WBF Poll: Which Sounds Better, Digital or Analog?

[Mike Lavigne]

Did you try/buy the MasterSound (EK53927) 1994 remastered SBM gold CD ?

I have EK64425 Gold CD SBM 'Mastersound' of 'Couldn't Stand The Weather'. is that different than the catalogue number you list? if it's worth $400 maybe I need to sell it. I have lots of these gold editions.

looking at Discog; the first number is the longbox,

http://www.discogs.com/Stevie-Ray-V...ble-Couldnt-Stand-The-Weather/release/4875674

the second is the same without the long box.

http://www.discogs.com/Stevie-Ray-V...ble-Couldnt-Stand-The-Weather/release/2658549

also have the early SACD ES65871.

been years, maybe 10, since I did any comparison of these discs.

 

[sbo6]

Puhleeze, you don't even know me or what I have covered in my 30 years of reviewing, seen or heard or you are calling me myopic? Puhleeze... Or reviewed digital from day 1.

Off the top of my head, and that list includes Bruce, MikeL, MEP, Rockitman (?), Ki, and myself, there are but a handful of people here or anywhere who have and use all three music formats and have valid comparison points. (Sorry if I left someone out!) So I'm aware more than most of what's out there. And as a matter of fact am currently reviewing the iDSD DAC and options with the ExaSound and Lampi coming soon. Not to mention comparing 4X DSD with 15 ips tape. So that certainly covers the ballpark. Along with four to five shows a year, visits to,dealers, etc.

Oh and in conclusion, I'd hardly call setting up a complete computer based digital front-end easy unless maybe you arena techno geek.

Thanks and you can win by having last word. Me I'm outa here.

No Myles, you win. It's clear by your comments above that you are the expert (reminds me of another thread from WBF) and you also know me - and all us part-time audiophile scrubs who know nothing irrespective of any education, musicianship or experience we may have. Your comparison of digital vs analog circa 1988 which you believe holds up 25 years later and finite opinion that budget analog beats digital says it all. Enjoy quoting your experiences from 1988 with a Sony CD player vs a Linn Sondek playing Human League, we're all ears..

 

[TBone]

Those are worth $400? wow ...

>>been years, maybe 10, since I did any comparison of these discs.<<

Me too, although SRV was an artist I'd witness 4 times, for some reason, he's rarely on my playlist.

Silly really, some of his original CD's sound beautifully open & dynamic to my ears.

ex: like this superbly recorded original CD mastered by Bob Ludwig, (EK47390) 1991 (tr#.2:The Sky is Crying)


My counterpart LP pales by comparison.

 

[amirm]

You don't have to re-read them if you have already. But some folks might enjoy reading them in the order that I posted them (I took my time to make the link's order as comprehensible as I can). > I did my research too, and I also read them all. ...Plus much more; but them seven links were the ones I picked @ the end.

Hi Bob. I read a ton of articles about audio yet you manage to unearth links I have not seen . So I read the first one and unfortunately it is completely wrong. It has this common graph:

What he shows about digital is just flat our wrong. Take a 1 Khz sine wave, convert it to digital, and playing on a CD player. The look at the output from the analog jack on the CD player. It will look just like the original waveform on the left. It never, ever looks like what he is showing. If it did, you would think someone, some place, would have measured a CD player outputting such, rather than a graphic created in a paint program.

The reason what he shows is impossible is not because of anything related to audio. It is the nature of signals. That sine wave he shows with steps in it indicates at those increments, the signal immediately jumped from one value to the other. For a signal to go from one value in zero time to another value, you must, let me repeat, must, have infinite energy. And of course no real signal does.

A signal that is similar to that is a square wave. A symmetrical square wave at 1 Khz has harmonics going to infinity at odd multiples of 1 KHz, i.e. 3 Khz, 5 KHz, 7 Khz, etc. Those harmonics based on above, must go into infinity or else, you don't have a square wave. You can choose to truncate that signal at any point, and if you do, then your signal deviates from square wave. The more you truncate those harmonics, the more it will not look like a square wave.

Turns out that is exactly what every CD player/DAC does. They have a filter on the output of the DAC whose job is to eliminate all of those extra harmonics. It is called a reconstruction filter. Apply that filter and 1 Khz now looks just like 1 Khz and that is that. It cannot have steps in it because there are no harmonics to generate them.

We can see this effect if we create a square wave using a computer, record it and then play it on a CD player. Stereophile does this in their CD player tests. Here is a random example:

The CD player was instructed to play a pure, square wave. The numbers were computed to be completely accurate. Yet what comes out of the CD player are those wavy corners. The corners get wavy because the CD player's DAC is filtering the high frequencies. When it does that, the math, not anything to do with digital or analog, says that it cannot be square wave.

Now, the reproduction is not perfect because it is limited to the resolution we have determined for our system. There are errors that can be measured but they don't show up like his made up graph. Here is stereophile again, feeding a DAC a sinewave at -90 dbFS:

Yes, it doesn't look like a perfect sine wave. The reason is that we are feeding the DAC such faint signals. No analog system can even reproduce such a clean signal as it will be overwhelmed by noise. A tape deck would be doing good at -80 db. The little jumps in there also is due to lack of dither. If we tolerate a bit of noise, they can look like analog noise just the same.

Cutting out all the technical bits, he is attempting to explain away theory of relativity from knowledge level of elementary school kid. It is junk marketing material. Not any kind of technical article. The person is unqualified to write such things.

Digital audio concepts unfortunately are complicated. They immediately involve a ton of math and signal processing. They are not approachable as a topic so I wish people would not. They can say they have a preference against it which is fine. But don't try to get technical and mislead when a simple measurement would invalidate everything you just wrote.

The man even gets the analog side wrong. Analog by definition has noise and lots of it. There is no way what you feed it, is what comes out, putting aside all kinds of other distortions it also piles in for good measure. Waveform comparisons of real devices would paint a horrific picture of analog against digital. It just happens that we either like the distortion, or it gets masked.

 

[rbbert]

No Myles, you win. It's clear by your comments above that you are the expert (reminds me of another thread from WBF) and you also know me - and all us part-time audiophile scrubs who know nothing irrespective of any education, musicianship or experience we may have. Your comparison of digital vs analog circa 1988 which you believe holds up 25 years later and finite opinion that budget analog beats digital says it all. Enjoy quoting your experiences from 1988 with a Sony CD player vs a Linn Sondek playing Human League, we're all ears..

I don't know why we still argue about this , but ti's unfair to apply this characterization to Myles or the other forum members he mentioned, they all have pretty up-to-date high quality digital playyback as well as analog. The fact that they prefer their analog playback doesn't mean everyone or indeed anyone else should necessarily share that preference, so as I said I don't know why we continue to argue. Post your preference, discuss reasons for it if you want, but get over the idea that any of us will convince any others who is "right" or "wrong", because in this area it's obviously an individual choice.

 

[amirm]

So that we are firmly grounded here, this is a random measurement of a turntable (Linn Sondek LP12) from stereophile: http://www.stereophile.com/content/linn-sondek-lp12-turntable-lingo-power-supply-measurements

The test tone on the LP was 1.003 KHz test tone. What should have come out in the spectrum analyzer output above, should have been a single, sharp, vertical line at that frequency and absolutely nothing else. What came out is something else:

1. The tone is in there alright in the middle but it has widened substantially. That wide "skirt" is due to random speed modulation. It is ton of variations but its frequency is limited to +- 50 Hz.

2. The shoulders are broad even if we just go down 10 db from the peak. The variations have create huge amplitude of random low frequency distortions. We are talking orders of magnitude more than we see in digital system jitter.

3. We then have symmetrical spikes on each side of the tone. That indicates speed variations that have fairly precise causes that occur at X/second. They are sudden jumps in speed and therefor create distortion spikes at those exact frequencies. These rise up to -60 dB FS.

4. Going way past the center frequency, the signal never gets clean, establishing a base level of distortion+noise that is 70 db FS below input.

What this says is that this analog system was not remotely capable of creating a sine wave that was fed to it. This again invalidates those made up graphs of analog simply reproducing what is given to it as in the Counterpoint graph. It can't do that or even come within shouting distance.

By any objective measure, these are horrible, horrible measurements.

Thankfully the story does not end there. If we look at the distortions that MP3 creates, they too look horrific (not on sine waves but complex content). But we don't hear them easily because the codec attempts to create distortions that are masked. Same is happening here. The above graph has very narrow scale. So a lot of those distortions get overshadowed by the main tone. Not all though. Some are coming through and may be what people like to hear.

I know if you love analog you are rolling your eyes over this. That's fine. I just ask that we don't try to all of a sudden sound technical and pretend to explain how the technologies work when we don't have a grasp of either. Just stay in preference domain and we are good. Don't try to get technical .

 

[mep]

I don't know why we still argue about this , but ti's unfair to apply this characterization to Myles or the other forum members he mentioned, they all have pretty up-to-date high quality digital playyback as well as analog. The fact that they prefer their analog playback doesn't mean everyone or indeed anyone else should necessarily share that preference, so as I said I don't know why we continue to argue. Post your preference, discuss reasons for it if you want, but get over the idea that any of us will convince any others who is "right" or "wrong", because in this area it's obviously an individual choice.

Well said Rob. It's all a preference. I'm convinced we all hear different just as everyone's eyesight is all over the map. Birds of a feather tend to flock together. I'm not sticking my head in the sand and ignoring digital. I wouldn't have gone through the time and expense of setting up my digital system if I truly didn't care and didn't want to keep an eye and ear on what is going on with digital. Is digital getting better sounding to me? Yes, it is thanks to DSD and I know that starts another digital food fight. I'm not sure that it's a coincidence that those who prefer analog over digital prefer DSD over PCM.

 

[mep]

So that we are firmly grounded here, this is a random measurement of a turntable (Linn Sondek LP12) from stereophile: http://www.stereophile.com/content/linn-sondek-lp12-turntable-lingo-power-supply-measurements

The test tone on the LP was 1.003 KHz test tone. What should have come out in the spectrum analyzer output above, should have been a single, sharp, vertical line at that frequency and absolutely nothing else. What came out is something else:

1. The tone is in there alright in the middle but it has widened substantially. That wide "skirt" is due to random speed modulation. It is ton of variations but its frequency is limited to +- 50 Hz.

2. The shoulders are broad even if we just go down 10 db from the peak. The variations have create huge amplitude of random low frequency distortions. We are talking orders of magnitude more than we see in digital system jitter.

3. We then have symmetrical spikes on each side of the tone. That indicates speed variations that have fairly precise causes that occur at X/second. They are sudden jumps in speed and therefor create distortion spikes at those exact frequencies. These rise up to -60 dB FS.

4. Going way past the center frequency, the signal never gets clean, establishing a base level of distortion+noise that is 70 db FS below input.

What this says is that this analog system was not remotely capable of creating a sine wave that was fed to it. This again invalidates those made up graphs of analog simply reproducing what is given to it as in the Counterpoint graph. It can't do that or even come within shouting distance.

By any objective measure, these are horrible, horrible measurements.

Thankfully the story does not end there. If we look at the distortions that MP3 creates, they too look horrific (not on sine waves but complex content). But we don't hear them easily because the codec attempts to create distortions that are masked. Same is happening here. The above graph has very narrow scale. So a lot of those distortions get overshadowed by the main tone. Not all though. Some are coming through and may be what people like to hear.

I know if you love analog you are rolling your eyes over this. That's fine. I just ask that we don't try to all of a sudden sound technical and pretend to explain how the technologies work when we don't have a grasp of either. Just stay in preference domain and we are good. Don't try to get technical .

Does that mean that all turntables measure as bad as the Linn? Today's and some of yesterday's tables have much better speed regulation than the Linn that was measured by SP. Also, we don't know if the LP that was used for this test was pressed off center and if this played any part. The good news is that we don't listen to 1 kHz sine waves, we listen to music.

 

[amirm]

Amir this has zero to do with vision but all about cognition. You need to read some new textbooks. When I'm back home Sunday, I'll post a few titles that certainly contradict your assertion about short term memory and processing.

You should also read the aforementioned book The Brain That Changes Itself.

http://www.amazon.com/Brain-That-Ch...226793&sr=8-1&keywords=The+brain+that+changes

It is nearly 450 pages and it says this in the description:

"An astonishing new science called neuroplasticity is overthrowing the centuries-old notion that the human brain is immutable. Psychiatrist and psychoanalyst, Norman Doidge, M.D., traveled the country to meet both the brilliant scientists championing neuroplasticity and the people whose lives they’ve transformed—people whose mental limitations or brain damage were seen as unalterable. We see a woman born with half a brain that rewired itself to work as a whole, blind people who learn to see, learning disorders cured, IQs raised, aging brains rejuvenated, stroke patients learning to speak, children with cerebral palsy learning to move with more grace, depression and anxiety disorders successfully treated, and lifelong character traits changed. Using these marvelous stories to probe mysteries of the body, emotion, love, sex, culture, and education, Dr. Doidge has written an immensely moving, inspiring book that will permanently alter the way we look at our brains, human nature, and human potential."

None of this relate to the topic at hand. I am happy to read it if you can quote appropriate section that relates to what we are talking about.

As far as the brain is concerned, there is no difference between senses and for example leading researchers in Germany and Israel are using sound to teach blind people to see or taste to teach people to be able to stand. It's all about primary vs. secondary neural networks and disspelling the ages old notion that only certain parts of the brain are responsible for specific actions. As you know, when we reach 12 (this also happens in the third trimester of development in utero) the brain prunes those pathways that we don't use and strengthens by myelination the networks we do use. That's why from when we are born to 12 or so, there are critical,things we must do from year to year to fully develop our brains and physical and mental capabilities. We all know the one about speaking a foreign language before 12. Well those pruned pathways never go away and these people through much effort are able to help these people gain lost abilities and senses. And the key to this is that all senses are just another input to the brain.

They are other inputs to the brain and precisely the reason we need to eliminate them to create proper tests. Visual cues do indeed manufacture aural perception. A pretty plate of food tastes better than an ugly one even though the food by definition is identical.

The question we ask ourselves though, is if two steaks taste the same. In that comparison, you can't make a pretty plate out of one, and ugly one out of the other and then ask people which they like better. We want to have a controlled test where we only change *one variable*. Then if there is a difference found, we know it is due to that one variable. If someone likes the steak in a pretty plate, we don't know if it is due to the fact that it was pretty or it was a better steak. This is why we do the test blind. We take away the knowledge. We can still let you look at your gear by the way as the test is not literally blind as you know. But we can't let you know that you are hearing your preferred system in one instance, and not in the other.

 

[TBone]

A 1995 LP12, with all due respect to this iconic table, but I've heard enough early LP12 (and taken 'em apart to setup/fix) to understand its many inherited compromises (hence the many expensive updates/fixes later) ... I'd suggest that many of today's better entry level players, especially those more rigidly built, could potentially measure much better. Even in 95, many considered it a compromised affair in direct comparison to better built alternatives.

 

[amirm]

Does that mean that all turntables measure as bad as the Linn? Today's and some of yesterday's tables have much better speed regulation than the Linn that was measured by SP. Also, we don't know if the LP that was used for this test was pressed off center and if this played any part. The good news is that we don't listen to 1 kHz sine waves, we listen to music.

Answering the last part, any piece of music can be decomposed mathematically into a combination of pure sine waves. So what you see there, most definitely is also being applied to every tone in the make up of your music. What happens though is that the distortion will be frequency sensitive. The amplitude of the error increases with frequency. So in that sense 1 Khz is being very generous to LP. In digital, we test at 11 Khz by way of reference.

As for other measurements, I would love to see them. Here are some other ones I found. This one from a "u-turn AUDIO Basic Turntable" which he says is better than Proj-ect:

This is the rumble and noise. The thing is not being asked to play anything but is spitting out those spikes and noise. A CD player sitting there of course doesn't produce anything like this.

Here is another set: http://www.audioasylum.com/audio/vinyl/messages/68/683183.html

Second harmonic at just -36 db? A DAC designer will commit suicide out of embarrassment at numbers twice as big . 1.12% distortion???

-57 db. Cassette deck noise levels were around -60 db.

-46 db?

As I said, the technical side of analog is horrific. As an analog lover, you just don't want to go there.g

 

[NorthStar]

Hi Bob. I read a ton of articles about audio yet you manage to unearth links I have not seen . So I read the first one and unfortunately it is completely wrong. It has this common graph:

What he shows about digital is just flat our wrong. Take a 1 Khz sine wave, convert it to digital, and playing on a CD player. The look at the output from the analog jack on the CD player. It will look just like the original waveform on the left. It never, ever looks like what he is showing. If it did, you would think someone, some place, would have measured a CD player outputting such, rather than a graphic created in a paint program.
The reason what he shows is impossible is not because of anything related to audio. It is the nature of signals. That sine wave he shows with steps in it indicates at those increments, the signal immediately jumped from one value to the other. For a signal to go from one value in zero time to another value, you must, let me repeat, must, have infinite energy. And of course no real signal does.
A signal that is similar to that is a square wave. A symmetrical square wave at 1 Khz has harmonics going to infinity at odd multiples of 1 KHz, i.e. 3 Khz, 5 KHz, 7 Khz, etc. Those harmonics based on above, must go into infinity or else, you don't have a square wave. You can choose to truncate that signal at any point, and if you do, then your signal deviates from square wave. The more you truncate those harmonics, the more it will not look like a square wave.
Turns out that is exactly what every CD player/DAC does. They have a filter on the output of the DAC whose job is to eliminate all of those extra harmonics. It is called a reconstruction filter. Apply that filter and 1 Khz now looks just like 1 Khz and that is that. It cannot have steps in it because there are no harmonics to generate them.
We can see this effect if we create a square wave using a computer, record it and then play it on a CD player. Stereophile does this in their CD player tests. Here is a random example:

The CD player was instructed to play a pure, square wave. The numbers were computed to be completely accurate. Yet what comes out of the CD player are those wavy corners. The corners get wavy because the CD player's DAC is filtering the high frequencies. When it does that, the math, not anything to do with digital or analog, says that it cannot be square wave.
Now, the reproduction is not perfect because it is limited to the resolution we have determined for our system. There are errors that can be measured but they don't show up like his made up graph. Here is stereophile again, feeding a DAC a sinewave at -90 dbFS:

Yes, it doesn't look like a perfect sine wave. The reason is that we are feeding the DAC such faint signals. No analog system can even reproduce such a clean signal as it will be overwhelmed by noise. A tape deck would be doing good at -80 db. The little jumps in there also is due to lack of dither. If we tolerate a bit of noise, they can look like analog noise just the same.
Cutting out all the technical bits, he is attempting to explain away theory of relativity from knowledge level of elementary school kid. It is junk marketing material. Not any kind of technical article. The person is unqualified to write such things.
Digital audio concepts unfortunately are complicated. They immediately involve a ton of math and signal processing. They are not approachable as a topic so I wish people would not. They can say they have a preference against it which is fine. But don't try to get technical and mislead when a simple measurement would invalidate everything you just wrote.
The man even gets the analog side wrong. Analog by definition has noise and lots of it. There is no way what you feed it, is what comes out, putting aside all kinds of other distortions it also piles in for good measure. Waveform comparisons of real devices would paint a horrific picture of analog against digital. It just happens that we either like the distortion, or it gets masked.

You are right Amir, there is no name attached with that short article...another proof that the Internet is NOT all true. ...I didn't know; it all sounded and looked legit @ first glance. I guess Digital is not so bad @ the end.

_______

* By the way, did you read the second link, with Dr. Sean Olive being interviewed on the Analog/Digital state of the affair?

 

[sbo6]

I don't know why we still argue about this , but ti's unfair to apply this characterization to Myles or the other forum members he mentioned, they all have pretty up-to-date high quality digital playyback as well as analog. The fact that they prefer their analog playback doesn't mean everyone or indeed anyone else should necessarily share that preference, so as I said I don't know why we continue to argue. Post your preference, discuss reasons for it if you want, but get over the idea that any of us will convince any others who is "right" or "wrong", because in this area it's obviously an individual choice.

Agree 110%. I didn't argue for or against, simply stated IMO $ for $ digital is better - again, IMO. Please also note that I never said digital is better, in fact the best source I have heard has been analog, truly stunning playback.

 

[Dre_J]

"...I am format agnostic, a full-time music-lover and a part-time audiophile. I listen to as much digital as analog, and prefer the better sounding music. I work hard at hunting down the best sounding copy of the music I love. Most of the time (as of mid-2015), I prefer the analog version when both are available."

Oh My!

 

[garylkoh]

I have EK64425 Gold CD SBM 'Mastersound' of 'Couldn't Stand The Weather'. is that different than the catalogue number you list? if it's worth $400 maybe I need to sell it. I have lots of these gold editions.

looking at Discog; the first number is the longbox,

http://www.discogs.com/Stevie-Ray-V...ble-Couldnt-Stand-The-Weather/release/4875674

the second is the same without the long box.

http://www.discogs.com/Stevie-Ray-V...ble-Couldnt-Stand-The-Weather/release/2658549

also have the early SACD ES65871.

been years, maybe 10, since I did any comparison of these discs.

I don't have the SRV, but I thought that the longbox of Roger Waters Amused To Death was quite a bit better the standard short box.

 

[amirm]

A 1995 LP12, with all due respect to this iconic table, but I've heard enough early LP12 (and taken 'em apart to setup/fix) to understand its many inherited compromises (hence the many expensive updates/fixes later) ... I'd suggest that many of today's better entry level players, especially those more rigidly built, could potentially measure much better. Even in 95, many considered it a compromised affair in direct comparison to better built alternatives.

The new state-of-the-art turntables do seem to go to heroic levels to squash speed variations. Shame there is no objective data to analyze how far they have come.

I am confident though that no matter how much progress they have made, digital still leaves analog in the dust as far as pure numbers.

 

[TBone]

Agree 110%. I didn't argue for or against, simply stated IMO $ for $ digital is better - again, IMO. Please also note that I never said digital is better, in fact the best source I have heard has been analog, truly stunning playback.

I understood your intent; I was once entrenched in the analog is superior by default camp (still am to a degree), but that was when I couldn't see the forest from the trees.

 

[NorthStar]

I see only goodness all around, nothing but just greatness everywhere.

* Blues ? music is playing right now, ...on the analog radio. :b

 

[TBone]

I am confident though that no matter how much progress they have made, digital still leaves analog in the dust as far as pure numbers.

No doubt ... not just the TT, every component within the turntable chain has similar issues; I was shocked by some tonearms cumulative decay spectrums (miller audio research) which looked atrocious.

 

[NorthStar]

1. ? Link number one deleted (no name with the article - not good etiquette in our hobby).

2. ? [B]www.npr.org/2012/02/10/146697658/why-vinyl-sounds-better-than-cd-or-not[/B] --> Link number two, with Dr. Sean Olive.

3. ? www.soundonsound.com/sos/feb10/articles/analoguewarmth.htm

4. ? www.mix-engineer.com/audio-philosophy/digital-vs-analog-recording/

5. • http://skeptoid.com/episodes/4303

6. ? www.quora.com/Audio-and-Acoustical-Engineering/What-is-the-highest-fidelity-medium-ever

7. ? www.quora.com/Is-it-true-that-record-players-have-the-best-sound-quality

Ok Amir, link number one is gone here. ...You sure it should be completely wiped out of the entire Internet?

If you have time*, perhaps you can comment, briefly, on the second link?

* Most people don't (won't) take the time to read too much stuff; first, life's too short to spend too much time on audio forums, and second, us humans we love to learn what we already know (keep repeating it, for a change). ...No? :b