1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

24bit vs 16bit, the myth exploded!

Discussion in 'Sound Science' started by gregorio, Mar 19, 2009.
346 347 348 349 350 351 352 353 354 355
357 358 359 360 361 362 363 364 365 366
  1. bigshot
    That is mostly true of ephemeral pop music. It isn't true at all for classical, jazz, classic country or classic pre-1980s reissues. All of those genres have experienced huge improvements in the past ten years. Sony in particular has worked wonders on back catalog RCA, Columbia, and Epic titles. Likewise, many 70s rock bands are going through their albums and remixing for multichannel. Just about anything remixed by Steven Wilson sounds better than any previous release. Digital restoration techniques have made a big impact particularly on pre-hifi stuff, and they are working from the original metal parts, not dubs. 78rpm recordings generally are mastered much better than they ever were in the LP or early CD era. There's some wonderful work being done, but it's on classic legacy recordings intended for adults with good systems, not throwaway pop stuff for kids with portable earphones.

    I'd say less than 20% of what I buy is not an improvement over previous releases of the same material. I just got a expanded version of Prince's 1999 that I'm looking forward to. His stuff is crying out for first class reissues.
    Last edited: Dec 5, 2019
  2. gregorio
    1. Yes, I always considered him to not only be very well informed but he also backed it up by producing high quality recordings. My only criticism of him was his adherence to the hi-res bandwagon. I didn't know he'd recently actually done a hi-res vs CD test and admitted he couldn't distinguish them. My respect for him has increased for publicly admitting this and of course, my only criticism of him is no longer valid!

    2. I've been part of the hi-res industry and I've been saying that for over a decade, as this thread evidences. Many others have too but I agree, as an owner of a hi-res label, rather than only as an engineer, I don't recall ever hearing a hi-res label ever admitting this.
    Yes. I'm not sure exactly what your CurveEQ is displaying but it appears there's little material above about 16kHz. So the difference (spike) indicated, an extra 13dB or so between 17kHz - 22kHz, would be about right for many noise-shaped dither algorithms.

    Noise-shaped dither usually employs about 1.5 bits of dither, which is about 9dB of white noise throughout the spectrum (up to the Nyquist limit) but if we shape that noise, move some of it from the critical hearing band and concentrate it in the least sensitive hearing band (17kHz and higher), then you'll end-up with roughly +13dB or so of noise in that high freq band. Remember that noise shaping doesn't reduce digital noise, that's not mathematically possible, you get exactly the same total amount of noise with 1.5 bits of noise-shaped dither as you would with 1.5 bits of triangular (standard) dither but it's concentrated in a smaller (high) freq band and therefore must have proportionately higher amplitude in that band. While +13dB or so of noise above about 17kHz might appear a bit drastic, it's not, it's inaudible.

    Human hearing sensitivity has a sharp fall-off starting around 12kHz - 14kHz, You can see this starting point even in the 1933 Fletcher-Munson loudness contours, which indicates that at around 14kHz -15kHz hearing sensitivity is about 20dB below peak sensitivity. By 17kHz sensitivity is, at the very least, -30dB below peak sensitivity but that's a best case scenario (teenagers with very good hearing), the average twenty something would be at least -40dB and probably lower. So although +13dB of noise might appear to be quite a lot, compared to the hearing sensitivity roll-off, it's insignificant (and inaudible), which of course is the whole point of noise-shaped dither in the first place.

  3. Isloo
    @gregorio - Great thread! I only ever dabbled in HiRes a few years ago. However, I couldn't tell the difference between HiRes and CD quality, so gave up on it. To be honest, I can't tell the difference between 320 Mp3 or 256 AAC and red book, so I save the space. It's really interesting to learn some of the theory behind all of this and to be a little less uninformed. A big Thank you to you and all of the other contributors to the thread who have been kind enough to share their knowledge.
    TheSonicTruth likes this.
  4. TheSonicTruth
    Again, great pitch, bigshot! The late Chick Lambert and you have a lot in common, and he'd be proud of the above :wink:

    I may have alluded to this some while ago, but my issue with remasters is not always one of improvement(that is for surr a subjective measure), but that with many remasters a change has been made to the sound.

    I'd prefer the slightly rolled-off-high-end original version of something, as opposed to something that has had V-shaped EQ and who knows what else applied to it, in a fancy new jewel case.

    Same with remixes: I'll take certain original stereo Beatle albums, where John or Paul is way off in one channel, opposite the meloldy on the other side, with slight background hiss, VS a remixed reissue where those vocals are perfectly centered down to one-tenth of one dB, and elements of the melody are panned modestly to both sides, and dead sterile silence in the quiet passages. That slightly hissy original stereo version is canon, compared to the cleaned, and sometimes loudened up, remix, which makes a sixties legend sound sickeningly modern - a la One Direction, etc.

    Succinctly, bigshot, and Gregorio: in most cases, I'm not always after a collection that sounds 'better'(again, subjective opinion!), but rather, what WAS - and contemporary to the era it came out.
    Last edited: Dec 6, 2019
    Chris Kaoss likes this.
  5. bigshot
    A lot of recordings were far from perfect on original release, particularly mid to late 70s LPs during the oil crisis. Mixing and mastering was done to suit the limitations of the LP format, and when the same masters are used for CDs, the warts can be revealed. Whether or not a release is the "best" or not is both a subjective determination and one that varies from case to case. I don't think it's possible to say that all remasters sound bad or all remasters sound better. The only thing you can say is remasters in general sound different to varying degrees and you have to hear it and decide for yourself.

    The things that can be generalized about is that Sony has done great work improving the sound of the RCA and Columbia library. RCA's dynaflex LPs (prime era David Bowie) deliberately introduced distortion in the mastering to make it sound better on crappy stereos. That is fine if you have a crappy stereo, but if you have a good one, the remastered CDs sound a LOT better. That is also what the problem is with modern ephemeral pop music. It's *designed* from the ground up to sound good on crappy stereos. I played a recent funk CD on my system and the bass nearly blew me out the door. It had been mastered to compensate for lousy bass response in cheap ear buds.

    The trick is to look for the masterings that are designed to be played on good systems. That is easy to do if you are interested in classical and jazz, because just about everything in those genres is designed to be played on first class stereo systems. If you want to listen to current pop music, you are probably out of luck because the audience for that doesn't have fancy stereos... they listen on their phone using cheap buds. In that case, your best bet is to just focus on the music if you like it, and try not to think about the sound quality. It's not apt to ever be released sounding perfect. And if it was, it sure wouldn't sound the same.

    Which Beatles remix are you referring to? Rubber Soul? The stereo version of that was remixed back on first CD release because Martin wasn't happy with the stereo mix. Other than that (and the recent multichannel mixes of Sgt Pepper, White Album and Abbey Road- and Let It Be Naked) all of the Beatles CDs have always sounded basically the same as the masters. The difference between the first release and the recent CD remaster is negligible. I have both and I've compared.

    If you want truly authentic sound, not better sound, you should get a Marantz receiver, a Technics turntable and buy up used vinyl. You can't get more medium-fi goodness than that.
    Last edited: Dec 6, 2019
  6. TheSonicTruth
    No need for sarcasm. My original CD issues are fine for my purposes, as horrid as they must sound to you, lol!
  7. TheSonicTruth
    Such as? I have the first three Van Halen(late 70s Roth era) CDs, quite likely transfers of LP master tapes, and they sound great to me on whatever I listen to them on - my full-sized home stereo, my car system, ripped as MP3 on my phone, whatever.

    What "warts" should I be looking out for? LF roll-off below 50Hz?
  8. bigshot
    That isn't sarcasm. If you are listening to 70s music and you want it to sound exactly as you remember it- not changed nor improved- a turntable is the way to get that. I have phonographs to play my 78s. It's a specific sound signature that recordings in that era were designed to work with. When these recordings get released on CD, they remaster them for CDs and the way the expected audience listens to them. Same goes for streaming. Music is mastered for the medium.
    Last edited: Dec 6, 2019
  9. TheSonicTruth
    When I listen to albums of which I own both the LP and first issue CD, I hear very little difference between them, aside from the absence, on the CD, of background noise and crackle associated with the vinyl. The CD volume *is* a tad higher, but that probably has more to do with the CD deck having a slightly higher output than the turntable.

    Timbre/EQ wise, their signatures are similar, save perhaps for a tad more bottom on the CD. Perhaps *some* processing was done to the master destined for CD, I don't know. But it doesn't seem like enough to make a huge difference, compared to the sound of the LP.

    Now, compare that 40 year old but gently used LP to the recently remastered CD reissue, the biggest difference I hear is how much LOUDER the remaster CD is than either the original CD or the LP.

    Do I consider that remastered CD to be an improved version? No! Just a louder version. Not an important enough criterion for me to want to keep it, except to make these points to fellow music fans and friends.
    Last edited: Dec 6, 2019
  10. bigshot
    I've got over 10,000 records and 10,000 CDs, and I can't make any generalization like that. I think you make stats like that up to support your agenda.

    If your LPs don't sound as loud, you can simply turn up the volume.
  11. TheSonicTruth
    No agenda. The remastered CDs of many popular artists(The Who, Billy Joel, etc) ARE louder than the 30 year old original CDs and the LPs.

    According to what? My EARS. Sorry if my ears aren't a "scientific" enough tool for judging differences in how loud different sources are.
  12. Chris Kaoss
    It isn't a matter of volume level at all.
    I did comparison between the first queen releases and the 2011 remastered ones.
    To me, the former sounds better couz they are more dynamic and natural then the latter.
    It doesn't belong to all remastered old pieces, but with the first releases of cd's of queen or that decade, you would hear it the best.
  13. TheSonicTruth
    Ok, try to wrap your head around the following(because nobody else here seems to be able to!):

    I have two CD issues, of the same 1970s era album that I like: 'A' = a 1985 original CD release, and 'B' = a remastered reissue from within the last ten years.

    I put 'A' in my system and play it, and adjust the volume for comfortably loud.

    Next, I remove 'A' and put 'B' in and play it, without touching my volume.

    The music is now UNCOMFORTABLY LOUD, and I must lower my volume setting to approximate the listening level I set previously, while playing 'A'.

    Why must I set my volume lower for CD B than I set it for A? Because version B of that album IS LOUDER than version A!!

    It's not my 'perception', it's FACT! Just as I am watching the sun rise here in CT as I am typing this!
  14. Chris Kaoss
    This is exactly what i wanted to express. :wink:
    Plus, the music isn't as dynamic, quieter and louder parts, as in the former original records.
  15. castleofargh Contributor
    As you might be able to guess from his profile pic even if you're not yet familiar enough with him, @TheSonicTruth is always grateful for the opportunity to turn any given topic into one about the Loudness War.
346 347 348 349 350 351 352 353 354 355
357 358 359 360 361 362 363 364 365 366

Share This Page