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

    Wow Ian, you rock!

    Thanks for taking the time to do this and for the time you have spent on the thread! I'm glad we have found a plausible explanation for what is going on so we have a better understanding.

    I know the original video helped me out a lot to confirm what I already suspected, it gave me validation.

    I know I will be a little more hesitant to buy a vinyl new release without knowing the pedigree of it first.

    Thanks again!
     
  2. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    This narrower stereo image causing TT Meter to read more DR made me remember Stereo vs Mono and how many times Mono recordings or mixes are described as sounding more "powerful" by so many people... it seems to be a similar phenomenon. I wonder if the narrower separation of the stereo image is allowing for more concise and focused transients on certain frequencies and that is causing TT meter to look at it as increased DR, and it is in fact perceived by most listeners as being more dynamic...

    That would make TT Meter actually a better tool by giving it a certain Human quality in the way it hears the music :D
     
  3. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    A couple of things more...

    First, I believe there were a few cases where CD and LP releases had very similar DR ratings (please correct me if I'm wrong, this is not something I usually follow). So, what can be said of those situations about TT Meter and Crosstalk? Maybe the vinyl rips were made from better cartridges or converters?

    Second, I find this part of Ian's post to be very important:

    As we all know, during the actual vinyl mastering a lot could be done to change the sound of the source file... Ian, is it at all possible to check this, maybe asking the people who cut the record what exactly was done?
     
  4. Ian Shepherd

    Ian Shepherd Forum Resident

    Glad you guys found this helpful, I think I'll post it to my blog as a follow-up.

    A few replies:

    - There will be a certain (tiny) amount of crosstalk in the analogue playback of a digital file, but the TT meter works directly from a digital source, where there is no crosstalk.

    - The TT meter already treats left and right independently, the crosstalk in vinyl arises during the cut and playback

    - A narrower stereo image may mean the centre information is relatively more prominent, adding "solidity" to mono elements (often bass, drums and voice). However it will also decrease the impression of space, 3-dimensionality, "air" and impact for elements which are hard-panned (eg. often rhythm guitars). So overall I think it's swings and roundabouts in terms of the "human quality". While it may influence our perception of the dynamics, it doesn't affect the actual dynamic content.

    Once again - if the source master is compressed and limited, nothing can genuinely undo that processing.

    There are plenty of situations where we can imagine the CD and vinyl might have similar DR ratings, for example a narrower stereo image to begin with (crosstalk will have less impact on the waveforms) or little sub-bass in the source (RIAA filtering won't cause as much phase smearing, which will influence the waveform less) but honestly, it's not worth over-thinking all this stuff.

    The TT meter is unsuitable for measuring vinyl rips. Let's move on ! This conclusion doesn't reveal any deep truths, or undermine anyone's love of vinyl, it just says "this is not a suitable tool for the task". Time spent trying to figure out exactly why might be an interesting intellectual exercise but not one that will help us enjoy the music more. (Of course I realise the irony of saying this, having posted a video on the subject - but that's the point ! The video is trying to say - "this doesn't work, let's go and spend time listening more carefully instead").

    I feel the same way about quizzing the cutting engineers - yes there's a lot that could have happened at the cut, but I'm sure that it didn't. I spend my life listening critically to fine details in music recordings - I'm satisfied that this was a flat transfer, with the exception of the change in image and EQ - both of which could very plausibly be due to playback variables alone. These two factors only, coupled with the need for RIAA filtering, explain what's going on here perfectly well.

    Ian
     
  5. vinylphile

    vinylphile Forum Resident

    Good points Sergio - I was wondering these exact things myself.
     
  6. vinylphile

    vinylphile Forum Resident

    It may not be a perfect tool for the task - but unfortunately outside of buying both and comparing ourselves it's pretty much all we have. If I found it completely useless I would have stopped looking at the numbers long ago - kinda like most hardware specs.
     
    Leif, WNind and SergioRZ like this.
  7. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    Ian, what I meant by "Human Quality" was not to say that it sounded better because of the Crosstalk effect narrowing the stereo image, or that it sounds more "Human" (whatever that might be) nothing like that at all, what I meant is that the TT Meter is actually behaving a little like most Humans do when such audio effect is perceived as having increased "punch", better defined transients, a more powerful presentation or more Dynamic Range (as these are the things we see many times being mentioned about the virtues of mono recordings or mixes), so in a way I find it interesting that TT Meter is basically saying the same about its music and likes a good mono mix :D :D :D
     
  8. Ian Shepherd

    Ian Shepherd Forum Resident

    So you're kidding ? In that case yes, absolutely :)

    Personally I can't bear pure mono, so I disagree with all those comments !
     
  9. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    Ian, when you say "There are plenty of situations where we can imagine the CD and vinyl might have similar DR ratings, for example a narrower stereo image to begin with (crosstalk will have less impact on the waveforms) or little sub-bass in the source (RIAA filtering won't cause as much phase smearing, which will influence the waveform less)" it becomes clear that many factors that occur very much before vinyl playback are at work to change the way TT Meter will read DR. I mean, you're talking about recording techniques, mixing options, even musical content (artistic options)... again, this all points to the TT Meter tool being just as vulnerable to that as it might be to vinyl playback related audio effects... all these factors are independent of the source being analog or digital...
     
  10. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    Not kidding at all, I'm clearly saying that TT Meter seems to perceive that audio effect in a similar way that we Humans do. Maybe the algorithm was designed that way for this very same reason? It can be argued that this exact type of approach makes the tool more accurate for real world music playback when the sole objective is to know how good it sounds and not to delve into endless technical mining...

    I have no preference for Mono mix or recording... did I say anything to make you disagree?
     
  11. Ian Shepherd

    Ian Shepherd Forum Resident

    In that case sorry, no. I don't have time to debate any further - the TT meter was designed to measure the peak-to-loudness ratio of digital sources. It does an OK job of that, and with care you can draw conclusions about them. Beyond that, it's stupid !
     
  12. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    So if TT Meter already processes each channel separately, that helps us understand something more about what happens.

    If there is added crosstalk on one of the files (considering a comparison between two music files from the same source and all that), this means there is more information being processed for each channel separately... this simple fact probably accounts for some dB difference but it seems like 4dB is a lot for this reason alone. Doesn't it?

    So let's say it might be enough for 2dB difference which then TT Meter adds from both channels to make it a total of 4dB? (I have no idea)

    If that was the case, then what we're talking about is in fact a 2dB measured difference (at least for "data file" level analysis, playback would be something different). Now that doesn't seem so scary... :) Maybe that's basically the same perceived difference we would hear durinhg playback, depending on the music content and how much certain types of sounds or frequencies are present?

    I wouldn't neglect the possibility that TT Meter is perceiving the crosstalk effect as added dynamics for the very same reason we Humans many times perceive it that very same way when listening to mono sounds. Also, it seems plausible that TT Meter might do that for any source file regardless of analog/digital source, regardless of CD / Vinyl rip... for the TT Meter Tool it's just crosstalk, it can be music that is mixed that way recorded, mixed and mastered 100% digital, doesn't really matter...
     
  13. Ian Shepherd

    Ian Shepherd Forum Resident

    No, maybe I wasn't clear. Those factors may influence how the TT meter reads the same music on vinyl as opposed to digital, introducing yet more uncertainty. The TT meter reads what it reads, 100% reliably, on digital lossless sources. (The usefulness varies, depending on your point of view) After that, all bets are off.
     
  14. Ian Shepherd

    Ian Shepherd Forum Resident

    Sorry, you seem like a nice guy, but this is the only thing you've said today that I agree with.

    No offence, but you need to stop imagining what you hope is true and just accept - the TT meter is not suitable for measuring vinyl.
     
    Mij Retrac, kevnhuys and ridernyc like this.
  15. SergioRZ

    SergioRZ Well-Known Member

    Location:
    Portugal
    Interesting, so what you're saying is that TT Meter for some magical reason ignores any audio effects such as "narrower stereo image" or "little sub-bass in the source" unless it comes from a turntable... in that case it just adds 4dB of DR to the reading.

    And you expect me to take that seriously?

    You tested it yourself, 100% on digital domain, narrowing the stereo image, and the result was 4dB increase in DR as you yourself tested and described here. Couldn't that narrow stereo image be just part of the original mix? How would that be 100% reliable (as you say)? How does that say TT Meter is good for 100% digital source and not good for a vinyl rip if your own test shows that the exact same behavior occurs working exclusively with the digital source and no vinyl rip in sight?
     
  16. Thurenity

    Thurenity Listening to some tunes

    Ian,

    Been lurking for awhile now as much of the tech aspects of the DR meter is likely over my head. But something to consider : when someone like myself does a needledrop, my drop is no longer analog, it's digital. For all its flaws with rumble and crosstalk and distortions, it's not vinyl anymore. It's basically its own separate entity...a "snapshot in time".

    So, with that in mind, if it's unreliable to measure the drop (and I do understand some of your points and even did some of my own tests), why would the meter be reliable to measure a CD rip? Couldn't a mastering engineer have taken a brickwalled MP3 source, for example, tweaked the EQ a bit when pressing the CD and improved the numbers and you've got essentially the same problem? Without knowing the source of the CD or what the engineer has actually done, I would think these numbers would be in question regardless of the source as we've both shown that it can be "fooled".

    Same for SACD / DVD-A rips too -- I think that the "use your ears first" mantra really applies regardless of the source.
     
    vinylphile, SergioRZ and dkmonroe like this.
  17. dkmonroe

    dkmonroe A completely self-taught idiot

    Location:
    Atlanta
    I was wondering this as well.
     
    SergioRZ likes this.
  18. dobyblue

    dobyblue Forum Resident

    Are you attempting to refute something that wasn't said? That's quite puzzling.\

    The vast majority of modern rock/pop CDs from major labels have excessive amounts of dynamic range compression applied, hence the article is pretty redundant. You DO want a separate vinyl master created for the vast majority of modern releases, otherwise they sound like gash.
     
  19. Stefan

    Stefan Senior Member

    Location:
    Montreal, Canada
    I've also been lurking through much of the thread but appreciating the information and generally positive tone.

    Your point is certainly a valid one and in a way, it backs up Ian's premise. The process of cutting the digital master to vinyl and then playing it back alters the information in a variety of ways. Thus comparing your needledrop snapshot to the original digital file with TT meter or anything else is definitely apples and oranges. It would be no different from taking a CD rip, altering it in some way and then comparing the altered version to the original. Any difference measured would be the effect of the alterations.

    As I noted in another thread earlier this year, for the TT meter to be in any way valid with vinyl rips you would need to compare only vinyl rips and not vinyl to CD, and only if all the rips being compared were done using the exact same equipment. Even then, you have no idea what variables the cutting of the vinyl introduced. So in the end, it's more or less useless with vinyl, which of course is the original point here.

    We can debate the usefulness of TT meter for CDs (personally I hate how some folks have reduced assessments of sound quality down to this one isolated factor), but for vinyl, it seems to be a complete waste of time. Life is too short. Time to enjoy some music. :)
     
    Mij Retrac, Ian Shepherd and ad180 like this.
  20. Ian Shepherd

    Ian Shepherd Forum Resident

    Well said, Stephan.

    This isn't an analogue versus digital argument, or an anti-vinyl argument. Yes, the results of a digital file could be skewed by applying extra processing, but that's not how most people use it - if they did, there would be no consistency at all in the DR database, for example.

    The TT meter will always give the same result when used on the same digital file. (In that sense, I say it's 100% reliable, although you may question the value of the measurements).

    But it may well (probably will ?) give entirely different results when analysing vinyl rips from different decks/stlyi, even when playing the same piece of vinyl.

    For that reason it will give unreliable results at best when measuring vinyl rips, and certainly you can't compare those results to numbers measured from purely digital sources.
     
    tmtomh and Thurenity like this.
  21. Ian Shepherd

    Ian Shepherd Forum Resident

    My article was aimed at people who are interested in having their music mastered. Ie. I hope to encourage them to have great, non-smashed masters made, so the CDs don't sound bad.
     
    Tullman and Mij Retrac like this.
  22. Mij Retrac

    Mij Retrac Forum Resident Thread Starter

    I'm not arguing the validity of wanting a separate master from the digital/cd version especially if the master is badly compressed and limited.

    I am saying it isn't a necessity to make one and most new albums are done without them. Now, do the records sound like crap? Yes, but that isn't the point I was making. I know of several new releases that have the same crapy squashed mastering on the record and the CD. It's a shame but unfortunately a reality.
     
  23. dobyblue

    dobyblue Forum Resident

    It hasn't been my experience, and will take the word of Ray Janos and Stan Getz over someone else telling me that they're probably lying if that's okay with you. I also know from direct contact with the project managers of the Live Trax series that they are being mastered separately for vinyl from the original sources to be more dynamic than the digital releases.

    Several =/= majority. There are many examples of both.
     
  24. Mij Retrac

    Mij Retrac Forum Resident Thread Starter

    I'm not sure how you got that I am calling anyone a liar out of my last comment.

    I merely said I know of several new release vinyls where they didn't have a separate master for the vinyl and CD. I didn't say all new releases were handled this way. If Ray Janos and Stan Getz make separate masters for their vinyl, great! Although I don't know how Stan Getz could make those claims since he is no longer with us. I merely stated that I have unfortunately purchased many new release records that sound like they used the same master as the crushed CD since they sounded equally bad. Most have been in the Pop and Hard Rock genres. I only know of a couple of acts (like DMB that you referred to) in those genres that are known to spend the extra time and money on separate masters and I applaud them, I wish more people did that especially considering how much more they charge for the record.
     
  25. ClausH

    ClausH Senior Member

    Location:
    Denmark
    Getz is a mastering engineer.
     
    Mij Retrac likes this.
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine