Why does vinyl (analog) sound better?

Discussion in 'Audio Hardware' started by ivan_wemple, Jun 22, 2006.

Thread Status:
Not open for further replies.
  1. coopmv

    coopmv Newton 1/30/2001 - 8/31/2011

    Location:
    CT, USA
    A pre-recorded open-reel tape that was properly transferred from the original master is still the best analog sound, period.
     
  2. Chili

    Chili New Member

    Analog is a continuous form of recording, by definition.

    Digital approximates the waves with impulses, creating a series of step functions (the number depends on sampling rate). Now your smooth little wave is all squared away (time domain). Each impulse if then converted into the frequency domain, through a DAC. So...what you are hearing is in fact continuous and smooth...but frequencies are being constantly left out because of the impulse approximation that digital is fundamentally based on. This does have a large impact on phase and timing, which our ears can detect to 9 decimal places.

    If you ask me, digital is kindof like reaching around your back to unzip your pants. Analog begins and ends continuous, and stays in the frequency domain.



    Bad analogy, but if you have to go there, we have no way of storing samples of information at the rate of brain signals anyhow. Analog is so much more convenient :edthumbs:
     
  3. toptentwist

    toptentwist Forum Resident

    Location:
    Houston, TX

    I didn't rule out mastering entirely... but I still don't understand the argument
    that the container is irrelevant... unless the original signal was "small" enough
    to fit inside the 44.1/16 "envelope"...


    Recognizing that the curious here wanted me to repeat my experiment with
    several different masterings, I dug out the following


    1.) My original "Rumours" LP (purchased in August 1977)
    2.) My Nautilus "Superdisc" LP (it has a 1980 date - but I purchased it in a used
    bin cheaply sometime in the last 10 years)
    3.) My old CD (purchased sometime in the last 10 or so years cheap in a used
    bin - date unknown - my guess is early 90s based on the fact
    that there is no SPARS code)
    4.) Recent CD remaster (2 CD set that came out about two years ago)
    5.) DVD-A


    I'm really not sure WHAT the Nautilus "superdisc" is... it mentions something
    about it being "digitally remastered" on the rear sleeve. I vaguely recall reading
    something about it circa 1980 that claimed the band worked closely with the
    company that made it (which may have been just marketing hyperbole).

    I wasn't able to locate a stop watch - but I wanted to get some sort of
    approximate data out here.

    I repeated the experiment several times for each recording - trying to
    make sure that I was comfortable with my measurement (I'll freely
    admit that my measurment is half-baked without a stopwatch - but
    given that I still had to start and stop counting - I'm not sure there
    is a really good scientific way to do this).

    For grins, I repeated the experiment using two different rigs:

    Rig 1: Pioneer 563a -> Dynaco Tube Amp -> Tannoy Speakers
    Marantz 6300 TT with Grado cartridge -> Dynaco Tube Amp -> Tannoy Speakers

    Rig 2: JVC 723 DVD-A Player -> Harmon Kardon Receiver -> Cerwin Vega Speakers
    (no vinyl playback capability for Rig 2)

    At first I was going to limit myself to just the stereo version on the DVD-A
    but then at the last minute, I decided to repeat the experiment with the
    multichannel mix. Not a fair fight - but yet another data point.

    These were my numbers:

    DVD-A on Rig 1 : Three seconds
    old CD on Rig 1 : Two seconds
    new CD on Rig 1 : Three seconds
    original LP on Rig 1 : 2.75 seconds
    Nautilus LP on Rig 1 : 3 seconds

    Forgetting the precision of my measurements, clearly the "old CD" was
    the worst performer. The whole thing sounded like I threw a blanket
    onto the speakers.

    To my surprise, the DVD-A and the "new CD" seemed to be equal in
    length. But if I had to choose a favorite - I think I would choose
    the DVD-A because it somehow sounded cleaner to my ears. The
    'dit dit dit dit' on the high hat seemed forced to me on the "new CD".
    Almost like what I would get if I jacked up the top slider on an EQ.

    What was mostly definetly clear was that the "new CD" is much much
    better than my older CD.

    Were the vinyl numbers up there with the DVD-A and the "new CD" ?
    No. But they were much better than the old CD.

    I'm not even sure if "Rumours" is a really good example for
    this type of shoot-out but it happened to be something
    I had a few different copies of - and I started showing
    people the comparison about 5 years back (when the
    remastered CD didnt exist)

    I repeated the same experiment using Rig 2 and I felt like the
    numbers were identical to Rig 1 - with one exception - when
    I tried using the multichannel mix on the DVD-A, I felt like there
    as extra half second for the splash.


    Now why do I think theory predicts this ?

    It has something to with simulation tools that I had access
    to in college. You could do things like change the sampling
    frequency and watch the damage it does to a signal.
    By watch, I literally mean watch - since the tool would
    plot the results every time you changed a parameter.


    And why are square waves used in labs??? That's easy,
    because you can spot a problem very quickly. You could
    get the same result if you sent a bunch of different sine
    waves into the system and plotted the results for each.
    But its much easier to feed a square wave and then
    magnify the corners of the square and look at the
    damage (if any).

    And this may be semantics, but isn't an impulse
    a special type of square wave ? Its a square where
    the top of the square is very small. And don't
    drum beats look like an impulse when viewed on
    a scope ???
     
  4. dekkersj

    dekkersj Member

    wags,

    In theory, the 16/44 format is good enough for our ears. 16 bit provides a 91 dB dynamic range (if properly dithered) and the bandwidth is even somewhat larger than what humans can hear (so the risetime of the timedomain signals is limited but this is not audible). But there is a catch. One needs a brickwall filter (in fact 2, one at the AD and one at the DA side) and they don't exist. If it exists, it is useless since you have to wait forever before a signal comes through. So you have to deviate from theory and make filters that are look alikes. When a FIR filter is used one creates pre and post echos, and they can be hearable. A ripple in the frequency domain of 0.2 dB wil produce a pre and post echo in the order of -40 dB relative to the original signal. Today, it is believed that this is the main problem in digital audio.

    What you see now on the market is chaos. Every manufacturer tries his solution and make sometimes bad designs. But they may sound better for some people.

    To overcome the problem, one can go higher in samplefrequency. The filters are easier to make and so on.

    Regards,
    Jacco
     
  5. william shears

    william shears Senior Member

    Location:
    new zealand
    Square waves, round waves, bent waves, brain waves yadda yadda yadda!

    Vinyl sounds better 'cos it looks better. it smells better, it feels better, it spins better. It probably tastes better but I havn't got 'round to that yet.

    Hey science is groovy, don't get me wrong. We made it to the moon..whoo-hoo!
    But after we've used our stop-watches to time bongo decay and inverted our phases so they stick out of our flanges at 43.789 nautical degrees PRECISELY maybe we should kick off the techno boots and give a thought to what the singer or the guitar player or even the bongo player is laying down.

    I would bet that the most emotional musical moments for 99% of the members here have come from shards of candy pop blaring from a midget speaker in a $25.00 radio. Or some cool riffing that you dug and shall dig for evermore from the beaten-up turntable in your parents front room that smelt funny when it was left on too long.

    I don't know if the experiment was ever tried but my bet is that had you got ol' big-brain Einstein in the parlour and given him a glass of Thunderbird wine and slapped on a 45 of Jerry Lee, or a Louis Prima record he would have cut some rug. The science would have gone right out the window. He would have been suckered like a kid at a magic show.

    Its all in the hips.
     
  6. dgsinner

    dgsinner New Member

    Location:
    Far East
    In theory, bees shouldn't be able to fly (the ratio of wing area to mass is too small IIRC). In theory, standard 'redbook' digital should sound better (to me) than analog vinyl. In my experience, nine times out of ten, it does not. I started coming to that conclusion in 1999 when I got rid of more than half of my CD collection (probably around 2,000 discs). I don't miss 'em at all. I stopped listening to them after a couple of plays. They weren't good enough for my ears. They still aren't.

    Dale
     
  7. dekkersj

    dekkersj Member

    But it is silly to think that the format itself is responsible for the bad sound. Or you must introduce distortion in the listening room to get a sound that is good.

    But I think you missed my point. That the cd format is not ideal due to practical limitations. Nowadays we are capable of producing cd's that are perfect (for human ears that is) and the reproduction is good enough as well. But not all DAC's will have this performance.

    Where it all boils down to is that we humans want to hear distortion. Apparently. That's why analog systems, especially vinyl, sound better...If and only if the same master is used. Otherwise this could be the most important factor.

    Regards,
    Jacco
     
  8. Scott Wheeler

    Scott Wheeler Forum Resident

    Location:
    ---------------

    "But it is silly to think that the format itself is responsible for the bad sound."

    It is now. It wasn't back in the mid 80's. Much of the problem stems from the folks that declared the format perfect back then without listening to it critically. You tell the audiophiles it's perfect and it sounds like crap you loose their trust.

    "Or you must introduce distortion in the listening room to get a sound that is good."

    Distortion into the listening room? Do you mean distortion into the playback chain? You don't "have to" but to assume it will always result in degradation is every bit as silly. Ditortion is an inherent part of recording and playback. Stereo is a form of illusion not a form of reproduction. There are good reasons to believe certain distortions will enhance the illusion. Lets also not forget that recording and playback chains, because they are based on anillusionary presentation are always going to be inherently full of distortion regardless of transparency or lack there of, of certain parts of the chain.



    "Where it all boils down to is that we humans want to hear distortion. Apparently. That's why analog systems, especially vinyl, sounds better..."



    No, I think you are missing the bigger picture here. humans don't want to hear "distortion." If that were the case we wouldn't use live music as a standard. What it really all boils down to is that we have a universal system for recording and playback that uses transducers and stereo as a format to create an illusion of live music. It is not a form of *reproduction*. What it all boils down to is that it's a false assumption that in the creation of the illusion of live music through stereo playback all distortion in the chain is a bad thing. If one considers the inherent flaws in stereo playback recorded and played back over transducers, it should make sense that some of the inherent distortions in vinyl may enhance the *illusion* of live music during playback. That doesn't mean we like distortion per se, it means we like a better illusion regardless of what creates it.




    "If and only if the same master is used. Otherwise this could be the most important factor."

    The factor that is all too often ignored by both camps.
     
  9. Chili

    Chili New Member

    We want to hear certain types of distortion maybe. CD is not distortion free by any means. The most painful noise imagineable has to be digital clipping, and the absence of frequency's between the gaps of impulse approximation...also a form of distortion.

    Its similar to the reason some people prefer tube amps to solid state amps....distortion preference. Tube amps have a soft compression and higher order harmonics of distortion, when driven heavily. Solid State amps contain mostly 2nd order harmonics, and violently clip when driven hard. As the last guy said, its all an illusion anyway, and theres plenty of distortion introduced in every stage from the recording studio to your ears.
     
  10. dekkersj

    dekkersj Member

    Yes, you are right. That is not the way to go. In retrospect, there are mistakes being made. Stupid mistakes. I am not aware of all details, but that there were flaws, that's true.

    [...]

    Well, for the vast majority live music is not the standard. Unfortunately. It is a slim balance in terms of distortion that is regarded as "wanted". A bit of distortion is often referred to as "containing more detail". For what it is worth.

    Totally agreed.

    Regards,
    Jacco
     
  11. dekkersj

    dekkersj Member

    The absent frequencies are the frequencies that are above the Nyquist frequency. Below that criterium there is no information lost. So, if we can hear up to 20 kHz, and the Nyquist frequency is at 22050 Hz, there is no information lost for our hearing system. By definition.

    Regards,
    Jacco
     
  12. soundQman

    soundQman Senior Member

    Location:
    Arlington, VA, USA
    Good points, but I've always understood tube amps to have more second-order harmonic distortion, but that this kind of distortion gives a sense of liveliness to the sound, and is relatively pleasing to the ear compared to other types of distortion. At least it supposedly sounds better than higher-order distortion, which I thought was more characteristic of solid-state and digital. Maybe I have it wrong. Perhaps there is higher order harmonics with tubes, but it is lower in amplitude or pushed so high up in frequency that it is less audible. Soft compression and clipping with tubes - yes. NAD introduced solid state circuits that imitate that soft-clipping characteristic and use it as a big selling point in their products, which are solid-state.
     
  13. Steve Hoffman

    Steve Hoffman Your host Your Host

    Guys,

    You could go on forever with this (and can if you wish; I don't mind).

    Thing is: A recording is just an illusion; it's not real. The fantasy of a live person singing or playing in your listening room rather than a recording of a person is (or should be) the goal of any recording and playback system. Whatever recreates FOR YOU the "illusion of life" is a system not to be sneezed at. That's quite a goal to accomplish and whatever it takes to fool the ear is what it takes, simple as that. Analog usually fools us more, therefore, for music playback, it's better. Doesn't mean it's more accurate or more this or more that, but if it takes some tubey harmonics, etc. to fool our little ears, well, that's the way to go. One false step in the chain will shatter the illusion.
     
  14. dekkersj

    dekkersj Member

    Amen.

    My point is more about the claim that the cd format itself is the main dissonant. That it is not good enough or that it has a sharp edge or whatsoever. Or that it is dull sounding.

    Regards,
    Jacco
     
  15. Grant

    Grant Life is a rock, but the radio rolled me!

    Once again, clipping is NOT inherent in the CD spec.
     
  16. LeeS

    LeeS Music Fan

    Location:
    Atlanta
    Because CD does not capture everything Wags. It is an old format with real material limitations - a breakthrough for its day but limited. I have done far too many comparisons in real time between acoustic events in the studio and then playback on CD in 16/44, 24/96, 24/192 and DSD. DSD always wins to my ears but24/96 can be a big step up over 16/44. It's not subtle, it's not very gear dependent. It is material.

    And as much as I like DSD, an analog tape can be better still. :righton:
     
  17. Driver 8

    Driver 8 Senior Member

    After half a lifetime of listening to music, I'm here to say that it's not. I don't really profess to understand the Nyquist theorem or the theory behind digital sampling in general, but it is blatantly obvious to my ears that redbook CDs just don't have the same "bandwidth" as vinyl lps. Maybe a higher sampling rate could solve this problem - I really don't have enough personal experience with SACD to pass judgment on that.

    To me the most painful noise imaginable is when a dirty or scratched CD locks into a digital loop.
     
  18. That doesn't really matter for me, I can only hear up to about 17 KHz anyway, so even if all the new formats extend the frequency response, it is doubtful I'll hear it.

    However, I am happy to conceed that may SACDs are very carefully mastered, because the labels are aware that SACD buyers care a lot about sound quality. So, as always, the difference may simply be in the quality of the mastering...

    The much bigger problem is mastering engineers not taking advantage of the capabilities of what CD can offer, rather than the CD format itself being fatally flawed.
    Do you mean comparing a DSD transfer of an analog tape recording to the original tape? Or comparing a DSD recording and an analog tape recording of the same thing?

    My guess is that it is hard to do a fair comparison of recording the same material because don't analog and digital recordings require different mic-ing techniques?
     
  19. LeeS

    LeeS Music Fan

    Location:
    Atlanta
    Some are, some are not...but that does not take away from the real benefits of extra data from faster sampling.

    The latter although the former can sound darn good. You generally use the same mic setup on both recordings. At least that is what we do with our acoustic work.
     
  20. dekkersj

    dekkersj Member

    As a SACD fan I am glad to read this kind of things. But if you look closer than one must conclude that the bandwidth of SACD is not that much larger than cd. It is not even clear how bandwidth is defined if you take into account the large amount of noise.

    To clarify my DSD interest: I share the same coffeemachine as the guys who developed SACD.

    Regards,
    Jacco
     
  21. Gary

    Gary Nauga Gort! Staff

    Location:
    Toronto
    I personally don't care about benefits of extra data from faster sampling if the SACD (or DVD-A) is mastered poorly. By that, I mean that extra compression or EQ (etc.) could be added for a "WOW" factor.

    Additionally, deficencies in the tape used could come to light. I have heard two SACD's that are (in my mind) "defective" in this way.

    I still own one of them....
     
  22. Robin L

    Robin L Musical Omnivore

    Location:
    Fresno, California
    So the real question is: what matters in creating that illusion? How does our hearing mechanism work? What are the specific factors that tell us "This Is Real"? For me, ambient retrieval and overall timing is crucial. Mid range accuracy and resolution are so important because most of the "This Is Real" cues are in the midrange.
     
  23. dgsinner

    dgsinner New Member

    Location:
    Far East
    This is the crux of the matter and what confuses me about all the people who seem to want to 'prove' to me via scientific analysis that one thing (digital) really does sound better whether I realize it or not.

    People, you are doing the listening. If one thing or the other sounds better to you, fine. A scientific analysis goes no distance whatsoever in determining what I like, or in this case, what I think sounds better.

    Dale
     
  24. lukpac

    lukpac Senior Member

    Location:
    Milwaukee, WI
    Don't forget noise reduction.
     
  25. bugmenot

    bugmenot New Member

    Location:
    Nowhere
    it isnt supposed to. its supposed to show that perhaps your reasons for justifying this practice may not be correct, but that doesnt mean your preference is incorrect at all.

    you can think vinyl sounds better all you want and i dont think anybody directly feels like contesting that. "think" is subjective. "vinyl sounds better because it has the little squigglies that digital does not" is not subjective, and can be argued.
     
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine