Why does vinyl (analog) sound better?

I'm looking for a technical explanation, here.

According to the principles of Nyquist, the standard CD digital sampling rate can faithfully reproduce frequency content up to 22.05 Hz. Audio frequencies above something in the 17kHz range are inaudible.

So why does vinyl sound so superior to CD? By the way, if you disagree with this basic premise, there's probably not much you can contribute to this post. I accept it as fact.

- Ivan

p.s. I'm a CD guy, because I don't like vinyl surface noise, and vinyl is so, well, "fragile". But still, I acknowledge that vinyl sounds better.

 

This is a case by case deal as discussed on the forum ad nauseum. I played Gordon Lightfoot's "Summertime Dream" vinyl once and I was underwhelmed with the sound quality, and the CD actually sounds better in this case.

 

Because God made it that way.

 

The generalization is wrong. Some vinyl does indeed sound better than a CD, but most doesn't. A properly mastered CD played on a decent CD player sounds fantastic! BTW what are you playing your CDs on?

 

Well said. In some cases, vinyl is superior, in others....not really. To me, Brothers In Arms sounds better on vinyl......Bat Out Of Hell on cd is...ummm...less crappy than the vinyl.

 

While Amy Grant's "Lead Me On" album is great sounding on either vinyl or CD, the same with her "Straight Ahead" album.

 

Dennis,

Good question... I need to fill in my profile. Until I do, I have a Marantz universal player, a Cambridge Azur integrated amp, and Pinnacle bookshelf speakers. The model numbers elude me when I'm not sitting next to them, hence the "blank" profile. For cabling, I have Grover interconnects and "stock" (but heavy gauge) speaker cable.

I have plenty of CDs that indeed sound incredible, even on my modest system, but I've been subjected to countless A/B "blind" shootouts via audiophile friends, and the vinyl always wins, even in cases of LP vs. needle drop of the same LP. This last statement motivated my question (I'm not trying to be a smartass). If I digitize an analog signal at a sampling rate which captures all frequency content within an audible range, and then play the result through a good sound system, it doesn't "hold up" relative to the original analog signal. Why?

 

I think you're confusing different with better. I'm sure you can pick out the vinyl over the CDs...the noise alone would make that a snap (crackle, pop)

 

Not when comparing to a needle drop.

 

Ivan, if you haven't read it already, Steve started a thread here:

http://www.stevehoffman.tv/forums/showthread.php?t=76174

that goes on ad nauseam about this topic. Like religion, if anyone claims to KNOW exactly which is better and why, he's probably missed either the mark or the point. I understand the thirst for an answer, but it's probably better to relax and listen to the music. I love vinyl but I'm not going to beat my head against a CD agnostic's wall. It starts to hurt.

 

Well said! And thank you for the link. I'd stumbled across this once before, and got through only a few pages... I'm reading further now.

 

I don't think there is ever going to be a satisfactory answer in this debate, Each format has its advantages and disadvantages, its own characteristic distortions and trade-offs. And each listener has preferences. Some can tolerate turntable rumble and stylus-friction background noise, others not. Some don't mind tape hiss, others do. In my own case, I like the lower noise and cleaner tighter bass of CDs but I like the smoother less-fatiguing sound (to my ears) of LPs overall after prolonged listening. I have examples where a particular CD wins out over the same recording on vinyl and vice-versa. I think what I have noticed is a greater tendency for CDs to exhibit a kind of distortion, glare, or harsh brightness that is more irritating than the more euphonious distortions present in vinyl playback. People are generally right when they say what matters more is the mastering rather than the format. However I remain suspicious of the redbook standard anyway and won't join in with those who say the media or format absolutely doesn't matter. If for no other reason than the preponderance of actual CD mastering jobs in practice seems to have an inferior track record to analog mastering to vinyl. But even the redbook standard may be inherently flawed - I and many others suspect the resolution is too low for true hi-fi, but this is difficult to prove, and it is the kind of assertion that can provoke heated argument.

So I would say the resolution limitation together with sloppier mastering techniques typical in the world of CDs are the twin culprits.

 

Because with digital you are taking pieces (sampling) of the waveforms of these audible frequencies, not recording the whole thing. You have all the frequencies but not all of the waveform. Analog records the whole waveform. Understand that's oversimplified and I'm using the word "waveform" for our convenience, but it's the gist of the point. Waveform... or think the "wave" of the vibration of a given sound in the air. The argument comes along when we ask, how many pieces of a waveform do we need to keep and reproduce in order the fool our hearing? Some claim the CD has enough pieces and others still hear the loss. SACD and DVD-A keep many times more "pieces" of the waveform, and you might say that's why they can sound more "natural" than a CD to some listeners. You'd also have to consider how many of those pieces are accurate, because digital makes "mistakes" just as analog does (in very different ways!). Frequency response isn't the issue with digital of any quality.

If that's not terribly clear, I suppose we might liken this to our eyesight: Digital is like film or TV, showing you snapshots of the timestream. How many snapshots do you need to pass before your eyes at what speed in order to fool your eyes and achieve that persistence of vision? They once thought 14 or so was good enough in early eras, but by the time sound was adopted, there was agreement that 24 snapshots every second was fast enough that you would see natural motion and wouldn't sense any flicker.

These are oversimplified, again, but it might answer you as far as you need to know unless you want to learn the technical side in depth.

 

Excellent analogy- many people I know say just becuase they know the figure that the human brain can't tell the difference after 24fps (movies) which is utter horse excrament. The human brain CAN notice in the high hundreds of fps often naturally and if not after one knows to look for it.

 

This is the same wrong chain of inference that people who don't get the Nyquist equation and its implementation in digital audio always give.

No, it doesn't, unless you have devised a system of infinite sampling rate and resolution.
Which is impossible.

You answer is so 'oversimplified' as to be well in to the realm of 'erroneous'. Digital audio is one area where 'common sense' is commonly incorrect.

I would suggest you purchase and read a book called 'Digital Audio for the Recording Engineer' by Mr. Nika Aldrich. Within he lays out in step-by-step detail how digital audio *really* works.

 

I stand corrected... in my next quick post reply on the subject I'll author an entire book so I may avoid any factual oversimplification, obfuscation or error? I don't doubt Mr. Aldrich's work is accurate whereas my post is loose to put it mildly, but if you would be of help to Ivan's questions as I attempted to be, could you post a correct and specific reply in "layman's terms" within a few short paragraphs?

 

With vinyl you have an actual physical sound being made with a needle scraping across a surface (fingernails on a chalk board?) which is then amplified. Pretty simple process really, but quite unique to the "vinyl system".

With digital a sound is being recreated mathematically but not physically. This is just a theory of mine on why vinyl has such nice presence.

Also to my ear, 16/44 resolution is clearly lacking and empty sounding. 24/96 is a great improvement and sounds a lot more like analog.

Tone

 

No need for that...just reading one will do. Or if not a whole book, one of the many
articles and websites out there that explain the ins and outs.

Having done so repeatedly in the past, with little apparent traction, I now simply recommend books and a personal voyage of discovery. Aldrich discusses the 'missing information' myth in his book. And there are dozens, if not hundreds, of accurate websites discussing the Shannon/Nyquist theorem and digital audio that tread the same ground (try Google and see). Or you can ask the same questions on a site where others are perhaps more inclined to pedagogy than me -- such as the prosoundweb.com forums (http://recforums.prosoundweb.com/), or http://www.hydrogenaudio.org.
I suspect a good search of stevehoffman.tv would also turn up the paragraphs you want, for that matter. It's not like the myth hasn't been dissected before. It's depressingly deathless.

 

Not really. The wax cylinder system preceded it. The tape system also involved physical contact of surfaces, though the 'waves' are represented as arrangements of magnetic particles rather than grooves in plastic.

The sound is certainly recreated 'physically' when the D/A converter reconstructs the waveform, and it is then passed as electrical values to the loudspeakers for transducution to motion of air -- just as with vinyl.

Yes, well, many have claimed that; few have done the comparison rigorously. 24-bit (or 32-bit floating point) has definite technical advantages during recording and digital processing, but its benefits as part of a delivery format to be played at home are dubious. Higher sampling rates even more so.

 

"Better" is always subjective. I also hate surface and other vinyl noise. That's why I do needle drops!

 

I have done quite a few blind tests both professionally and casually. It's not hard to hear the resolution problems of 16/44.

If you yourself have done blind tests and can hear no difference then you get the benefit saving a lot of money on high end equipment.

 

Lucky me!

And lucky you if you can set up rigorous blind tests of 16/44 vs other formats, where you aren't actually comparing different mastering EQ, or different filters.

 

It is to you. But the person posting this clearly doesn't have your knowledge nor was he along with you through the years to read all your conversations on the matter; he didn't nab detailed books on the subjects, he asked people here, all of whom are free to be more inclined to pedagogy than yourself, or to be as abstract as they please; he didn't yet ask about for discussions examining specific issues like a "missing information myth" or the Nyquist Equation and popular misinterpretations of it; and as one of my qualifiers expressed, my reply to Ivan wasn't intended to substitute for more in-depth study such as a technical book or technical article or technical dissertation. My brief general analogies in reply to Ivan were never intended to satisfy whatever your requirements are. I feel rather belabored for your frustrations.

 

OK, without going into tons of math I will try to answer your question. Some people (not all people) are very sensitive to the effects of the anti-aliasing filter that has to be in all digital systems if you do not want to have unwanted aliasing in your recording. Aliasing is the presences of false frequencies in the recording. If the media is CD then at some point in the record playback chain there has to be at least one application of an AA filter that cuts of all frequencies above 22.05 KHz. Depending on the nature of the filter (again a lot of math to explain in detail) you get phase anomalies, leading edge arrival anomalies and ringing. All of this amounts to distortion that some people are very sensitive to (that is not say that analog systems do not have distortion, they do, just a different type of distortion). As you push up the digitizing frequency the effect of these digital distortions goes down. There are other digital distortions do to digitizing such as jitter and quantization distortion, however, IMHO the one above is the major reason why some people do not like CD's.

 

I do appreciate all of the contributions to this thread, but yes, the more technical and scientific explanations are what I'm seeking. While the principles of digital audio theory are not my forte, I am an electrical engineer with a pretty solid understanding of linear system theory, frequency domain analysis, etc., and I did mention Nyquist in my original post.

Thanks,
Ivan