Complexity of sound

June 18, 2022
 by Paul McGowan

Thinking a bit recently about needle drops and the sound of vinyl (as I mentioned in my earlier post).

It occurs to me that if one can fully capture something without loss then logically the capture method is better than what’s being captured.

It’s only been recently that digital capture has gotten good enough to grab what’s on analog without much change, yet for some time now we’ve been able to capture perfectly the sound of vinyl (Fremer’s been doing that for years).

Which says to me that pure analog as captured by a microphone is far more complex than a vinyl reproduction of it.

As I write those words it seems rather obvious to me that of course that’s the case. That vinyl, for all its wonderfulness and loyal followers who prefer it to digital, could never capture and reproduce all that comes from a microphone.

Not to diminish the magic of vinyl because that’s obvious to anyone with a great setup.

No, this rant is just an observational rambling about what’s possible in the world of perfecting audio capture.

We’re so close.

Subscribe to Paul's Posts

44 comments on “Complexity of sound”

  1. There are two basic problems to be solved: on the recording side the problem that a single microphone cannot resolve the directions of the arriving sound-waves as our ear-brain system does with the help of the pinnae and technically spoken the HRTF. And on the stereo (!) reproduction side: you must cancel inter-speaker crosstalk when you want avoid inherent comb-filter effects. This article in Copper shows what’s actually possible when music is to be stored in grooves:

  2. I think that’s logically right, but in my experience there’s still one or two things which question something of it without clarifying the whole thing further.

    First, a needle drop of a very high end vinyl rig doesn’t capture really the whole thing, just close enough to be impressing. It can’t, as its outcome is depending on the different performances of different ADC DAC combinations it’s recoded/played back with and the additional losses within a digital rig. And not many such combinations play on a level of a very high end vinyl rig in certain characteristics that make vinyl playback special (but may be better in others).

    Second, also digital doesn’t capture yet all a mic provides or all an analog recording provides. One can easily check this when e.g. comparing an SACD reissue from Analogue Productions, made by the same mastering engineer from the same tape as the vinyl releases. The differences are extreme (pro vinyl or tape) when comparing the best sounding releases, providing the relevant potential. But it needs very high end vinyl rigs to make this very obvious.

    I’m with you, that digital must get better at the end and should be better already, given theoretical preconditions and logic. For everyone who really compares himself on suitable gear, results are partly different and there still remain questions, especially when comparing the illusion playing back the media on dialed in home setups. Feeding a mic into a digital recorder and analog tape and played back without much care on openness/air or 3D illusion, will on the other side probably focus advantages more towards the digital side.

    Too many are theorizing, too few are listening on suitable gear with suitable recordings to compare imo. And no, a 20k only turntable sitting on a normal shelf, more or less properly set up, unfortunately isn’t fully suitable to show all and the extreme differences, as digital is very very good already.

    There are also forums with folks doing such comparisons on insane digital and analog gear.

    1. There is a basic principle also most relevant for audio: GIGO! 🙂 I am pretty sure that a single driver crossoverless widebander loudspeaker can perfectly reproduce a human voice captured by a microphone covering the frequency range of a human voice – no mixing or mastering involved! Things get complex and loaded with all kind of distortions and phase problems when trying to capture more than a single sound source with a huge (extended) frequency range from 16 Hz to 25 kHz (or an even higher frequency characterizing the most steep transient) and simultaneously the room sound (concert hall ambience/ characteristic reflections) using many microphones and finally reproducing a strange mix of all these mic-inputs via a multi-driver multi-way loudspeaker with a most complex and power demanding crossover additionally inducing all kind of phase problems. And the turntable most prone to any kind of unwanted vibrations and having itself a characteristic resonance frequency seems to be the most complex system having to mechanically scan grooves more or less “smooth” and more or less deformed by thermal effects during pressing and storing.

      1. Record players are crazy systems doing impossible things under impossible conditions, they just can’t sound competitive anymore, but strangely enough they do.

        The for many in the scene most competent turntable manufacturer told me, the platter and its isolation is 80% of the sound of a turntable. From my experience I’d second that. According to him, it really starts to get less enough resonant with a platter weight of over 32lbs, complete isolation from the spindle and its contact to the chassis, complete non resonant platter material and hard, vinyl-like or vinyl contact to the record, all this on a chassis cancelling down to at least 4-6 Hz.

        That’s why I say one can’t compare comparisons made on differently tweaked out setups. You’ll hear very different things.

        And still, even on the best tables, the other quirks of the analog and vinyl cutting and manufacturing process remain. That this is still competitive might give a hint on other quirks still present within digital processes.

        1. I wonder how the turntable manufacturer mentioned above would have argued the most innovative approach of the laser turntable (launched by Finial, USA, and finally perfected by ELP Corp., Japan, reeding the grooves optically so there is no friction from a stylus causing wow & flutter and all kind of resonances – and inherent distortions and poor channel separation. 🙂

          1. That would be interesting indeed, I never did. I know his experience and opinion of different kind, but never asked this, will do next time. He for sure will have heard one and also the optical cartridges, I never did.

            I found a review here which is quite mixed, with a quote from famous mastering engineer Doug Sax at the end, which somehow contributes to todays topic 😉


              1. Many thanks. A most „funny“ review not even mentioning the phono preamp used and not mentioning the parameter settings for the MM-phono section. I regularly enjoy my LT having both output options unbalanced for connecting my PS Audio GCPH or balanced for direct input in my preamp. And my vinyls are cleaned using ultrasonic tools. 🙂 If you read this „review“ between the lines you get the basic design goals and relevant findings – however what is the meaning of „dull“?. And how does a stylus remove dullness? 🙂 Having actually four classic turntables (ReVox B790, Technic SL-7, Luxman PD 444 and TW Acustic Raven and in sum total six tonearms of most different designs and a dozen cartridges including DS Audio’s DS 002 optical cartridge I can run all kind of comparisons. An ADC with phono input even allows to use my DSP for loudspeaker and room „corrections“. 🙂 However I haven’t yet met an audiophile buddy or an audio dealer who has personal experience with listening to a LT.

  3. So if vinyl doesn’t catch everything a microphone does, but digital does, then how can you be so sure that a microphone with its transducers are capturing everything exactly as is being presented to the microphone? My head is spinning with the logic that is being presented. Anything that is being sampled by definition is not capturing the whole signal – some part of that signal is inferred or averaged. Now whether that inference or averaging has any sonic value is another story.

    If what you are saying is that the transducer in a microphone is far more sensitive and complex than what the transducer for vinyl can reproduce or react to then that helps slow the head spin.

    1. When it comes to measurements discussions, Paul is a strong advocat of checking the listening result mainly for judging the effect vs the theory.

      This also helps here, but surely not everyone has the environment for such comparisons and so has to rely on theories of others, that’s what feeds the theories. If you google for comparisons of digital, tape, vinyl e.g. on the whatsbestforums, you’ll find several who seriously did this on gear and in rooms, most of us just dream of.

      1. I don’t disagree with anything you or Paul are saying. I was looking at things from strictly a pragmatic thought process. Microphones sound different, electronics sound different, cables sound different, speakers sound different, environments sound different. Then add in the manipulation of the capture, usually done by a human….

        The what’s and whys some have to know. They speculate, they obsess, they never stop. People become entrenched in their points of view and never waiver.
        People also pick their preference and look for ways to justify those preferences. Human nature says get as many to agree with you as possible. Others just listen for the enjoyment even while finding the technical side somewhat fascinating.

        Or to put it more simply, if something sounds like crap to you even if it’s the perfect recording on the perfect format played thru the perfect playback system you’re not going to listen.

        1. I understand and agree with what you say with picking preferences and trying to justify them. I think that’s valid for those who picked a preference before listening to suitable gear of both kinds or those who don’t always listen to both and don’t have a similar big collection of both media, ideally some of which make sense to be compared.

          I think the music lovers who listen to both for good reasons and on similar level have little interest in and need of justification of each concept they listen to. They just tell what they hear and continue to listen to digital, tape and vinyl in parallel.

          I personally would say, under a certain level of gear and without the will to find out and get the many recordings/pressings where this makes extreme sense, for those who start, it makes no sense for sound quality reasons to go the vinyl path. Otherwise it does. Digital meanwhile is great, too.

    2. My head is spinning with the *lack* of logic that is being presented; especially that second sentence…”then logically the capture method is better than what’s being captured”…WTF??

      1. Yes, this argument can be read extremely often and if the implied preconditions were true, it would be quite striking.

        Precondition 1:
        a vinyl rip, on suitable tweaked out setups, indeed sounds the same as the vinyl playback

        Precondition 2:
        a rip of an analog tape sounds the same as the vinyl or tape playback

        Unfortunately both preconditions are not true (they may be in certain environments)

        Assumed they were true, I’d understand the logic. Then vinyl playback would not contain information that’s missing on the equivalent digital media, then it would just add things which contribute to sound that’s perceived better by some (which is part of it, but only a part imo), Then digital could still not produce that sound, just record it, but in an ideal world of a perfect recording, this production of vinyl artifacts wouldn’t be favored anyway. The problem here…recordings are not perfect yet, that’s why many think, vinyl artifacts still help it sound better (as Doug Sax RIP mentioned)

      2. 😀

        After a while of thought the only thing I can come up with is to head to the Golf Course and chase a stupid little white ball around. After a hole or two start with the cold brew libation.

  4. Until a few months ago I’d never heard the term ‘needle drop’ which I learned here. Perhaps it’s more commonly used in the U.S. than the U.K. Wikipedia tells me that it is also known as a ‘vinyl rip’ which to my mind seems more appropriate.
    I’m with Mike and struggling a bit with the logic here. Call it what you will but it’s a copy, and to my mind a copy can never be as good as the original. I’m basing this logic on photocopies and tape copies. Each subsequent copy of a copy will lose something and whilst modern technology may have made that loss infinitesimal it’s still unavoidable. If something is added that’s not original either.
    I’m also finding it confusing when talking about capturing perfectly the sound of vinyl. Only the other day, in fact quite frequently, we talk about all the imperfections involved in this audio game, now we’re expected to believe part of it is perfect? There were a couple of hard to find CDs that became available but once I discovered they were produced from the vinyl I lost interest.
    Because I found this post confusing I may well have got my logic wrong in which case I hope someone will put me straight.

    1. You’re right, it all makes no sense when you read it. Why do some think vinyl with its quirks should sound better than digital, why someone should have the goal to record this on digital, why analog sourced SACD’s are the best quality ones for some…imo you have to dig deeper (and listen on suitably environment) or simply ignore it, it doesn’t help, it just confuses.

      That’s why I try to give a few examples today how one could listen for it.

      Another one here: This official needle drop was probably made under ideal conditions with a quite good vinyl rig.

      Unfortunately both, the LP as well as the rip are not available anymore cheaply. I have both and listened to both. The LP beats the rip easily, it’s rather night and day than anything else, and the used turntable to rip of, is good enough that this shouldn’t have been the result.

      You have to know what to check and listen to in a proper environment and make your conclusions. Theoretically based opinions are too diverse and strong and even practical reports often based on too different gear to help. You have to do it yourself under the most meaningful conditions you can ensure. If you can’t, don’t rely on theories you read, better rely on experiences you read, made by people who follow no dogma as they spent as much in the 5 to 6 digit range on their tape machine, turntable and their digital combo. They just tell what they hear and that nearly always still goes in one direction. With exception of the most famous mastering engineers I’d say, it goes in a different direction when you ask the usual professional scene.

    2. Hi Rich,
      I have two CDs that were burned on TDK blanks & I also have the original ‘factory’ CDs…everything in 44.1/16 as far as I know but maybe the copies were up-sampled, I’m not absolutely 100% certain.
      Anyway guess what…the copies sound better, cleaner, clearer than the originals…go figure!

      1. Ha, my first reaction was no way, but I do believe you.
        Of course it’s all down to how hard the error correction is working.
        No, I just threw that in but it sounds plausible, like all good stories.
        I know factory CDs are pressed and burned ones are made differently. Actually that’s always impressed me how its done, being able to burn and then rewrite over a cd. And as we both know just an extra dB or two can make it sound better, but I can’t figure, and without all the facts I doubt we’ll be able to. The thing is we’re talking digital and all sorts of manipulations can take place. Just look at some of those photoshopped digital pictures, I never knew they looked like that 😉

  5. The problem is that we are trying to capture an experience and not simply sound. When we remove the other senses, it becomes much harder to focus on the music. The addition of a familiar sound like turntable rumble can even help us ignore unfamiliar sonic distractions. None of this is simple.

    1. That’s true but the live experience is very different from home audio one and the very best we can hope for is a close approximation.
      I was listening to Gerry Rafferty’s cd ‘On a Wing and a Prayer’ the other day, a gorgeous album with a full sound. I’ve got a nice stereo image with the female backing vocalists over to the right. It reminded me of when I saw him live. I was sat in the auditorium far over to the right, so far over I couldn’t see all the musicians on stage and there were lots. Gerry had the most crumpled suit on I’d ever seen, it wasn’t even linen. So I was in far from an ideal central position but it didn’t spoil my enjoyment.
      Now I don’t know if the following is true because I don’t have any first hand knowledge so it’s based on my speculation and assumption. At a gig you have a big stack of speakers either side of the stage. I imagine all the concert sound, the same sound, comes from each side, they’re not trying to create a stereo image because you see it. The backing vocalists are over to the right but their sound is all around you, no imaginary pinpoint image because you can see them. I suppose therefore at a concert we have a mono sound that looks like stereo because it is real. On that basis recreating the actual concert sound at home you’d just end up with a mono system. As you inferred, without the visual references home audio is already facing an uphill struggle, which when you think about it, it manages to overcome remarkably well.

  6. I don’t know the answer to this, but I wonder: How close did direct-to-disc recordings come to the mic feed? Pretty close, based on how great those albums sound. The recording engineers back in the day usually recorded the sessions on a backup tape at the same time. I’d be interested in their thoughts. Both direct-to-disc and a master tape sounded more accurate than a traditionally mastered LP made from tape. Generation loss is perhaps the most important factor, and with digital there isn’t any.

    Those engineers might even have an opinion on how close a digital recording comes to the mic feed. A recording engineer I know thinks 24/192 AIFF, recorded through his ADC/DAC using his preferred digital-audio mixing software, sounds exactly like the Earthworks QTC-1 mic feed. Nothing else does that trick as perfectly, including 24/192 with other ADCs. And, again, there is no generation loss all the way into our home listening rooms.

    1. Although there seems to be loss in digital copying (several engineers stated, that the production media sounds clearly inferior to the primarily recorded file), copy loss doesn’t seem to be the main problem. The main digital losses seem to happen before and after the copy processes.

      But I’d guess a mic feed into DSD should be best in most aspects, then maybe PCM, then tape, then vinyl, probably even under extreme audiophile listening conditions. Why the replay of the production media of each at home then doesn’t mirror that sequence, is the quest.

      I talked with a few (personally very audiophile playback orientated) label owners who do DSD and analog recordings about this and they don’t know either, but they finally decided to record in analog for vinyl and in DSD for digital and they prefer the analog vinyl (although they might prefer the DSD feed in the studio).

      1. I’m confused about your comment that there “seems to be a [generation] loss in digital copying.” That’s demonstrably not true and it’s easily debunked. I believe that Bob Katz many years ago did a test where his 100th-generation copy was bit-for-bit a clone of the original. Maybe I’m confused about what you mean.

        Another interesting anecdote: A mastering engineer acquaintance told me 15 years ago that when he gets back a preliminary run of CDs from the manufacturing plant for his approval, he does a rip with iTunes to AIFF. The QC process is that he loads the ripped AIFFs into his DAW and lines up the rips with his original tracks that were used for CD replication. He would line up the waveforms and invert the phase on one to produce a null. If the null wasn’t perfect for the whole album, he would tell the replicator to start over. No CDs from this audiophile label were ever released commercially that didn’t pass the null test.

        1. Yes I understand this and I shouldn’t try to explain, as I can’t. I have heard it myself, I have heard and read about it, but hardly anyone can explain what gets lost and how in digital processing.

          If you haven’t heard Paul’s interview with mastering legend Bernie Grundman, do so.

          He also touched this topic and said something like „I don’t know how and why, but something seems to get lost in digital on the way from the recording to our media.“ He produces analog and digital and should be a reliable and independent source of information. All mastering engineers I talked to or read about, really still doing both analog and digital on a high level (which quite some of the most famous do), stated more or less the same.

          Not to keep back, that much also gets lost on the way of the vinyl or tape production process…different things seem to get lost than in digital.

          1. The losses and problems in digital transfers are down to the power supplies.
            Okay, so that’s tongue in cheek, but there could be some truth in it, worth considering?

          2. Yes, jazznut, that was a great interview with Bernie Grundman. I respect both Paul and Bernie immensely, but Bernie is wrong when it comes to his hypothesis that the processing that goes into making a CD — submaster, glass master, duplication, etc. — causes a generation loss. That hypothesis is easy to test objectively and it has been falsified. As I mentioned, rips from commercial CDs will null to the original 16/44.1 files that the mastering engineer used to create the submaster sent out for replication.

            If Bernie’s ears tell him that his CDs don’t quite sound right when compared to the original digital files played through his mastering system, then maybe the fault is his CD playback system, not generation loss. Occam’s Razor.

            Personally, I gave up on CDs over 10 years ago because on my system and on a few friends’ systems I could demonstrate that CDs ripped to AIFF files sounded better than the CDs they were ripped from. (At the time I was using Pure Music’s memory-play feature on a Mac Mini.) There’s a lot to go wrong in CD playback, and other things to go wrong in computer-based playback. Audible differences will depend on the CD player itself and on the computer-based playback system.

            The point, however, is that the differences in digital playback alluded to by Bernie are not caused by generation loss. It’s an interesting and plausible hypothesis, but it has been falsified.

            1. Actually it turned out Bernie had spotted an error caused by some of the transfer equipment that the manufacturer had missed. It was very subtle but bits weren’t bits.

            2. What I personally heard was that a file copied on another harddisk without change in power supply sounded clearly different. But that’s probably another topic again, depending on harddisk speed and noise. A can of worms. And that can of worms probably is, what also influences digital media production processes.

              I think no one fully understands those various duplication influences yet, so I just recognize his experiences of the differences and skip his attempt to explain for the moment 😉

  7. Dear Paul,

    “[V]inyl . . . could never capture and reproduce all that comes from a microphone.”

    I’m a little bit confused about your reference to vinyl as a “capture” medium.

    I think of tape as an analog capture medium, but not vinyl. I think of vinyl as a copy or mass reproduction medium.

    Unless you’re referring to direct-to-disc recording, in which case vinyl is a capture medium.

    With warmest regards,


  8. For me the final quest is:

    I assume (others might know this for sure or not), a mic feed into DSD must sound better than into tape or D2D in the studio, but how does this match my experience how much better an LP of an analog recording sounds than the SACD produced from the (I think analog edited) DSD capture of the same tape (the latter quite completely missing the airy ambiance and 3D spaciousness, dynamic playing etc.).

    To verify this, I also compared many very good digital recordings pressed on vinyl and SACD with no generally different mixing/mastering, and often there’s hardly difference at all, just in minor characteristics, at least not at all in that range with those digitally sourced recordings.

    When certain digital recordings are on the other hand mixed differently (analog) for vinyl, than their digital media, then again the vinyl is far superior (I have examples).

  9. Sorry I am late to this, had yard work this morning. This is a very difficult question to answer. To do so I want to break it into two parts: analog only and then digital.

    Let’s say you listen to a singer live 10 feet from you. How can you then tell if a capture chain ( microphone, cable and microphone preamp ) accurately capture that music. It is very difficult to do this. I could take the output of the preamp and feed it into an oscilloscope and a spectrum analyzer and study the signal, however, I cannot feed the person’s voice directly into these instruments. The only thing we could do is to have the singer go quickly in a sound proof box, sing again into into the microphone and have the preamp hooked up to an amp and speaker and you listen to that. Obviously, this is not a good experiment because we have added an amp, speaker and cables to the chain ( but no recording ) and we are depending on your acoustic memory. Thus, IMO, it is very difficult to determine if a microphone captures music 100% accurately.

    Now let’s talk about digital.

    We know that when you digitize that you have to use an anti-aliasing filter to have it sound right. In CD’s this filter removes the frequency content above ~ 20 KHz. We know that the frequencies above 20 KHz are essential to having impulse signals ( i.e. a drum strike ) with sharp leading edges and no pre-ringing in the signal. We have learned that humans are extremely sensitive to leading edge arrival times and thus the sharpness of that edge is very important. The sampling period used for CD’s is one sample every 22.68 micro seconds. If we reduce that period by a factor of 64 and take one sample every 350 nano seconds ( this is what is done in DSD used for SACD’s ) we find that impulse signals look much better. Thus what I think we can safely say is that digital recording is coming very close to high quality analog recording ( i.e. an RTR deck at 15 i.p.s. ).

    Finally, when it comes to vinyl I believe that the magic is there.

    1. I have a digital pipe organ with multi-channel output. It sounds very realistic. When I supplement the sound with a pair of DPA 4006 omnidirectional mics positioned in the middle of the room feeding a Gordon balanced discrete two-channel mic amp that feeds my Pass Lab monoblocks which feed a pair of Von Schweikert full-range loudspeakers, the sound of the organ becomes warmer and richer. The DPA mics are very accurate. The Gordon amp and the Pass Labs are very accurate, but render a somewhat tube-flavored tone. The overall sound is richer and more pleasurable. By using the microphones and additional amplification circuits, the organ sounds better than just listening to the digital organ playback system alone. In this case, system complexity adds an extra dimension that is pleasurable, although not as accurate. Sometimes more is more.

    2. The funny thing is, those stupid people without a clue, favoring this silly technology which is just pleasing due to less information, who are famous for creating the best sound for decades and don’t know how the real thing sounds, master the SACD’s you’re proud to listen to 😉

      But they at least listen and compare and don’t just repeat cliches of the 80’s 😉

  10. My system is digital only, but I have to say that I’ve always thought that most vinyl rigs I’ve heard had better cymbal definition and decay than digital.

    1. You have ears, that’s indeed an immediately noticeable characteristic. It’s one that anyway made big progress in digital. The times of artificial and bloodless sounding cymbals in digital are over. It’s close meanwhile and good enough to be absolutely satisfied without comparison.

      I remember some years ago here in the forum when some said, what vinyl provides more there in terms of decay, air or similar is artificially added, belongs to (not applied) compression etc.…just until a noise optimized update of the DS firmware improved the DAC in a similar way…then it suddenly was real and the true thing again.and everyone praised the progress.

      The best digital and vinyl gets closer than the die hard camp defenders want to realize.The thing is, the die hard dogmatists of both camps rarely have experience with high quality gear of the opposed camp, that’s why they discuss as they do. They decided for their camp on theoretical basis mainly or by comparing with too low level gear of the opposed camp.

      I don’t say everyone will favor vinyl. But those who prefer digital and made the comparisons have other arguments and talk about other characteristics than those who obviously repeat old theory.

  11. There are a few people who have top notch digital rigs and vinyl rigs, and most of them prefer vinyl. One guy on the what’s best forum has more money in his digital rig, but still prefers the sound of vinyl.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram