Vinyl v. digital curiosity

Discussion in 'Audio Hardware' started by SKBubba, Oct 3, 2018.

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

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

    I still hear the flutter, and I doubt the table is built to work with 45s.
     
    2trackmind likes this.
  2. Pinknik

    Pinknik Senior Member

    It could do 45s. I can barely hear an issue with the first version, but things need to be fairly wide out of whack for me to notice.
     
  3. Grant

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

    I spend a great deal of time and effort to manually/visually center a 45 onto the platter before doing a needledrop of it.

    Maybe it all comes down to those who are OCD and those who aren't.
     
    Pinknik likes this.
  4. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    What was it Philips said in 1983? Oh yes - "Pure perfect sound forever". And I was won over, bought pretty much the first CD player (CD104??) and tried to force myself to like the sound of someone scraping the back of my eyeballs. And yes - it measured so much better than vinyl. Lower noise and lower distortion by orders of magnitude - so it just had to be better, yes? Well, no. Certainly not in 1983 and for at least two decades thereafter.

    Indeed Kate Bush has issued new recordings because she could not stand the awful screechy sound from the early generation A-D converters. Which of course measured properly.

    Well, OK - fast forward. Digital audio technology is massively superior now. Digital resolution and processing engine speed is in a whole different league. By orders of magnitude - it obeys Moore's Law in the same way as computers. I have now ripped my CD collection of thousands to a Synology NAS drive (and oh boy was that a job!), and I subscribe to Tidal HiFi. And it sounds truly superb. No question.

    And you know what? Most of my listening is to vinyl. Yes it is technically lousy as compared with digital. It might make a signal to noise of 70dB on a good day. And tracing distortion of 0.5% on average with any arm alignment. In fact a cartridge has an intrinsic noise. For example a Denon DL103 with 40 ohm coils has a maximum possible SNR of 68dB as a result of the noise from the (admittedly unusually high) 40 ohm coil resistance. There is wow and flutter. The hole might be punched off-centre and give periodic change in musical pitch. And warps and ripples cause intermodulation distortion sidebands on the audio.

    OK - all that says vinyl should sound totally rubbish. Scraping a tiny bit of diamond across a piece of soft moulded vinyl is a crazy idea (in the same way that exploding petrol/air mixture is a mad way of propelling a motor vehicle). But you know, none of that matters, because something in your brain ignores all that imperfection and just listens to the music.

    One clear benefit of vinyl is that you put a record on the turntable, and you listen to an entire side. You might well turn it over and listen to the other side too. With digital it is all too easy to dip into tracks. It is so easy to pick and choose. Of course that has benefits in broadening musical taste. But to get under the skin of a recording artists output, vinyl wins for me, hands down.
     
    Kiko1974, jeffnesh, VQR and 5 others like this.
  5. Grant

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

    But, just because something measures adequately does not mean it will sound transparent.
     
  6. bresna

    bresna Senior Member

    Location:
    York, Maine
    But the whole point of this thread is that what you are reporting does not have anything to do with the A/D converter because if you digitize one of your LPs, even with a so-called ancient, crappy digital converter, the resulting CD will sound just like the LP. If the A-D converters were adding screechy sound, this needle drop would be screechy too. Digital audio doesn't add any screech - it has to come from the original source. So whoever mastered Kate Bush's first CDs likely didn't know what he or she was doing... not uncommon in those early days either.

    Remember that DCC was making CDs back in the early days of A/D converters and yet we can still sing the praises of these discs because they sound so faithful to the music. I've never heard anyone describe any of them as "screechy". :)
     
  7. Carl Swanson

    Carl Swanson Senior Member

    I think the point he's making is, more than two is superfluous.
     
  8. Carl Swanson

    Carl Swanson Senior Member

    You're a victim of poor mastering, not a bad medium.
     
  9. Mal

    Mal Phorum Physicist


    When applying the sampling theorem, you're not handling an infinite set of sine waves individually - you are dealing with a signal that can mathematically be considered as comprising an infinite set of sine waves.

    This elegant mathematical equivalence opens the door to being able to rationalise a band-limited analogue signal as a set of discrete instantaneous samples (amplitude measurements) and being able to transform the signal from one form (analogue) to the other (digital).

    The 'more than two' is from the definition of the cut-off frequency at which aliasing begins - you need to sample at more than two sample intervals per period of a sine wave to be able to reconstruct it without aliasing.

    Again, I use a sine wave here because it simplifies things into an elegant model - while true mathematically, it doesn't mean that the input signal is really a sine wave. After all, there's no such thing as a perfect sine wave in nature, just infinite sums of the things...

    The sampling/reconstruction process is based upon a mathematical model that involves an infinite set of sine waves but that is not a practical limitation since the signal is treated as one continuously varying function in the analogue domain or a finite set of samples in the digital domain. Mathematically, this is an equivalence for a band-limited input signal.

    In practice, the output is not identical to the input but not because there were an infinite number of sine waves to track and we couldn't keep up. Instead, the limitations are because we can't band-limit the input without adding filtering artefacts, we can't sample at perfect intervals nor can we quantise with perfect accuracy which distorts the encoded waveform and we can't reconstruct without adding more filtering artefacts. The mathematics doesn't have any of these real-world problems to deal with and so is error free.

    These limitations in digital audio can be reduced to the point where it is effectively transparent under normal listening conditions although under enough scrutiny a difference will always be apparent (ie. studying the noise floor).
     
    Last edited: Oct 8, 2018
  10. Eno_Fan

    Eno_Fan Staring into the abyss: Brockman BIF, Pilbara WA

    Location:
    Izieu, France
    Excuse my paring this down. You're suggesting that sampling our continuous analogue signal at integer intervals of time captures it perfectly as long as sampling frequency is 2fn, yes? If so, then please explain the fate, post-sampling, of the signal between samples if this is a faithful transcription of our original fn signal that "...perfectly reconstruct(s) the original waveform". Do you contend that all the inter-sample information that existed between those samples in our original signal is not lost in the sampling process? Continuous- and sampled-time signals forever differ, not unsurprisingly, in that the former are defined temporally everywhere whereas the latter are only defined at the sample instants. That is, by unavoidably ignoring what is going on in these inter-sample areas the sampling process throws away information about the original signal. Digital sampling is intrinsically lossy (indeed, that is why it is so-called -- it is a "sample'). Neither does the reconstruction ('decoding') of this sampled signal rectify this aspect of the ADC, 'encoding' side of things -- these lost data are merely interpolated. Aliasing here (or indeed in the sampling) is not the point -- reconstruction of the sampled signal in the time-domain bears only an approximate 'join-the-dots' relation to that original signal and nothing more. It is not perfect, just as a line drawn through a series of data points is insufficient to characterise a phenomenon when the number of those points is too small. Infinite bit-depth per sample and perfect band-limiting of the input signal would fix this, but neither is possible in practice as opposed to theory.

    Are the two statements below not actually saying the same thing in different words?

    "...Sampling and reconstructing an analogue waveform (encoding/decoding) is not a game of patching the original together from incomplete parts"

    "...it's a game of mathematical wizardry involving discrete, regularly spaced, instantaneous amplitude measurements (samples) and this sample-set's relationship to the original analogue waveform".

    Not terribly different from MPEG-1 Audio Layer III coding -- just temporal sampling instead of perceptual coding. Different sorts of lossy.
     
    Last edited: Oct 8, 2018
  11. Mal

    Mal Phorum Physicist

    No because instantaneous measurements (samples) don't constitute parts of the original since they have no time variance.

    It's a magic trick the complex number Universe allows - trying to explain it by imagining a pile of sticks and a piece of string will only end in tears.
     
    VQR and missan like this.
  12. HDOM

    HDOM Well-Known Member

    Dont forget, we can develope a great digital file, whit no noise or jitter, forget the cd for a while, and thing how much we can develope from a "new" digital effort!
     
  13. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    http://www.clir.org/wp-content/uploads/sites/6/AD-Converters-Pohlmann.pdf published twelve years ago. Sonic differences between different professional A-D's of extreme specification are decided between by ABX testing. For archival digital storage of analog audio from wax cylinders, shellac 78's, vinyl and tape.
     
  14. bresna

    bresna Senior Member

    Location:
    York, Maine
    I read that whole paper, dry as heck I might add, and all it really says is that you should use a modern converter because they are better than older converters and that you should use ABX when comparing them. It does not say that audio produced from old converters will sound bad and it certainly doesn't say that it will be screechy.
     
  15. bluesky

    bluesky Senior Member

    Location:
    south florida, usa
    I have never personally experienced 'the process' of mastering LPs or CDs.

    I've read...
    LPs are mastered much differently than CDs because of the medium involved (keeping the needle in the vinyl track). For better... or for worse, there is a lot of manipulation/massaging to the originally recorded sound on the studio analog recording to get it playable on LPs. Mastering CDs is a pretty straight forward process. (???) Not as difficult as analog for LPs.

    I love the mastered 'sound' of LPs. Just sounds much better to me. OK, I can understand that it's not as 'accurate' as the original sound on a CD but it usually does 'sound much better' to me, but not always. Sometimes the CDs sound better. Analog/Digital - two different sounds.

    Something like that... I guess.

    But just put on a NEMS Black Sabbath LP and buckle your seat belts... your going for a ride!! :)
     
    Last edited: Oct 8, 2018
  16. Chris Schoen

    Chris Schoen Rock 'n Roll !!!

    Location:
    Maryland, U.S.A.
    That's why when the cd sounds like ****, I can at least enjoy the vinyl, which in many cases is better.
    - You are right though, the cd ("perfect sound!") is only as good as the mastering...
     
  17. HDOM

    HDOM Well-Known Member

    Have you guys, ask your selft;

    What live music sounds like? Is it more vinyl or digital like sound?

    Or neither!
     
    ilo2 and Fishoutofwater like this.
  18. xfilian

    xfilian Forum Resident

    Location:
    Essex, UK
    I agree. And to my mind, listening pleasure has nothing to do with absolute accuracy or endless pages of science. Listening pleasure is all to do with sounds flowing through those two things stuck on the side of your head and whether those sounds elicit positive feelings. The best vinyl elicits those feelings to me far better than CD or any digital format. You cannot quantify or explain that with numbers and science.

    At the end of the day, what does it matter? This debate will never end because we are all respond differently to certain things. A bunch of people on here respond in a similar way to me to vinyl, others respond in the same way to digital. Its all good. :)
     
    bluesky likes this.
  19. beppe

    beppe Forum Resident

    Location:
    Venice, Italy
    I trust my hears and my brain. Problem isn't vinyl or digital, problem is bad mastering and loudness.
    Too much artists and sound engineers want music played loud and flat.
    Please GO HOME!
     
    Leonthepro, HDOM and Fishoutofwater like this.
  20. fogalu

    fogalu There is only one Beethoven

    Location:
    Killarney, Ireland
    A famous classical pianist (I can't remember who it was) once remarked that he could never tolerate piano recordings (tape or vinyl) until the CD arrived. He was extremely sensitive to pitch variations and it didn't matter how good the analogue equipment was.
     
    HDOM likes this.
  21. John Dyson

    John Dyson Forum Resident

    Location:
    Fishers, Indiana
    I have existence proof that at least some of the 'problem' with digital audio is actually the lack of ANALOG DolbyA decoding Below is a post that I just made to HydrogenAudio with examples of DolbyA decoding of legacy material from digital media. I also have a repository with a few such examples (I don't think that I can offer 30second snippets here like on HydrogenAudio):

    They probably should remaster. Here are 3 30second snippets of a signature ABBA song.
    This demo shows EXACTLY a distributed CD, DolbyA equivalent decoding, and original vinyl version.

    If you listen to the fully DA decoded copy -- it is amazing -- even better than the original Vinyl version (generally more dynamic range/etc.) The entire MamaMia-DADecoded.flac in full detail is on the repository below. I might have left a few more examples on my repo also (not sure.) A real DolbyA HW unit can do similar to my SW (maybe a bit more distortion), purpose here is not to talk so much about my decoder -- it is about the fact that LOTS of DolbyA material has leaked out.

    Also added a Carpenters's example from HDtracks. You'll notice MUCH MORE spatial depth in the decoded copy. Absolutely NOTHING was done to the Carpenters recording except DHDA decoding it.

    Attached: NOT ATTACHED HERE ON HOFFMAN
    First -- MamaMia-Vinyl.mp3 Ancient vinyl rip
    Next -- MamaMia-UnDecoded.mp3 From 1992 ABBA Gold CD, direct conversion to MP3
    Final -- MamaMia-DADecoded.mp3 Decoded from exactly the same CD 'mastered' to sound similar to the vinyl, but NO sweetening -- just simple M+S spatial widening by 1.414 to match vinyl (from the same file, using my V3.8A decoder.)

    Also Attached (Carpenters Top Of The World, from HDtracks, Singles album):
    First -- CarpHD-UnDecoded.mp3 From Carpenters Singles Album
    Next -- CarpHD-DADecoded.mp3 Decoded directly from Album file -- NO further processing.


    Here is a repository with a few example also:


    Hightail Spaces

    John
     
  22. Chris Schoen

    Chris Schoen Rock 'n Roll !!!

    Location:
    Maryland, U.S.A.
    Sounds like a curse. :shake:
     
    fogalu likes this.
  23. HDOM

    HDOM Well-Known Member

    I had heard 24 bit 48khz that sound it really good!
     
  24. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    Hmm. My wisdom has always been that this was the best way to digitise a mono source, because summing bears the risk of removing high frequencies if any phase error is present. What is your opinion about this?
     
    cmcintyre and Grant like this.
  25. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    No, analogue has a limited frequency range just the same. You can't store a signal with unlimited frequency range, it's against the laws of physics. So bandwidth limitation is not a specific shortcoming of digital, but a fundamental property of our universe that surfaces in both analogue and digital applications.
     
    ElevatorSkyMovie likes this.
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine