconstructive comments re: vinyl.

All my design work is run thru a pair of carefully calibrated Genelec 8260s using DSP.

DSP is very good at improving speaker and room issues but it does not alter the quality of the source material.

The vinyl rig still kills it more than 50 percent of the time.

Personally I want digital to be better, I couldn't be bothered with the hassle of vinyl at home, and I have over ten TB of digital files.

But is still sounds lean and 2D to my ears.

DSP also has its price. NOTHING comes close to an analogue only system with pure class a amps directly connected to a driver.

 

 

Which is how it's been for decades. A digital chain not fully "debugged" sucks the warmth and richness out of a track- it takes listening to a, yes, sorted rig to appreciate what the true nature of the recording is.

 

It's not DSP that's doing the damage; directly, that is. Processing activity in any area of the system very, very easily injects interference in the analogue related circuitry; which then cripples the SQ. Very careful attention to detail can make all the difference - the sort of thinking that, say, PeterSt uses.

 

Fear not, digital can deliver the full bundle - the technology 30 years ago could do it, as I personally found out. But only if all the i's are dotted, and all the t's are crossed ...

[One and a half]

6 minutes ago, fas42 said:

 

Which is how it's been for decades. A digital chain not fully "debugged" sucks the warmth and richness out of a track- it takes listening to a, yes, sorted rig to appreciate what the true nature of the recording is.

 

It's not DSP that's doing the damage; directly, that is. Processing activity in any area of the system very, very easily injects interference in the analogue related circuitry; which then cripples the SQ. Very careful attention to detail can make all the difference - the sort of thinking that, say, PeterSt uses.

 

Fear not, digital can deliver the full bundle - the technology 30 years ago could do it, as I personally found out. But only if all the i's are dotted, and all the t's are crossed ...

 

Can you give an example of a digital system with dotting i and crossing t that has stood the test of 30 years?

[sandyk]

19 minutes ago, One and a half said:

 

Can you give an example of a digital system with dotting i and crossing t that has stood the test of 30 years?

 

 Other than his own ?

 

 

 

 

 

 

 

 

Sorry Frank

 The Devil made me do it.

[One and a half]

17 minutes ago, sandyk said:

 

 Other than his own ?

 

 

 

 

 

 

 

 

Sorry Frank

 The Devil made me do it.

Anything that Frank has experienced and known to work for 30 years. I'm all ears and eyes and keen to know. If it concerns digital audio, going back that far, then USB is out of the equation.

 

IMHO, USB has the bandwidth, but S/PDIF (either balanced or SE) has the 'lowest noise transmittable content'. Jitter & clocking can be controlled to a degree, largely depends on the DAC's receiver and clock circuits. The amount of malformed signals is pretty well minimal, for example RME makes a big noise about their stability of clocks, so do Mutec. Here's a good paper from Wolfson about high performance S/PDIF receivers from 2006. These days, although care is required for power supplies that feed receiver circuits are still essential, the receiver design should be well mature by now.

 

Members report reasonably consistently that a preference toward AES3 or coax inputs are better sounding than the USB inputs. I don't think the clocking advantage of USB transmissions is enough to overcome the PHY requirements and quality of transmissions needed for USB to work properly, then there's the isolation problem. Just look at the power supply lines in close proximity to the signal data at 480Mb/s...how can that be split, especially at the connector?  S/PDIF can be isolated simply with transformers, the speeds are well below awkward limits, so the design of the transformer is a well established technology. 

[sandyk]

41 minutes ago, One and a half said:

Anything that Frank has experienced and known to work for 30 years. I'm all ears and eyes and keen to know. If it concerns digital audio, going back that far, then USB is out of the equation.

 

IMHO, USB has the bandwidth, but S/PDIF (either balanced or SE) has the 'lowest noise transmittable content'. Jitter & clocking can be controlled to a degree, largely depends on the DAC's receiver and clock circuits. The amount of malformed signals is pretty well minimal, for example RME makes a big noise about their stability of clocks, so do Mutec. Here's a good paper from Wolfson about high performance S/PDIF receivers from 2006. These days, although care is required for power supplies that feed receiver circuits are still essential, the receiver design should be well mature by now.

 

Members report reasonably consistently that a preference toward AES3 or coax inputs are better sounding than the USB inputs. I don't think the clocking advantage of USB transmissions is enough to overcome the PHY requirements and quality of transmissions needed for USB to work properly, then there's the isolation problem. Just look at the power supply lines in close proximity to the signal data at 480Mb/s...how can that be split, especially at the connector?  S/PDIF can be isolated simply with transformers, the speeds are well below awkward limits, so the design of the transformer is a well established technology. 

 Garry

 It could be interesting to try something like the attached with proper terminations and 75ohm BNC connectors at each end.

 They are designed for video distribution, have a very wide and flat response and a gain of 2.

Silicon Chip magazine used them in a Video Distribution project for S-VHS etc. distribution.

Regards

Alex.

MAX496-MAX497.pdf

[One and a half]

3 minutes ago, sandyk said:

 Garry

 It could be interesting to try something like the attached with proper terminations and 75ohm BNC connectors at each end.

 They are designed for video distribution, have a very wide and flat response and a gain of 2.

 

Regards

Alex.

MAX496-MAX497.pdf

They certainly have the bandwidth, the transformer though requires no PSU and are totally passive. So long as the voltage is around the 0.5V for S/PDIF it's not a problem.

[sandyk]

39 minutes ago, One and a half said:

They certainly have the bandwidth, the transformer though requires no PSU and are totally passive. So long as the voltage is around the 0.5V for S/PDIF it's not a problem.

 

Agreed, but the amplifier would also extend the workable distance, as well as retaining  an excellent waveform requiring less processing in the Input receiver.