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

Hi-Res 24/94 vs Flac vs CD vs Mp3 files download comparison

Discussion in 'Sound Science' started by christian u, May 26, 2014.
1 2 3 4 5 6 7 8
  1. BartSimpson1976
    That article raises the bar very high in terms of utter nonsense BS. 
    MQA is very nice in terms that the size of lossless files will be smaller.
    Besides from that lossless is lossless and you won't hear a difference whether
    it is 16/44.1-FLAC, 24/96-FLAC, WAV, ALAC or MQA.
    But I don't want to open Pandora's Box here...
  2. gregorio
    No, that's impossible. Firstly, even if MQA did have more bits/resolution, it wouldn't make any difference as we can't get anywhere near using even half the resolution of 24bit anyway! Secondly, if the original recording/mix was 88.2/24, re-encoding it with a different codec, any codec, cannot produce more resolution, at best only exactly the same resolution. Both of these points are irrelevant though, because MQA does not have more bits than 88.2/24, it has less!
    That's not necessarily true with MQA, from what I understand. As I understand it so far, MQA doesn't necessarily just compress the data, as do FLAC/ALAC, it can also process the audio itself. Some form of audio compression, EQ and/or other processing is/can be automatically applied and that would effect an audible difference.
  3. Roseval
    What I understand is MQA can do two things:
    Lossy compression of Highres recording
    Compensating for time-domain errors

    It compresses Highres audio into a 24 bit / 44.1 or 48 kHz PCM format.
    As the result is PCM, it can be treated like any PCM format.
    It can be contained in any lossless format like WAV, FLAC, ALAC, etc.

    The lossy compression is rather complex.
    It is based on the assumption that below -120 dBFS recordings contain random noise.
    Compress the part above 44/48 and store this in the bits 20-24.


    As this space is limited, it is a lossy compression.

    As the lower bits are used to store the compressed part, you won’t have the original 24 bit dynamic range.
    The result is a 24 /44.1 or 48 kHz PCM audio file.
    This can be played on any DAC with any media player.

    It is more or less equivalent to a 16/44.1 redbook format (CD quality).
    The clever thing is that all what is stored below bit 16 appears as random noise.
    To expand it to the original sample rate, a decoder is needed.
    This can be implemented in software (media player) or in hardware (streamers).

    The other claim is that MQA can compensate for time-domain errors of both the AD converter used to make the recording and the DA converter used for playback.
    This means that a DAC must be equipped with some additional hardware processing the audio and applying a specific algorithm tailor made for this specific DAC.
    It implies that for each revision of the DAC, this as to be done anew.

    The biggest advantage of MQA is probably the substantial compression of Highres.
    If you run a streaming audio service like Tidal, Qobuz, etc.  it saves on bandwidth compared with streaming Hires in its original size.

    The downside is of course as it is a proprietary protocol; somebody has to pay the license.
    In the end of course this will be you.
  4. droopy1592
    Most vinyl has been made from DIGITAL masters for a good long time now. 
  5. gregorio
    I'm not sure that's really a "clever thing", except maybe in terms of marketing. There is no such thing an original 24bit dynamic range and everything stored below bit 16 is already just noise. In commercial recordings, with no exceptions as far as I'm aware, everything below about bit 10 is just noise.
    It appears to be quite a clever marketing ploy though. Store in bits 20-24 all that info which in higher sample rates is inaudible, so it makes no audible difference how lossy that info is. And, I can't see how they propose to compensate for supposed time domain errors of converters, particularly of the ADCs used or even if they can, how that would make an audible difference either.
  6. Ancipital
    Ah yes, reigns surpremely (sic), because taking a pristine digital master, forcing it through an RIAA curve, then cutting it to an archaic lossy medium that inherently suffers surface noise, restricted dynamic range and of course, has to have that RIAA curve beaten back out of it, is somehow magically superior to just playing a properly noise-shape dithered 44.1/16 digital file derived from the same source. The world is truly an amazing place!
  7. Mifi
    Some people likes the sound of a RIAA. Probably has to do with what you grew up with. I just recently heard a real to real 1st generation copy of an analog master recording.  It sounded incredible to my ears, but when I was growing up, I also heard a lot of real to real, maybe it is in my DNA[​IMG] 
  8. pinnahertz
    "Real to real"...made me laugh! [​IMG]
    How big were your Reals?  Reel big!  And they were real reels.  
  9. Mifi
    sooooory, I'm not a native speaker but indeed the reel to reel where really big[​IMG]
    portable_reel_to_reel_tape_recorder_818931.jpg   this big   th1551tascam4track.jpg
    castleofargh likes this.
  10. old tech

    Yep, never ceases to amaze me that the "analog" crowd point to the vinyl record as their poster child, rather than reel to reel or the Hi Fi VCR.  It is a bit like a digital audiophile pointing to the MP3 over CD and hi res.
    Having said that, there is nothing wrong with preferring the sound of vinyl, in particular the character that medium imparts on the sound, but is it better?  Certainly not in any way it can be measured.
  11. pinnahertz
    The RIAA curve itself doesn't cause problems because the record curve and play curve are complimentary, the result is flat (assuming both are accurate, which is not always the case). But since the cutting stylus has a maximum velocity based on physical dimensions of it and the groove it's cutting, the combination of that and the RIAA record curve creates a rather non-flat maximum record level which is usually responded to in mastering with peak limiting that follows the maximum level curve.

    Rather significant distortions occur in playback because of a list of problems, many grouped into the tracking induced distortion.

    HiFi VCRs were HiFi in terms of frequency response only, which was why they were developed over the linear tracks. But the noise in the AFM carrier system mandated a companding noise reduction system, which failed to adequately hide the head switching noise components, limiting the serious application of the technology for any real high fidelity recording. It worked best when there was picture to distract from the sound issues. Some duplication plants used additional dynamics compression in the chain in a failed attempt to hide the noise.
  12. old tech
    They also bettered any consumer reel to reel in terms of dynamic range and signal to noise.  I had a couple units back in the early 90s and the only issue I ever encountered was that some tapes recorded on one did not play back as well on another machine.  I used to record CDs to the VCR (using high quality tapes) and would struggle to hear any difference between the recording and the original CD, and that was with a young man's ears.
  13. pinnahertz
    That's true for static S/N, but not for dynamic S/N...basically noise that occurs during modulation.  VCR HiFi tracks were actually worse than analog tape in that aspect, but with the right program material you might not notice.  With the wrong stuff it was awful. 
  14. pinnahertz
    Basic video tracking was a mess, entrance and exit guide alignment made a huge difference, but even at its best the head switch point created a sort of "click" that modulated with music.  Mistracking made it worse, and at some point the AFM signal couldn't be recovered, the deck would fall back to the linear tracks.
  15. christian u
    These sounds interesting;
    but it seems that they are still not available.
1 2 3 4 5 6 7 8

Share This Page