Accuracy Is Not the Answer

Before 1982, when the Compact Disc arrived, I didn't love LPs. Analog was already very old tech, and while every trick in the book had been applied to turntables and LPs, they still wowed & fluttered at 33 1/3 revolutions per minute. Vinyl's deficiencies were legion: warped LPs were more common than truly flat ones; surface noise, clicks, and pops sang along with the tune; LPs rarely had perfectly centered spindle holes; inner-groove distortions popped up at inopportune moments; and each time an LP is played, its sound quality degrades, if only ever so slightly. The LP format? Imperfect sound forever.

So in the days leading up to the introduction of the CD, digital looked like an instant cure for all of analog's ills. It had to sound great. We would finally have a truly quiet format with dead-on speed accuracy, wider dynamic range, razor-flat frequency response, and no wear issues. Digital's implicit promise was of a 100% transparent recording medium that would add to and subtract from the signal nothing at all. Digital had no sound. Music was about to be liberated from analog's gross colorations.

Or that's what I wished for. A few minutes into listening to my first CD (footnote 1), my heart sank. It certainly sounded different—but not dramatically better than an LP. I was confused. Why didn't this hyper-accurate new format produce more realistic sound? Maybe some analog distortions still lurked in the pits of my shiny new CD? Yup, that was it—a lot of early CDs weren't pure, all-digital recordings. Recorded and mixed in analog, they were only mastered digitally: AAD. Whatever, my highly imperfect LPs sounded better. I smelled a rat.

I had to wait a little longer to hear an all-digital—recorded, mixed, mastered—CD. But when that day arrived, my hopes were again dashed. A DDD CD was no better than an AAD or ADD CD. Not only that, pure-digital discs weren't all that much quieter. The low-level noise was still there, but this time it wasn't tape hiss or record-surface noise—it was mike-preamp noise, or the ambience of the recording venue. It was clear then, and it's still true: LPs' musicality trounces CDs'. If anything, my pro-digital bias should have favored CDs, but their sound couldn't hold my attention.

No matter—CDs were hugely popular in the broader market, and a good number of audiophiles loved 'em. But naysayers thought "Red Book" CDs were deeply flawed from the get-go. Many were convinced that the Red Book sampling rate of 44.1kHz was much too low, and that the CD standard's 16-bit resolution was inadequate. That's why CDs sounded so . . . digital. The analog faithful judged the CD as unworthy of true high-end status. I never felt CDs sounded bad, just that they failed to deliver the promised great leap forward in sound quality.

Ten years later, with the debuts of the two higher-resolution digital formats, SACD and DVD-Audio at the turn of the century, it was déjà vu all over again. They clearly sounded better than CD, but didn't really get us closer to more believably realistic playback at home. If I had any lingering doubts that two-channel stereo was the reason recorded music never sounded realistic, the small but steady stream of 5.1-channel releases weren't any better. It took a while to dawn on me, but it was becoming increasingly clear that there was absolutely no correlation between higher-resolution digital recordings and more viscerally realistic reproduction of music. Differences, sure, but sound-quality preferences are just that: subjective preferences. Analog playback may be highly imperfect, but it still has that special something digital never quite achieves.

Then they said that jitter was to blame—digital was perfect, but jitter was messing with the steady flow of zeros and ones. We're quick to find a bugaboo—the one thing that's ruining the sound—and to pin our hopes on the new fix that banishes or reduces the gremlin of the moment. Well, the CD eliminated all of the LP's problems, and still we didn't have perfect—or even better—sound.

Next? Room acoustics were obscuring the sound, so a number of room-correction systems were developed to tame the worst offenders. But they can't make the room disappear. Then came those for whom the prime villain is speaker-cabinet resonances, and some manufacturers responded with massive, structurally inert behemoths.

Again and again, as we rush to identify the reasons why the sound isn't better, engineers poke around for a fix and find solutions. That's great, but it's the LP-vs-CD argument repeated ad infinitum. We're not really getting anywhere. We still don't have hi-fis that can fool the ear into believing we're hearing a live violin or solo voice or rock band or orchestra.

Maybe it comes down to this: Making gear that's more accurate and/or measures better isn't the same as making better-sounding gear. Today's best gear can play louder, with lower distortion, and has wider bandwidth than the best of yesteryear's "Recommended Components." That's true, but a hi-rez file of a new recording can't match the bloody realism of a 1960 RCA Living Stereo LP played through a well-set-up turntable, 1980s-era electronics, and a pair of Quad ESL or Klipschorn speakers.

We must have missed some essential aspects of sound reproduction. I have no idea what those aspects might be, but there has to be more to the pursuit of ultimate fidelity than eliminating or reducing imperfections. We need to learn more about capturing and reproducing the gestalt of music. Analog may be far from perfect, but it still seems to convey more of that hard-to-define stuff that brings recorded music to life.



Footnote 1: Sorry, I have absolutely no memory of the title.
Share | |
Comments
FSonicSmith's picture
"I gather then that you are

"I gather then that you are not included in neither of the two groups."

Mr. JohnnyR, you genius, the above is a double negative. The word you meant is "either" and not "neither". Call "bullshit" where it is appropriate, sir. That would be the mirror. Or look at it this way. YOU are in NEITHER camp, the subjectivist or the otherwise intelligent objectivist.

Now before you get all mad and threaten to hunt me down etc, let's get down to some real business. When was the last time you heard a clean well pressed record on a really good vinyl rig? Just where does this belief that all records are full of ticks and pops come from? I rarely hear clicks and pops when I play my records. Occasionally, sure. But they are not nearly as predominant as so many of you digital-only seem to believe.

returnstackerror's picture
its not that simple

12 years ago, I started getting into Jazz esp the stuff from late 1950's to end of 1960's

 

I like the music but what enhances the experience is the simple recording chain that existed back then… with less knobs to tweak, less channels to manage… the music breaths more freely.

 

So to me,  there is this link between the recording process and the medium that is hard to seperate.

 

Does old vinyl sound so great because its vinyl or because of the simple recording chain.

 

Does new vinyl off old masters sound so great due to the masters simple recording chain.

 

Note: I am not saying its better than digital,  just that its not as crappy as some people will have you believe.

 

Does digitial (say a CD)  sound worse to some people, not because the medium itself (CD) is bad but because of the lack of simplicity in the overall recording chain (recorded at X, down sampled to Y, downsample not an multiple of the original resolution, 48 or more available channels as opposed to say 4, lots of mics as opposed to a handful etc).

 

My vote says vinyl/cd/dsd/dvd-a/bluray are all excellent... each has supposed technical limits (or not depending in your view point) but to me what makes or breaks it is the care in the production of the music that is stuck on the medium… not the medium itself.

 

Peter

 

 

Sultan's picture
Human Emotion

If one views a painting that is so beautiful it makes the viewer cry then the viewer is experiencing a response to visual stimuli. How does one measure that process? Similarly, how does one measure the reaction to beautifully reproduced music by only concentrating on sound waves or distortion?

ultrabike's picture
That is a good point. If

That is a good point. If beautifully performed music makes the audience cry, life is good!

If the same audience wants the experience again, and hears the same music being reproduced through a rig with a FR range of 168-2000Hz, and with distortion levels above 20%, then that audience may probably cry for a different reason...

MikeMercer's picture
I'm with him

I enjoyed this essay thoroughly.  I used to think more analytically when it came to stereo playback, but a recording session with Arif Mardin changed my view of Hifi from that perspective forever!!  Gotta find it, but I wrote about it in my review of the Zu Audio Omen Defs for Positive Feedback...

Sometimes the magic is in the imperfections (and I'm not talking muddiness or lack of clarity here - jeez I just posted about something similar on Head-Fi), the soul of the music.  Some gear that I love isn't the most tonally or dynamically accurate (using my own record as a reference) but it excites me!  Same w/ formats.  I don't discriminate.

DetroitVinylRob's picture
><((((º>`·.¸¸.·´¯ `· .¸.·´¯`·...¸><((((º>‏

Nice article, funny how divisive a personal preference of format can get. Niether are perfect and beyond all the science and numbers I return to the sound that has rewarded my mind, my body, and soul with beautiful music for over fifty years. A lot can be said about what is wrong, in my experience there is far more to focus on about what is right. So JohnnyR, you know they have doctors that can remove that huge bug that has crawled so far up your bum and give us all a bit of relief, really.

Happy Listener! ;^)>

JohnnyR's picture
Awwwwwwwwwwww

You stil holding a grudge about my other posts where you showed how little you really know? I guess anyone that doesn't fall into step and salute the fearless leader is suspect and "wrong" Look for a doctor yourself that can remove that rock between your ears that you call a brain.Funny how people make it personal when they are lacking in the sciences.I suggest you spend more time reading and studying how the real world works instead of clinging to outdated methods of music reproduction. Maybe watching too many Flintstone cartoons when you were little with the bird beak used as a phono stylus made you this way.

komo's picture
"but a hi-rez file of a new

"but a hi-rez file of a new recording can't match the bloody realism of a 1960 RCA Living Stereo LP played through a well-set-up turntable, 1980s-era electronics, and a pair of Quad ESL or Klipschorn speakers"

how can vinyl with small dynamic range that sounds "fat" and less detailed can produce bloody realism ? this is a very subjective opinion.

"We must have missed some essential aspects of sound reproduction."

i agree with this. look at the process : every musical instruments and vocal was separately close mic-ing recorded, all tracks then mixed with all kind of effects, overdub and heavy equalization. finally in the mastering process there's a process which is the biggest enemy for audiophile : multi band dynamic range compression. so with these all kind of processes don't expect that your speakers magically "disappear" in front of you.

"I have no idea what those aspects might be, but there has to be more to the pursuit of ultimate fidelity than eliminating or reducing imperfections."

i thought you are an audiophile ?

dsmalle's picture
audiophiles are subjective by design?

Analog, digital, it must be a matter of taste I guess. No matter what technical solution is used to store and reproduce sound, analog is the source and analog is what your ears require. If inbetween source and ear the music is scratched in shellac, stamped in iron or casted in a thin aluminium layer, what does it matter?

If the reproduction matches the source and you cannot make the distinction, then the medium does not matter. BUT the distinction is always there. You're sitting in a comfy chair with excellent headphones or with very expensive speakers listening to a band you know can't fit in the room you're in (don't even think about the symphonic orchestra). Neither do you have the atmosphere of the concert or the recording studio. 

Do we replace this with nostalgia to older media? And are our ears so perfect that we can hear almost unmeasurable differences? At the age where we can afford a stereophile's equipment, most of us don't have the perfect hearing anymore. 

It's not the ears. It's not the LP or the CD. It's all in what's between the ears...

wink

earwaxxer's picture
now I get it!

Now I get why Stereophile has Steve writing blogs. He gets the responses! Check out the other articles on this site. Maybe 1,2 or no comments. This one has over 40. Advertisers pay for hits.

Steven N's picture
Thank you, Steve, for

Thank you, Steve, for reiterating what some may consider “rubbish,” however speaks to the classic (and ongoing) quantitative - qualitative dialogue.

I do not know what the future holds for the capabilities of a quantitative approach as one able to account for qualitative discernment and the ascription of value.   The scientific method was not developed to answer the question of “why” things function as they do; rather, only “how” things function—and then only within the limits of the testing apparatus and the variables (controlled and confounding) within the testing environment.

The manner in which our senses are marketed to becomes disconcerting, insofar many will internalize this marketing (e.g., “perfect sound forever”) as truth and, consequently, disavow their individual capacity to discern, to decide for oneself what is valued at the particular moment in time, within the conditions one is situated (a potentially daunting task in and of itself to account for relative to one’s personal history and understanding thereof).

While we can be lead astray by our senses (the same senses used to develop test instruments and create meaning from the data they produce), still, they exist, for better or for worse to teach us, to inform us, to facilitate internal pathways to deepen our understanding both of our outer world and inner experience.  (Thank you, Mr. Hendrix).

I trust you will be at RMAF next week, Steve.  It’ll be good to see you.

Ciao4now,

Steven N

Monkish54's picture
Maybe not for you...

but accuracy is very much the answer for me. I've noticed a pattern in your write-ups and interviews. You seem to be anti-accuracy. You're one of the Audiophiles (Audiophools?) who don't look for accuracy, but instead look for enjoyment. I wouldn't say anything is wrong with that, but when you try to convince others to abandon the WHOLE IDEA OF HIGH FIDELITY, I think it's time to draw the line. Accuracy is what HIGH FIDELITY is all about!! If you want speakers that will make you happy, i'm fine with that. Why would I care? But when you try to convince people that accurate sources are bad, I start to question your articles. 

Audio Asylum Bruce from DC's picture
vinyl vs. digital

I suppose if you owned a lab, you could have as a goal owning playback equipment that produced the best measurements, even though that would seem kind of absurd.

But most people don't own a lab, so therefore most of them must buy stereo playback for some other reasons . . . probably as a source of enjoyment.

I own several linear yards of recorded audio in analog (vinyl) and digital (RBCD, SACD) format.  As a general matter, I would be hard pressed to say one format sounds better than the other.  I have very good sounding digital recordings and very good sounding records.  As a general matter, I would say that SACD sounds better to me than RBCD.

CD measures perfect, when the measures developed for vinyl are applied.  But that doesn't mean CD sounds perfect.  There are measures applied to CD (which have no relevance to analog systems) which show that it is not perfect.  How these measures relate to what we hear is less clear than how analog measurements relate to what we hear.  But they are not irrelevant.

Inexpensive analog can sound surprisingly good and lifelike in a way that inexpensive digital does not.

And for people obsessing over dynamic range, take an SPL meter in your favorite listening room and measure the ambient noise level.  If you are very lucky and live in the sticks, it will be 40-50 dB.   So, with that as your "noise floor" consider the merits of 60 dB dynamic range above that floor, the upper limit of which is actually dangerous to your hearing.  And also consider the linearity of your loudspeakers and their ability to reproduce that dynamic range without going into gross harmonic distortion, which will alter the timbre of the sound you hear.

If you do that, you will see that the dynamic range of a vinyl record is fully adequate for real world home reproduction.  In fact, very few recordings exploit that full dynamic range.

I would agree with you that vinyl playback's weakness is short and long term pitch accuracy.  Short-term pitch accuracy ("flutter") blunts the impact of transients, especially bass transients.  Long-term pitch accuracy ("wow") is more obvious and less tolerable.  However, reasonably good vinyl systems bring these down to a very low level.

By contrast, artifacts of digital reproduction (ringing of digital filters, jitter) introduce non-correlated noise into the signal, which is audible at extremely low levels.  And digital reproduction of high frequency transients is often pretty messy . . . just listen to how a cymbal or high-hat on a drum kit is reproduced.

In any event, digital vs. analog is a false choice.  Lots of legacy material is available exclusively on records, which can be surprisingly durable.  And, as digital moves beyond the physical disc into computer-type storage systems feeding a DAC, digital offers unparalleled convenience in both storage, access and acquistion of new material through downloads. 

jlaubza's picture
indeed, accuracy is not the answer

It has been fun reading the views expounded above - ears and perception are obviously different between people and each of us seems inclined to believe that our perception is reality - a somewhat Cartesian view of the world - but we should know better than that.

But digital sound is often different from analogue sound - that seems to be a fact and some prefer the rounded edges of analogue sound - I recall tests of square waves played back on LPs and on CDs. LPs can't handle the edge of a square wave - these are rounded off whereas with CDs, the edge can be seen on an occiloscope and obviously also heard.

Of course music doesn't generate square waves but rapidly rising transients are accurately reproduced by digital media whereas some rounding off of the transients occurs with analogue media.

And it seems that some like the rounded off sound, others like the sharp, accurate sound. Personally, I like my music without Rice Crispies - you know, the snap, crackle and pop of an electrostatically charged vinyl disc.

But what I have noticed, as my sound system has been upgraded, is how badly recorded many CDs are (and these include AAD discs) - I sometimes wonder if that is a reason why vinylphiles are so convinced their groovy systems are better - they simply don't hear how badly the recordings were made whereas with CD playback, recording deficiencies stand out more clearly.

audiofeelya's picture
The REAL Problem

I agree that analog has some positives as does digital. Now, the REAL problem is the consumer audio business model. which is.. offer the consumer marginal improvements (being generous, many products offer zero improvement.) and play off the placebo effect as it has been discovered that a persons perception has a major impact on subjective subjects (describing what they can and can't hear) instead of striving to achieve a truly considerable improvement from update to update. This deception game is so organized with expert opinions (no offense Steve I know a man's gotta eat) .and pseudoscientific nonsense. They have been able to steer clear of the video industry's "problem" having to actually produce results due to the fact that video is seen. There can be less deceiving  in general, what you can see is more concrete to us and much less subjective. So the real answer in producing better reproduced music is to stop "buying" (literally and figuratively) the sizzle and demand the steak! So if you have a good sounding system (to you) currently, don't bother trying to get it just a little better with overpriced gear that depends on the placebo effect and the return/exchange limit for the seller to stay afloat due to such marginal returns on performance. We need to allow non innovative companies to begin to depend on volume and majorly discounted prices if they can't offer better products or they can go under. if they can't move forward and we don't demand it, we will be missing out on what is truly possible. So until the lies stop from audio manufacturers and audio mags and fake reviews, we will continue to have the same argument for the next 25 yrs . It gets boring and obvious and thats why traditional home audio is such a sluggish sector. Just think, the only considerable innovation has been convenience, thats it! and it wasn't the audio companies that created it!demand more! stop the lies! Join the Audiolution! 

walshbouchard's picture
Its all subjective humbug

I Agree with John Atkinson

 

I am old enough to remember quad in fact, I remember first hearing it as 4 channel stereo in the UK by AR using 2 stereo Reel to reels in sync playing back though AR4s, I think, I was 17 and walked out of that demo thinking hi fis got a great future

That was 1970, 43 years later and hi fi snobs are still banging on about stereo being three dimensional etc.

NO IT ISNT. I admit some recordings can sound pretty good but it’s still only two channel

It’s alright for someone who can afford a couple of PMC BD 5 studio monitors and a £100,000 worth of amp and sources , probably does sound 3 dimensional then ? But for most of us who are limited to the real world of a £10,000 often second hand used system,

NO IT Doesn’t!

 

Not to diverge but it’s the same agreement with CD 44.1 and DSD or 24/96 PCM

DTS CD is better than CD , Vinyl is better than CD SACD is better than  DTS CD and Vinyl, DVDA is better than DTS CD and all of them are better than CD!

Why is there a debate over this? it’s like arguing in favour of a flat earth or creationism

Originally stereo was first conceived to be three channel , I have a load of 3 ch RCA Victor Living Stereo DSD SACDs and these are just bloody amazing even although the tapes are over 50 years old  Reiner’s Scheherazade being one of them . the third channel makes all the difference both in resolution and imaging , of course it does the existing two speaker have less to do , so what they do have too they do better

However nobody would wear 3 speakers and it was probably not possible at the time to put 3 tracks on a vinyl disc, so we have been living with a half-finished product for 50 years anyway,

The fact is the human ear in a concert hall or rock concert picks up sounds both direct from the source and reflected from over the shop and it’s pretty obvious that stereo while an improvement on mono can’t do that , multichannel can and does

 

End of argument!!

They reason multichannel has failed is because everybody is an MP3 freak and it’s hard to put four channels on an IPod  , about ½ % care about quality of sound  , it’s just market forces and practicality  .

If those figures were the other way around nobody would be supporting CD or Stereo

wandique's picture
hlaB, hlaB, hlaB

Vinyl LPs are better.

acuvox's picture
Absolute Ears

First, I hope we all can agree that there are aspects of sound that machines can measure far more accurately than we can hear, like steady state sound intensity to .01dB or frequency to .000000000001%.  

Secondly, there are aspects of sound that human hearing can discern far beyond the capabilities of machines.  For example, there are blind bicyclists who can ride between parked cars and moving traffic by hearing the reflections, diffractions and sound shadows of the stationary objects, even recognizing stop signs by the sonic signature of the octagonal shape.  

This is unarguably objective.  Either they ride safely or not, and there are no DSP algorithms that can keep up with that real-time performance.  If you want to contest this, I welcome you to ride on the bicycle the computer is steering by sound to prove your point.

Where it gets sticky is the adaptability of ears.  Surely if perception is so plastic there can be no absolute and all opinions are equal.  I beg to differ.  

We learn to hear over time scales from milliseconds to a lifetime.  The latter is the most interesting, because from birth to puberty our brain grows by adding neurons, and it grows auditory processing capability in patterns and proportion that are trained rather than genetically pre-determined.  The circuits create and wire themselves according to what is useful.

Recent studies show that conservatory trained acoustic musicians have larger volumes of spatial processing and white matter in their brains.  Their musical intelligence and spatial intelligence is developed physically to a higher degree, just like athletes have more muscle mass they have more neurons.  They listen intently to the sounds of acoustic music from childhood and carry a current memory of precisely what music sounds like and what it MEANS.   Unlike the audio testing of post-industrial "average citizens" by Edison, Bose, Sony and Toole, the musicians have never been satisfied by reproduction systems.

I have been conducting informal audio surveys for the last fifteen years, and the responses of acoustic musicians are tightly grouped with individuals raised in ex-urban environments with no speakers and motors.  I believe I have indentified the characteristics of Naural human hearing, and it points to quite different results than audiology and audio since 1932, in particular more consistency between subjects and an emphasis on spatiality, phase, transient response and time that are in the gray areas of audio measurement.  OTOH, audiological data says that humans are phase deaf.  

This is the resolution: humans with hearing developed using speakers in post-industrial background noise levels are phase deaf because speakers mangle phase and have uneven off-axis response.  Noise masks echoes so the mapping of envoronmental geometry by sound is not developed.  It gets worse if you listen to decimated sound (digital words), digital filters with ringing on transients, digital reverb and/or digital compression.  Individuals raised on acoustic sounds free of motor vehicles and motorized appliances use phase of transient reflections as the predominant determinant of acoustical spatial perception using the phase encoding of their pinnae.

Acoustic musicians have shown a preference for mechanical phonograph (Edison and Berliner) over electronic, LP over CD and find digitally compressed music (MP3, AAC and streaming) unsuitable for listening to music.  Perception of time and phase to microseconds in transient analysis and discrete decoding of echoes explains this.  Note that the physical frequency repsonse limit of the ear mechanism does not set an upper limit on time resolution, rather our hearing is optimized to supercede the frequency limit  by over a factor of ten to extract more useful information.

There are still issues to work out in DSD, but it is the format that potentially has flat frequency and phase response, minimal distortion and time resolution greater than human perception.  The latest reports are that double and quad DSD offer improvements, possibly from the relaxed filter requirements.  

The boutique record labels that are run by or for conservatory trained acoustic musicians offer SACDs and keep the signal pure.  Note that no mixing or mastering operations can be performed on DSD, it has to be decimated for operations as simple as a splice so major label classical SACDs are compromised.  Even the much vaunted Sonoma workstations must convert to PCM at the splice points.

I see this as a feature of DSD.  If you can't perform time smearing mathematical operations, the only production method is to do all sonic adjustments acoustically.  With acoustic mixing, balancing, spectral adjustment, peak limiting, spatialization and reverb, you can create an optimized acoustic event and capture it live.   This is the closest to absolute sound.

spacecase's picture
To OneMic:

I liked your original comment and thought the conversation went downhill from there.

I have not heard this perspective on vinyl in the time domain before--do you know if anyone else has written about it?

If I understand correctly, part of what you are saying is that since there is more than one aspect to sonic accuracy, the blanket statement that one medium or another is "more accurate" is not very accurate.  That observation would neutralize a lot of the vinyl vs. digital bickering on these forums and make them a lot more interesting to read...

Site Map / Direct Links