The digital difference – Can digital audio sound different?

[b_panther_g]

OK, this is something that I really want solved once and for all...

Can 2 well designed CD players sound different? If so, why?

Does anyone know about DACs and digital filters? Can someone give some incite about the variables (if any) that can affect the sound?

Experts this is your time to step up and show what you know. Just remember that we aren't all electrical engineers, so please keep it as simple as possible.

Personally, I believe they can sound different. But I am only human and this is hi-fi. Some people claim to hear all types of things that I think are nuts (the magic CD chip always makes me laugh).

Please share your opinions. But let's keep this civil, OK.

Thanks.

 

[Rip Van Woofer]

OK, this is something that I really want solved once and for all...

Can 2 well designed CD players sound different? If so, why?

No, if by "well designed" we mean it has flat frequency response and low (below the threshold of hearing) distortion. (Of course, you knew I'd say that! )

Does anyone know about DACs and digital filters? Can someone give some incite about the variables (if any) that can affect the sound?

In another thread I posted a link to a very good, understandable series of articles on digital audio. They should answer most of your questions:

Digital Audio: Theory and Reality

Also look thru the Audioholics articles. There was one on "brickwall filters" that was good.

DACs (Digital to Analog Converters) vary in measured performace to be sure, but at the current state of the art even most (if not all) of the cheapos are audibly transparent. Interestingly, an outboard DAC increases the chance (albeit slightly) of clock jitter-induced distortion. Something many of the high-end types who think a high priced outboard DAC is the cat's heinie seem not to realize.

 

[mtrycrafts]

OK, this is something that I really want solved once and for all...

Can 2 well designed CD players sound different? If so, why?

Does anyone know about DACs and digital filters? Can someone give some incite about the variables (if any) that can affect the sound?

Experts this is your time to step up and show what you know. Just remember that we aren't all electrical engineers, so please keep it as simple as possible.

Personally, I believe they can sound different. But I am only human and this is hi-fi. Some people claim to hear all types of things that I think are nuts (the magic CD chip always makes me laugh).

Please share your opinions. But let's keep this civil, OK.

Thanks.

Not much to add to Rips input. If they are audibly different, the specs or measurements will show it to be. Yes, measurements do correlate to audibility issues

 

[Rob Babcock]

I say yeah, they absolutely can sound different. Many machines are even designed with euphonic non-linearities built in. Can two good players be designed to sound identical? Yes, just as certainly.

 

[tbewick]

I'm as puzzled as b_panther_g. I don't really understand Rob Babcock's statement. I looked at a Marantz CD player and quoted frequency response was 20Hz-20kHz +/-1 dB (ie. highly linear).

My dad's LP's on his top-end Thorens record player all sound great but virtually all my CD's stink in terms of sound quality, be it using a 24-bit DAC or 1-bit MASH.

I think I'm switching to vinyl - most CD's are just sound rubbish, whatever the specs say. I find that applying some tonal adjustment is the only way to get the most out of a lot of my CD's. The only CD that springs to mind just now as sounding excellent is the Grandaddy album 'Sumday'.

 

[b_panther_g]

Hello Rip,

Thanks for posting the link to the articles. I've read them but they left me with more questions. Such as...

Interpolation Filters...
According to the article, the common interpolation filter is a FIR filter. The article says that complex FIR filters are extremely accurate but require more computational power. It also said that CD play back doesn't require complex FIR filters. But (here's what I'm wondering)...

If someone “breaks the rules” and uses a more complex FIR filter to more accurately represent the analog waveform, that should have an audible affect on the sound. Right?

Please let me know if I'm missing anything. Thanks.

BTW
I agree with you when you said...
No, if by "well designed" we mean it has flat frequency response and low (below the threshold of hearing) distortion. (Of course, you knew I'd say that!)

I just wonder if we have ever reached that point in the real world. If we had, wouldn't a well recorded CD of a solo instrument sound identical to the real thing (let's assume it's a mono recording being played back on a very accurate system)?

Enjoy,
Panther

 

[Rip Van Woofer]

If someone “breaks the rules” and uses a more complex FIR filter to more accurately represent the analog waveform, that should have an audible affect on the sound. Right?

Please let me know if I'm missing anything. Thanks.

Only if the less complex filter has errors that are above the threshold of hearing.

The thing to keep in mind whenever discussing questions of whether this or that component or circuit will sound better is that human hearing has limitations, regardless of how "golden" one's ears are. And those limitations and what is or is not audible have been well established by decades of scientific work. If a given component's measured errors are below those thresholds, it's transparent. It has no "sound" of its own.

Contrary to frequent claims by some, there are no audible differences that cannot be measured. 30 or so years ago that was the case, but no more.

Put another way: at the current state of the art the limiting factor for modern, well designed purely electronic audio components is ususally our ears, not the component!

"Purely electronic" does of course exclude electromechanical devices like speakers, phono cartridges, and microphones. All have variations in frequency response and distortion that are demonstrably audible.

 

[Rip Van Woofer]

I think I'm switching to vinyl - most CD's are just sound rubbish, whatever the specs say. I find that applying some tonal adjustment is the only way to get the most out of a lot of my CD's. The only CD that springs to mind just now as sounding excellent is the Grandaddy album 'Sumday'.

Do you listen mostly to pop/rock, etc? If so, there might be an explanation.

Past Audioholics threads and maybe even an article or two here have dealt with the problem of "hot" recording. Record producers have been insisting on louder, heavily compressed mixes that basically clip a lot. The idea is to grab your ears when it's played on the radio, I guess. The engineers, who generally know better, get overruled by the suits. It's been a subject of discussion in recording circles.

Sometimes the artists and engineers can prevail. Maybe that was the case with the Grandaddy record.

Since CDs have all that lovely flat response up to 22kHz and since clipping distortion is most objectionable at high frequencies, you get "ouch"! Vinyl rolls off at the top frequencies so it's kind of like turning the treble down. It's more "forgiving" of hot mixes.

(This is a simplistic explanation. Do a search to find out the gory details! I think Rob Babcock and Dan Banquer have posted on the subject with links to relevant articles.)

So...it's not the medium, it's the way it's used...or abused.

FWIW, I listen mostly to classical where this is less of a problem. Even when the CD is a reissue of old analog recordings, I do not feel any vinyl nostalgia. Indeed, a good CD reissue of a recording I already have on vinyl is ususally a revealation for the better.

I've still got my old Thorens, btw, for my old irreplaceable vinyl. Great tables!

 

[b_panther_g]

Hello again Rip,


I understand what you are saying...

As long as the errors are below the threshold of humans' abilities to hear them, it doesn't matter how much more accurate the device (or filter or whatever) is.

I agree wholeheartedly...
BTW, when I speak about errors, I am also referring to omissions. A digital filter can reproduce a representation of an analog waveform. But does that representation have the resolution to be audibly transparent?

This leads to the bigger question... If the current state of digital recordings can capture everything that we can hear, then why is there a difference?

As good as our audio systems are, there is still a difference between them and the real thing. Also, I'm not talking about a system with 2 or 5 or 10 speakers that's trying to reproduce an entire symphony. I'm referring to a relatively simple task. A system that sounds as real as a single piano playing softly.

I am aware that most of the problems lie with the speaker(s) and room acoustics. But I'm not willing to say that the recordings themselves are perfect. Nor am I willing to say that the players are perfect either.

Now if my ramblings actually make sense (LOL ), I'd love to hear your thoughts.

Do you think CD and CD players are perfect (or are audibly transparent)?

If so, do you believe all our efforts should be placed in improving the mechanical aspects of recording and reproduction?

If not, where do the deficiencies lie?


Thanks again.

Enjoy.

 

[WmAx]

This leads to the bigger question... If the current state of digital recordings can capture everything that we can hear, then why is there a difference?

As good as our audio systems are, there is still a difference between them and the real thing. Also, I'm not talking about a system with 2 or 5 or 10 speakers that's trying to reproduce an entire symphony. I'm referring to a relatively simple task. A system that sounds as real as a single piano playing softly.

I am aware that most of the problems lie with the speaker(s) and room acoustics. But I'm not willing to say that the recordings themselves are perfect. Nor am I willing to say that the players are perfect either.

The reason a recording will not sound exactly like the original event is that there exists NO STANDARDS for recording and playback. In a lab controlled environment, you can easily replicate(over speakers), as has been demonstrated in the past by John Dunlavy of Dunlavy Labs. But when you enter the non-standardized world, it's just not possible. Note that I am referring to reproduction, not realism. I have heard real sounding playback over a stereo system, but it does not have to be an accurate representation of the original event in order to sound real. Example: Recording was made in music hall A. When played back over certain stereo system it sounds real -- but it sounds like it is being performed in music hall B as opposed to the correct music hall A. It sound real, but it is not true to the original performance. Additionally, because when you typically go to a live music event, you can see, you are accustomed to the visual data input. Removing this factor(visual), even when the sound component is correct, may result in a different experience that may never seem 'real' to a specific person if they can not concentrate on just the sound.

-Chris

 

[miklorsmith]

Dissenting View

Great thread! For the flip side of the coin:

CDP's sound different. Mine has tubes in it and different tube types sound different. My burned, CDR's sound different from the originals. Audioheads worldwide sing the praises of vibration control and quality AC power. Non-OverSampling DACs are being talked about as being more "analog" sounding than upsampling or oversampling strategies. Hard-drive-based players have the potential to eliminate virtually all jitter.

These are factors. Yes, there will be an avalanche of argumentative views that say a $50 Samsung player sounds identical to one by Ayre because there's no hard data to show the difference. This is a convenient viewpoint because it allows for very cheap acquisitions without loss of sound quality and relegates the source to non-variable status, i.e. not needing of consideration/contemplation/upgrading/expense.

Why doesn't a piano sound like a piano? Because multi-driver systems, with their inefficiencies and crossovers that live where the piano plays kill the life and dynamics of the piano. And everything else, for that matter. Find a well-implemented, single-driver-based arrangement and see if you don't agree.

 

[WmAx]

Great thread! For the flip side of the coin:

CDP's sound different. Mine has tubes in it and different tube types sound different. My burned, CDR's sound different from the originals. Audioheads worldwide sing the praises of vibration control and quality AC power. Non-OverSampling DACs are being talked about as being more "analog" sounding than upsampling or oversampling strategies. Hard-drive-based players have the potential to eliminate virtually all jitter.

These are factors. Yes, there will be an avalanche of argumentative views that say a $50 Samsung player sounds identical to one by Ayre because there's no hard data to show the difference. This is a convenient viewpoint because it allows for very cheap acquisitions without loss of sound quality and relegates the source to non-variable status, i.e. not needing of consideration/contemplation/upgrading/expense.

Why doesn't a piano sound like a piano? Because multi-driver systems, with their inefficiencies and crossovers that live where the piano plays kill the life and dynamics of the piano. And everything else, for that matter. Find a well-implemented, single-driver-based arrangement and see if you don't agree.

Do you buy that rubbish by the truck load?

Nearly everything in your post was unsubstantiated and in conflict with scientific research. I find the claimed superiority of single-driver-based arrangments as being inherantly superior to multi-driver based systems especially entertaining.

-Chris

 

[miklorsmith]

Truckload

Everything's cheaper in quantity.

I know my views are not supported here and that's OK. I just have to chime in from time to time when it seems like somebody is looking for information and gets the same 4 guys every time saying the same thing that isn't representative of what everyone believes. By casting it all as truth and supporting each other, it comes across as being established fact.

And it is, to the purveyors of said thought. My points urge people to do some of their own thinking instead of just relying on someone else's opinions. As such, I think the advice about trusting your ears and not scopes and instruments is especially apropos.

Your condescending attitude towards others and their views cultivates sameness here. If that's what you want, you're doing a bang-up job. Kick out Mule and laugh at anyone with different ideas and you'll Almost have cornrows as far as the eye can see. Interesting to look at? Not to me.

I calls 'em how's I see's 'em. When I heard a 6-watt, single-driver setup kick the Christmas out of my spendier, Class A, multi-award-winning arrangement, it rearranged my thinking.

 

[b_panther_g]

Hey Chris,

Good points. I agree with many things you said.

1st I'd like to clarify something. When I say a reproduction “sounds real”, I mean that it is sonically indistinguishable from the live event.

For example, lets assume that we are in a lab with perfect acoustics. We take the best (most accurate) recording equipment and reproduction equipment available. We record a baby grand player piano, because it plays the exact same way every time. We then make a CD or DVD-A or whatever of our recording.

We now place the best (most accurate) speaker(s) next to the piano and play our recording. Are they going to sound exactly the same? Will a person be able to tell which one is a recording and which one is real?

I say yes, they will hear a difference. Even if we assume that our lab has perfect acoustics, there will still be a difference.

I am not willing to say that the difference in this case is ONLY because of the limitations of the speaker(s).

I know that the acceptable margin of error for speakers is currently about 10x that of other components. However, I'm not willing to say that other components are perfect. If they were, CDs, DVD-As, SACDs, and all future digital media will all sound exactly the same.

Now, if someone proves that the distortion from current speaker technology is so high that it masks most of the deficiencies of current CDs and CD players, that seems more logical to me.

But, I guess until someone performs such an experiment, we won't know just how close (or far) we are from perfect reproduction. (**Cough Gene. **Cough Clint)


BTW, Chris I'm not familiar with John Dunlavy's research. Could you provide a link. Thanks.

Enjoy,
Panther

 

[b_panther_g]

Hello miklorsmith

I think that we are talking about different things.

I'm not talking about an “analog sound.” The analog sound, as it is discussed on the Internet and in publications, refers to the sound of analog recordings. That means the “analog sound” boils down to distortion. Pleasing or not, I don't want any audible distortion - from tubes, DACs, or any thing else.

I'm interested in realistic audio reproduction. That is, audio reproduction that no one can tell apart from the real thing. I'm also interested in CD technology. Has the technology reached it's performance limit? Can any new CD players produce more accurate sounds than current players offer?

These are very different from the things you wrote about. Tubes and single driver speaker technology can not, at this time, compete with other technologies in terms of realism.

Enjoy,
Panther

 

[Cygnus]

Do you listen mostly to pop/rock, etc? If so, there might be an explanation.

Past Audioholics threads and maybe even an article or two here have dealt with the problem of "hot" recording. Record producers have been insisting on louder, heavily compressed mixes that basically clip a lot. The idea is to grab your ears when it's played on the radio, I guess. The engineers, who generally know better, get overruled by the suits. It's been a subject of discussion in recording circles.

Sometimes the artists and engineers can prevail. Maybe that was the case with the Grandaddy record.

Since CDs have all that lovely flat response up to 22kHz and since clipping distortion is most objectionable at high frequencies, you get "ouch"! Vinyl rolls off at the top frequencies so it's kind of like turning the treble down. It's more "forgiving" of hot mixes.

(This is a simplistic explanation. Do a search to find out the gory details! I think Rob Babcock and Dan Banquer have posted on the subject with links to relevant articles.)

So...it's not the medium, it's the way it's used...or abused.

FWIW, I listen mostly to classical where this is less of a problem. Even when the CD is a reissue of old analog recordings, I do not feel any vinyl nostalgia. Indeed, a good CD reissue of a recording I already have on vinyl is ususally a revealation for the better.

I've still got my old Thorens, btw, for my old irreplaceable vinyl. Great tables!

This is why CD's today (or most of them anyways) sound really bad, compared to ones that were made by bands ack in the 70's. Dynamic Range. Now-a-days, CD's dont have much dynamic range which means that everything on the CD is AS LOUD AS POSSIBLE! With a ton of cliiping to boot.

 

[miklorsmith]

Same!

I think that we are talking about different things. . .Tubes and single driver speaker technology can not, at this time, compete with other technologies in terms of realism.

I have never heard a complex, multi-driver arrangement that will play the sounds of piano with realistic dynamic range. It is widely assumed that flat frequency response is what is needed for realistic portrayal of instruments. While most single drivers will not deliver crushing bass or tickling highs, they DO get the midrange incredibly right. They are much faster on the uptake and decay more naturally than other systems.

These are critical elements in sound reproduction. A speaker with "flat" response will sound dead if it is not fast enough with sufficient transient speed to reproduce not only the frequencies of the event, but its timing and intensity. These are the elements where "flat", complex speakers fall short. Do we call this "accurate" nonetheless?

We are not talking about different things at all. I seek the realism of the event, as we all do. I've found that dynamics are more important to instrumental reproduction than the frequency curve of a transducer.

I made this determination immediately after building some Fostex boxes and I'll never go back. Realism and beauty can go hand-in-hand. How to get there is experimentation and following one's own priorities, as compromises are inherent to the search.

If you Really want to hear a piano though, and you don't at least go to a good shop to hear a high-efficiency setup, you are just cheating yourself.

 

[WmAx]

I say yes, they will hear a difference. Even if we assume that our lab has perfect acoustics, there will still be a difference.

BTW, Chris I'm not familiar with John Dunlavy's research. Could you provide a link. Thanks.

Enjoy,
Panther

Unfortunately, Dunlavy Audio Labs articles were removed from the website when the company shifted ownership(and subsequently went out of business). However, I can describe to you the basic test procedures as best I remember them:

-A musician or musicians were recorded in an anechoic chamber using measurment microphones.

-Speakers were set up in a lab environment(basicly in a large space with no room reflections).

-Musician or musicians who were recorded in the anechoic chamber were placed between the speakers.

-An acousticaly transparent curtian was put across the front so that no visual information was available.

-The recording were played, alternating with the actual musicians in order to see if subjects could identify which was the recording and which was the live sound.

The results were that it could not be identified(speakers vs. live).

-Chris

 

[Tom Andry]

Kick out Mule and laugh at anyone with different ideas and you'll Almost have cornrows as far as the eye can see. Interesting to look at? Not to me.

The discussion of the mule was in no way related to his love of a particular speaker or any other belief or opinion that he has stated here. It was simply because his writing style was so obtuse that it contributed little (some say it detracted) from the quality of the discussions on the forums. I don't think it was it suggested that his beliefs and opinions could not have a home here, just that he needed to state them more clearly.

I've been surfing these forums for a very long time and have seen you get in some pretty heated debates. I seem to remember that at one point you indicated that you didn't want to have much to do with this site (though that may have been someone else). I'd like to think that you'd realize that this site is one of the few (the only one I've ever found) that has given a home to those that don't believe in many of the high end audio beliefs (myths to some). While I for one welcome your opinions and continued participation, to suggest that you are being persecuted is akin to a fish complaining that there is not enough water on the land.

Too often a rabid audio nut will spend a week or two blasting the boards and it’s participants for not believing the intelligent chip (or magic rocks or whatever) generally contributing little to the discussion. You are well spoken (or typed), think through your opinions, back up your opinions with at least personal experience, and generally treat others with at least the same (if not greater) amount of respect that they have shown you. I have the impression that if an experiment that was conducted by Gene or Clint contradicted your beliefs and experiences, you would at least reevaluate your opinion. If some posts have hurt your feelings or seemed too harsh, I hope that you look past it and continue to contribute. If you decide not to, I’m sorry to see you go. I’m sure the wheat over at Audio Asylum is just a boring as the corn here.

 

[WmAx]

E I just have to chime in from time to time when it seems like somebody is looking for information and gets the same 4 guys every time saying the same thing that isn't representative of what everyone believes. By casting it all as truth and supporting each other, it comes across as being established fact.

I base advice on the state of knowledge of scientific research concerning the subject(s). So, of course when others tout things as if they are established and/or equal with that of scientific research, of course I may respond as you accused:

Your condescending attitude towards others and their views cultivates sameness here. If that's what you want, you're doing a bang-up job. Kick out Mule and laugh at anyone with different ideas and you'll Almost have cornrows as far as the eye can see. Interesting to look at? Not to me.

I calls 'em how's I see's 'em. When I heard a 6-watt, single-driver setup kick the Christmas out of my spendier, Class A, multi-award-winning arrangement, it rearranged my thinking.

Nice testimonial. However, you seem to have failed to recognize the difference between what it takes to establish someting as established fact vs. opinion/testimonial.

-Chris