Is the DR database really accurate for vinyl?

Discussion in 'Music Corner' started by Mij Retrac, Oct 3, 2013.

Thread Status:
Not open for further replies.
  1. Mij Retrac

    Mij Retrac Forum Resident Thread Starter

    The issue hear is the vinyl IS adding distortions to the sound and I believe it is affecting the DR number, maybe not 4db but maybe it is. The distortion that is added by the vinyl is typically distortion that is enjoyable by most vinyl lovers. There is nothing wrong with that its just the reality. The bottom line is and the biggest reason I posted this thread was because this proves that the DR database cannot be reliably used to determine whether the vinyl and CD released at the same time are the same master or not. I can't tell you how many times I read on this forum that someone is gonna hold out for the vinyl because its more dynamic than the CD assuming they used a different master. Then they show the numbers from the DR database to prove their point. If you like the sound of vinyl more than CD thats fine but the vinyl is adding things to the original master so if your a purist and like getting as close as you can to the original master, vinyl isn't always gonna do that for you.
     
    Chooke likes this.
  2. Bill Hart

    Bill Hart Forum Resident

    Location:
    Austin
    This is a fascinating thread, and kudos to Mr. Shepard, not only for his feedback but his graciousness in responding to the various questions and comments. I have a few questions/observations as well:
    1. Mike Fremer made a blog entry on his Analog Planet site comparing a digital file to the vinyl mastered from that same digital file (I think it was a Tom Waits album). Fremer concluded that he liked the sound of the vinyl better; he didn't attempt to explain the mastering process involved and the mastering engineer (I think it was Bernie Grundman on the vinyl side) didn't weigh in either. Fremer simply acknowledged that it may be a matter of 'euphonics'- that the vinyl (whether the mastering, the medium or the playback chain was not specified) adds something he likes. This precipitated an endless- and in my view unnecessarily rancorous debate on another forum- about what 'changed' between the digital-digital playback and the digital to vinyl version. Many (including me) attributed that to some 'special' mastering that was done just for the vinyl release. But, according to Ian Shepard, here nothing was changed in the mastering between the two formats. So, the difference in end results (not by the DR tool measuring results, but by ear) is solely the result of the vinyl playback chain.
    2. Is there a more accurate way to measure what we are hearing than this DR tool? And, in particular, comparing a digital recording with an analog recording made simultaneously from the same performance? Obviously, there will be many variables that differ in the process, from the recording itself to the mastering, but comparing a 'native' analog source all the way through might be instructive. Ralph Karsten, the maker of Atmasphere amps, is making analog recordings and claims that vinyl is capable of being 'cut' very dynamically-the limitations, he claims, are less format inherent, and more the result of the limitations of the cartridge and tonearm. (I suppose this would explain the forum favorite- the so-called "RL" cut of Led Zep II, which can be played back with aplomb over modern turntables, but apparently caused problems when first released, given the state of tonearms and cartridges on mass market phono players circa 1970).
    3. One observation on my part; when EMI released The Beatles boxed set on vinyl in the fall of 2012, there was considerable debate about its sonic merits, here and on other fora. Out of curiosity, I bought one record from that set, Revolver. Compared to an older Parlophone release of the same record (the latter from the first series old Blue Box), the "new" Revolver vinyl did not have all those attributes that vinylistas attribute to the format, e.g., warmth, dimensionality, etc. It sounded very "neutral" but lacked that 'euphonic life' that we associate with vinyl. So, at least in that case, the vinyl playback chain did not give the digitally sourced vinyl added 'euphonics.'
    4. A final thought: As Mij, the OP, suggested when answering whether there is a ''list' of what records have been mastered from digital file, this information is often difficult to come by; almost like food products that are often required to disclose ingredients, it would be helpful to have this information as part of the release, e.g. the Simply Vinyl reissue of Dylan's Blonde on Blonde, which was first released from a tape source, and later, from a digital file.
     
    Last edited: Oct 6, 2013
    eyeCalypso and Mij Retrac like this.
  3. kevintomb

    kevintomb Forum Resident


    Yes, I am just trying to guess what would cause the difference, but as far as the distortion thing, there was a study or something done, and people actually "prefer" a small amount of distortion to zero distortion when hearing music. I have no clue where I read that as it was years ago, but apparently they did some test, where they could slowly add harmonic distortion to a music signal, and most agreed it sounded Better WITH some amount of distortion than with none at all. But most agreed after a certain point, the distortion became a negative.

    Im not even sure where I found that at, but it was interesting to me.
     
    Mij Retrac likes this.
  4. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    This topic about the possibility of TT Meter not being accurate for all source material should not turn into a Vinyl vs CD or Analog vs Digital debate... these are very different issues. Doing that only introduces confusion and serves no meaningful purpose.

    The exact moment TT Meter is said to be unreliable for unknown reasons is the exact same moment where it becomes unusable for any Vinyl vs CD discussion, this is going nowhere for as long as some members try to push the original question into this Vinyl vs CD territory.
     
    floweringtoilet likes this.
  5. vinylphile

    vinylphile Forum Resident

    It's hard not to move into that territory when "distortion" is just assumed to be the culprit for what appears to be better numbers.

    However, overall I think we probably all agree that all that really matters is what we hear.
     
    eyeCalypso likes this.
  6. Mij Retrac

    Mij Retrac Forum Resident Thread Starter

    The main reason I posted this is because of the CD vs vinyl debate. Many people here buy the vinyl version of a new release because they think it has a different master because they look at the TT meter measurements. There have been many discussions and even arguments I have had on several posts where people swear that the vinyl has to have a different master because the TT meter says it's at least 3 to 4db louder. This disproves that and proves my theory that the TT meter is not a reliable way to differentiate masters when comparing CD and vinyl that came out at the same time.
     
  7. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    Mij, that's fine, I understand that and I would very much prefer that more people used their ears instead of these tools to be confident about their choices and preferences. But you should understand that the very same reasoning should apply to any data that TT Meter looks at, be it retrieved from a CD, Blu-Ray, Reel Tape, Vinyl, straight from a mix console, or whatever... this is not a Vinyl issue, this is not a Vinyl vs CD issue. It seems to be more some kind of EQ issue than anything else... or maybe it's no issue at all.

    As far as we know, right now, TT Meter might be always right, or always wrong, or maybe there's even no pattern to its readings... and we don't know "why" for any of those possibilities. How can we draw any conclusions from that?

    Ian and others might be on to something interesting about TT Meter, that's all it is in my opinion, and until that is cleared up and proven and fully understood it's not doing any good to mix that debate with the Analog/Vinyl vs Digital debate as it just confuses both debates and is not helping at all.

    Follow the White Rabbit! If you guys found TT Meter can be deceived by adding stuff to the source file, then do it in the digital domain, do some EQ tricks, add distortion or surface noise, even those "vinyl like" plug-ins might help, see if anything like that can make TT Meter get a +4dB reading... I'm sure TT Meter can't tell by magic that the data in the file cam from a turntable! Be scientific about it and undertand what it is before going the easy vinyl beating based on theories that are, at this point, VERY FAR for being any kind of facts...

    I've read a lot of bad information and generalizations in this topic about vinyl as a format and those who like to listen to vinyl as their best sounding format... it's not fair that this TT Meter thing is being used as a weapon to bash vinyl and people who love it, there is nothing there to support it, just theories and pure speculation.

    This ends up being very tiring and destroys the original topic purpose because after a few posts it all starts to look like a group of people desperately searching for technical reasons why we shouldn't like and prefer to listen to vinyl... I mean, what's the point in that? We just like it! It sounds great! I'm not constantly searching for technical reasons to justify why I love to listen to vinyl records (and there are plenty)... just the same way that I don't try to justify my love for Lasagna or Football :) Sure I once tried and tested and read a lot about this and became quiet well informed about the A vs D issue, and then I let it be... I'm not obsessing over it, I'm enjoying it! :)

    In this topic there was no need for this, if we all stick to the main issue (as we should, it's quite interesting actually), there's no reason why the focus should turn to Analog vs Digital here.
     
    Akhorahil, jimhb, Leif and 1 other person like this.
  8. vinylphile

    vinylphile Forum Resident

    Agree with every word you say.

    However, re the bold part at the top - I think the problem for lots of us is that we can't listen until we buy - and we can't compare unless we buy both. Therefore we do look to things like other listeners' opinions and DR values to try and make some kind of educated guess about which may sound better before shelling out our hard earned bucks.
     
    Akhorahil, Leif, Mij Retrac and 2 others like this.
  9. Mij Retrac

    Mij Retrac Forum Resident Thread Starter

    I really haven't seen a whole lot of vinyl bashing here myself. I see explanations as to why the vinyl will get different results on the TT meter and how the TT meter can't give a definitive answer to anybody interested in a new release as to which version may sound better or whether the CD and vinyl have different masters or not. If you consider relaying the facts about vinyl and some of its shortcomings bashing vinyl and vinyl lovers well I think you might be just a little sensitive on the subject. I actually buy more vinyl than CDs nowadays because I prefer the sound of vinyl when it's mastered right because it usually is higher resolution. However if I know that the original master is a 24/44.1 or 24/48 master then I think that there isn't enough of a difference in the resolution to make up for the things I don't like about vinyl so I will just buy the CD. These things only take place if there isn't a high resolution digital version of the album available however. To me that is the best way to listen to music in this day and age. I think it gets you the closest to what was going on on the studio which is what I'm looking for.
     
  10. Ian Shepherd

    Ian Shepherd Active Member

    OK, I've been playing around with this, and I think I know where this apparent difference in DR comes from. But before I get to that I'd like to make (and re-state) a few points.

    - I'm not saying (and this video doesn't show) that vinyl isn't more dynamic than CD, sometimes. There are definitely vinyl releases that sound more dynamic than the CD versions, because of different mastering choices. Others don't - for example this one. ALL I'm saying is that the TT meter isn't a useful tool to distinguish which is which.

    - Even with a great master, there may well be tweaks made at the vinyl cut, as I mention in the video. Examples include the RIAA curve (which is mandatory), and processing specifically to optimise for the format - for example extra de-essing, elliptical filtering (to prevent issues with phase and level in the bass) and changes to the overall stereo width.

    - None of these adjustments are "artistic" or "creative" decisions that will enhance or significantly change the perception of the original master, they are all technical issues based on the particular properties of the vinyl format. They're simply intended to make sure the material "translates" successfully to vinyl in the same way that CDs can't contain frequencies above 20 kHz, or peaks above 0 dBFS

    - However in the sense that they change the source a little during the cut, you could argue that they represent a separate "master". Personally I don't categorise them in that way, but there's room for semantic disagreement. For me, a separate master is (for example) one with less compression, or a markedly different EQ.

    - I don't know for certain that nothing was done at the cut of this record - I wasn't there. But I do know that the band were very happy with my digital master, and asked me for a version that could be used for the vinyl, as well. I've simply level-matched the two and made the observation that I hear no massive differences, beyond what I can easily ascribe to RIAA processing and variables in the replay chain.

    - Finally, anyone trying to argue that the apparent extra PLR (peak-to-loudness ratio) of the vinyl represents "genuine" musical dynamics needs to let it go. There are no extra musical dynamics in this digital master file to be "discovered".

    So, having said all of that, I do think I know where the apparent DR increase comes from in this case, as I mentioned at the start of this post.

    In the video, I mention that I can hear a change in the image. More specifically, I hear it as narrower, on the vinyl. This is quite normal - there is always a certain amount of crosstalk between the left and right channels, and it's also quite common for the cutting engineer to choose to narrow the stereo image in the bass region, or even overall, as I mentioned above. I have no idea if that was done in this case - the image on the CD is quite wide, so it's perfectly possible.

    Either way, I tried the same thing on the digital file, ie. narrowing the stereo image slightly. And lo and behold, the measured DR value increased to... DR12! Looking very similar on the meter to what we see on the vinyl.

    So, it may be as simple as that. Either a tweak during the cut, or just normal crosstalk, is sufficient to alter the waveforms so that their peak level increases enough to account for the "extra" DR points.

    And before anybody asks, no this isn't a "real" increase in dynamics, it's just a side-effect of the audio processing.

    Now to the really important part, for the purposes of this thread.

    The channel crosstalk I mentioned will certainly play some part in what we're seeing, even if the engineer also made a tweak. And it happens on every vinyl release, and just like the EQ, it depends on the playback system.

    In other words, the same record played on different decks may exhibit a different degree of image width, and measure a different DR value as a result - which is a point that's been made several times in this thread already. What's interesting to me is that this change alone is sufficient to cause the spurious DR boost, and it will happen to some extent whenever a piece of vinyl is played.

    Which just supports the whole point of the video - the TT meter can't be used to prove anything about the dynamics of a vinyl release compared to a digital release - or indeed to anyone else's vinyl rips.

    Maybe it could work up to a point for comparing your own rips with each other, but personally I have my doubts about that, too...

    Ian
     
  11. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    Now we're talking :) That's great information Ian! Who would have guessed that a narrower image separation could cause such a different DR rating? I'm impressed! I suppose this is a flaw with TT Meter algorithm... a serious flaw so it seems. Now, wouldn't this happen for every data file? I guess crosstalk is not exclusive to vinyl rips...
     
  12. ad180

    ad180 Forum Resident

    Location:
    USA
    Thanks Ian, that makes a lot of sense.
     
    Ian Shepherd likes this.
  13. Ian Shepherd

    Ian Shepherd Active Member

    No, not at all. The TT meter was designed to work with lossless digital files - which it does. You could call this a "drawback", but I can't think of a way of getting around it. As with all meters (digital peak, PPM, VU, LUFS) you need to use it with understanding, and when appropriate - and on a vinyl rip it's just not.
    Well, crosstalk like this doesn't happen in digital at all. On analogue tape it happens, but to a far lesser degree. So actually, it's pretty unique to vinyl.But what is true is that any kind of processing may mess with the results from the TT meter. Basically it should only be used with a lossless digital file and no processing. Otherwise, all bets are off.
     
  14. Steve Hoffman

    Steve Hoffman Mastering Engineer Your Host

    Why can't each and every file (digital and analog) be summed to mono first and then measured? Wouldn't that eliminate the ********?
     
    Matt I and SergioRZ like this.
  15. kevnhuys

    kevnhuys Forum Resident

    Location:
    Brooklyn, NY
    Yes! The TT Meter is measuring crest factor (peak vs average), not actual dynamic range. Most discussions of 'dynamic range' in forums like this are really about crest factor

    One thing to also consider, is that perceived loudness is a psychoacoustic phenomenon that doesn't track simply with crest factor. There's debate over how to model it properly for use in, say, tools like Replaygain , that try to 'normalize' perceived loudness across multiple tracks.

    Finally, here's a short but pithy list of the wrong ways to determine if a vinyl master sounds better than a CD master

    How do you know if a vinyl master is audibly superior than the CD master?


    Overall, TT Meters and its ilk, as well as waveforms, are pretty blunt tools for predicting how something will *sound*.
     
    Ian Shepherd likes this.
  16. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    You mean crosstalk of this order of magnitude, right? Crosstalk exists even in the digital domain, doesn't it? And as you mention, also on analogue tape there is some crosstalk (that's why in previous posts I questioned if there would be a similar behavior if the data file was transferred from analogue tape)... less than the expected crosstalk from vinyl playback on most systems but still a considerable amount, no?

    And this is one feature that you identified as being capable of messing with TT Meter's "math", who knows what else is at play here, for all files from all sources...
     
  17. Ian Shepherd

    Ian Shepherd Active Member

    No, because then it's not an accurate representation of the source file. To give an extreme example, imagine something where the drums were in anti-phase - yes, I've seen this...! When mono-ed they would cancel out completely, which would dramatically change the DR reading as compared with the same thing in stereo. The same factors would apply to different degrees to less extreme cases.
     
  18. Ian Shepherd

    Ian Shepherd Active Member

    No.
    No. Far less.
    Actually I listed a whole slew of them, in the video, and I've said repeatedly here that ANY form of processing could mess with them - especially the RIAA encode/decode process - basically anything that isn't phase linear (which is a lot of processes).

    But none of these happen in the situation the meter was designed for - lossless digital files.
     
  19. kevnhuys

    kevnhuys Forum Resident

    Location:
    Brooklyn, NY
    Mr. Ludwig makes some claims of 'fact' there that are anything but -- in particular, him citing Dr. Diamond's bogus work about the negative physiological effects of digital, is really sad.
     
  20. Steve Hoffman

    Steve Hoffman Mastering Engineer Your Host

    Agreed.
     
    Ian Shepherd likes this.
  21. kevnhuys

    kevnhuys Forum Resident

    Location:
    Brooklyn, NY
    Well, it's not quite that simple -- there is more than one way to calculate average RMS. If you read the TT Meter technical paper, you learn their choices in windowing, how to treat low-level peaks, , etc.

    Regardless of that, the point of Ian's demo is that the extra 4 dB in this case cannot have come from using a source with wider 'dynamic range' -- which is the assumption audiophiles often make when they see comparisons like this. Therefore the extra 'dynamic range' must come from some other process, and is possibly even artifactual. An analogy would be to look at the spectrum of an SACD vs the CD and see a band of 'signal' way above 20kHz, and conclude from that, that the source of the SACD had a wider frequency range than the source of the CD (when in fact the band you're seeing in the SACD spectrum is due to noise shaping).
     
    Ian Shepherd likes this.
  22. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    There is digital crosstalk. Every DAC shows a Crosstalk figure in tech specs, actually any digital circuitry of any kind gives quite a bit of importante to Crosstalk... but I understand what you mean, it's so little that it can be safely considered completely harmless for the purpose of this discussion.

    You did list/mention other factors, but one thing is to talk about them, another thing is to actively test each of them to get/share measured results (note I did not say objective results, at this point there are too many unkowns and variables involved for it to be truly objective).

    I think this last step you did is really something useful and can really help us all understand what is going on and how, if at all, TT meter or other tools can be safely used to compare audio files and in what controlled conditions it can or cannot be done.

    As for me, I'l just keep using my ears :D
     
  23. What about calculating the DR for the left channel and again for the right channel? Would that offer any sort of benefit in comparing the vinyl DR numbers to the CD DR numbers? I'm guessing it wouldnt.
     
  24. floweringtoilet

    floweringtoilet Forum Resident

    You would have some crosstalk from the left channel in the right channel and vice versa, even if you measured them independently, so it really doesn't get you anywhere.
     
  25. tmtomh

    tmtomh Forum Resident

    I just had to quote this again and say, Wow. The first and only explanation I've heard for the "vinyl DR factor" that makes sense.

    Thanks!
     
Thread Status:
Not open for further replies.

Share This Page