Home of the Squeezebox™ & Transporter® network music players.
Page 1 of 5 123 ... LastLast
Results 1 to 10 of 42
  1. #1
    Junior Member
    Join Date
    Dec 2008
    Posts
    26

    Served music sounds better?

    I had read claims in Stereophile and (I think) The Absolute Sound that music sourced from a hard drive sounds better than music sourced off a CD player. I was skeptical of these claims, and am skeptical of my own ears, which seems to confirm the fact.

    As soon as I started listening to music served by my new SB Duet system, I detected a certain more lively and engaging aspect to the music. I did an A/B comparison to assure myself levels are matched (they are), and on close A/B listening, I cannot identify any difference. Yet, the fact remains that my impression of music played back via SB is more favorable, and this is not a subtle impression, it is a definite emotional "wow".

    Here are the two different music paths:

    path A: Nakamichi MB-10 CD player digital out --> Tact 2.1S digital processor --> Mark Levinson 360S DAC

    path B: Netgear ReadyNAS Duo --> SB Receiver digital output --> Tact 2.1S digital processor --> Mark Levinson 360S DAC

    I cannot pin down any good reason why path B, with the Squeezebox, should sound better. Both paths use the same, very high quality DAC. Up to there, bits is bits, one would think.

    The audio mags claim music servers sound better because lack of jitter on hard drive playback; that doesn't make sense for two reasons. First the Levinson DAC has sophisticated dejitter circuitry, but more fundamentally, the problem with jitter arises with the nature of clock recovery from the S-PDIF waveform content. Both path A and path B use S/PDIF digital signal to transfer the music data.

    The other potential reason is the effect of error correction on CD playback. This would imply that the cheapie CD-ROM drive in my laptop computer does a better job of reading music data (i.e., lower raw error rate) than my CD player, giving rise to fewer uncorrectable errors (uncorrectable errors cause the software to "fake it" by throwing the data out and interpolating across the gap).

    Anybody know of comparisons in raw error rate in playback on CD player vs playback on a computer CD-ROM drive?

    --Steve

  2. #2
    Senior Member Themis's Avatar
    Join Date
    Dec 2007
    Location
    Montpellier, France
    Posts
    920
    I don't know (and I don't care) about what some people might think @ Audiophile.

    What is sure, is that CD players don't use their error recovery system causing interpolation, except on some rare cases with heavily damaged CDs.
    Most of the time, error correction is done way before the synchronous digital path.

    So, I think that whatever difference may be (for a signal going through an external dac) it can only come from the quality of the SPDIF conversion.

    Also, I believe the following is also true : anybody will find the same signal sounding "better" when it goes through a cheapish device (like the SB3 is, compared to some CD players). I've tested this with several of my friends.
    SBT - North Star dac 192 - Croft 25Pre and Series 7 power - Sonus Faber Grand Piano Domus

  3. #3
    Senior Member mswlogo's Avatar
    Join Date
    Dec 2006
    Posts
    700
    There is a related discussion here http://www.meridianunplugged.com/ubb...2988#Post82988 (See bullet #7) about A/B testing (on subtle differences) which claims it's extremely difficult because of how the brain works.

    Later in the thread he gives a great visual example of how your brain works and it goes like this. He puts a slide up on a projector that is blurry and you cannot quite make out what the image is. Then he puts up a slightly sharper slide of the same image. Now everyone can see a Dalmatian dog in the image. Now he puts the original slide back up that nobody could see the Dalmatian in and now everyone can easily see it. That's because your brain learned what cues were important and once learned you filter out all the distractions and focus on the information important to you.

    When doing audio if you hear some subtle effect like the placement of an instrument on the "Better" system you will hear it on the "Less Better" system just as well. The more you A/B the more it will sound the same !!!

    My feeling is pure redbook CD (if working properly) and served music (if working properly) sounds exactly the same. I certainly enjoy served music more because I can find what I'm in the mood for MUCH more quickly. Each can be enhanced through DSP (dither, upsampling, room correction etc.).
    Last edited by mswlogo; 2008-12-26 at 11:51.
    Transporter/DuetController > SPDIF > Meridian G68 > DSP6000, DSP5500HC, DSP5000

    "It's the speakers and room stupid".

    My Transporter Setup
    Hitch Hikers Guide to Meridian

  4. #4
    Senior Member
    Join Date
    Nov 2008
    Posts
    310

    Money back time

    If those two signal paths genuinely sound discriminably different, then doesn't it follow that at least one component in one of them is either broken, or so badly designed as to be worthless?

    What happens when you take the Tact box out of both chains?
    Last edited by JezA; 2008-12-26 at 11:55.

  5. #5
    Senior Member
    Join Date
    Nov 2005
    Location
    Berlin, DE
    Posts
    1,909
    What you need to do to figure out the difference is to use a SPDIF recorder on the output of both the SB3 and the CD player. Then you can compare them at the bit level to see if something funny is going on.

  6. #6
    Senior Member
    Join Date
    Mar 2007
    Location
    UK
    Posts
    1,304
    Quote Originally Posted by mswlogo View Post
    There is a related discussion here http://www.meridianunplugged.com/ubb...2988#Post82988 (See bullet #7) about A/B testing (on subtle differences) which claims it's extremely difficult because of how the brain works.
    Can I quote from the original: "In Bob's opinion, double-blind listening tests are useful for a limited set certain psycho-acoustic research areas, but generally are not appropriate for comparisons of different equipment. The problem is that the human brain does not respond in the same way to a repeated identical audio impulse. Meridian, apparently, have learnt how to perform such comparisons and, if I understood correctly, it involves learning (training the brain) what a test audio sample sounds like through repetitive listening so that comparisons are possible. Just listening to music will not work."

    If you believe the problem is about AB using repetitive sound clips versus unique sound clips then why draw a conclusion about sighted versus double-blind? This problem is presumably just as acute for sighted AB tests using the same clips.

    What about a double blind test where you swap between A and B whenever you feel like. Sometimes you listen to the same clip again and again and sometimes you listen to whole albums before switching sources. You listen to whatever takes your fancy, mostly "reference tracks" but throwing in some unknowns. Sometimes you listen intently and sometimes you do other things and listen in the background, suddenly becoming conscious of the sound at particularly impressive moments. Change the volume level at any time you're digging the music! At the end of this long process make a simple judgment as to whether A or B is superior or they are the same. This blind test, presumably, would have so many holes picked in it it would look like swiss cheese in five minutes!

    Yet this is the typical sighted listening methodology and most people are quite happy to draw firm conclusions from it. And it's indeed quite possible to perform this test double-blind...you'll be pleased to hear.

    But I say nit-pickers about blind tests should have the decency to nit-pick consistently!
    Darren
    Last edited by darrenyeats; 2008-12-27 at 05:24.
    Check it, add to it! http://www.dr.loudness-war.info/

    SB Touch

  7. #7
    Junior Member
    Join Date
    Dec 2008
    Posts
    26
    Quote Originally Posted by JezA View Post
    If those two signal paths genuinely sound discriminably different, then doesn't it follow that at least one component in one of them is either broken, or so badly designed as to be worthless?

    What happens when you take the Tact box out of both chains?
    As someone once said about pizza (and certain other activities): there's no such thing as bad pizza, it's just some is better than others.

    In this case, there's something more engaging about music being replayed via SB; it's not that music sourced from my CD player sounds bad. Indeed, it's very good. It's just that I find the sound of music sourced from SB to be more enjoyable.

    The amazing thing about the audiophile hobby is small changes that "shouldn't make a difference" sometimes *do*. Speaker cables, anyone?

    --Steve

  8. #8
    Senior Member Themis's Avatar
    Join Date
    Dec 2007
    Location
    Montpellier, France
    Posts
    920
    In this particular case, a simple test between the two interfaces should show whether there is a difference or not.
    In case of a difference, then we can use various tests to determine whether the difference is audible.
    In any case, there's nothing "unknown" here: if there are differences, they are easy to explain, and in NO case could they come from the source's method (CD or network).

    (i mean, we can still discuss about something else, if you wish, of course )
    SBT - North Star dac 192 - Croft 25Pre and Series 7 power - Sonus Faber Grand Piano Domus

  9. #9
    Senior Member Phil Leigh's Avatar
    Join Date
    Apr 2005
    Location
    Cornwall, England
    Posts
    9,991
    Quote Originally Posted by bhr1439 View Post
    I had read claims in Stereophile and (I think) The Absolute Sound that music sourced from a hard drive sounds better than music sourced off a CD player. I was skeptical of these claims, and am skeptical of my own ears, which seems to confirm the fact.

    As soon as I started listening to music served by my new SB Duet system, I detected a certain more lively and engaging aspect to the music. I did an A/B comparison to assure myself levels are matched (they are), and on close A/B listening, I cannot identify any difference. Yet, the fact remains that my impression of music played back via SB is more favorable, and this is not a subtle impression, it is a definite emotional "wow".

    Here are the two different music paths:

    path A: Nakamichi MB-10 CD player digital out --> Tact 2.1S digital processor --> Mark Levinson 360S DAC

    path B: Netgear ReadyNAS Duo --> SB Receiver digital output --> Tact 2.1S digital processor --> Mark Levinson 360S DAC

    I cannot pin down any good reason why path B, with the Squeezebox, should sound better. Both paths use the same, very high quality DAC. Up to there, bits is bits, one would think.

    The audio mags claim music servers sound better because lack of jitter on hard drive playback; that doesn't make sense for two reasons. First the Levinson DAC has sophisticated dejitter circuitry, but more fundamentally, the problem with jitter arises with the nature of clock recovery from the S-PDIF waveform content. Both path A and path B use S/PDIF digital signal to transfer the music data.

    The other potential reason is the effect of error correction on CD playback. This would imply that the cheapie CD-ROM drive in my laptop computer does a better job of reading music data (i.e., lower raw error rate) than my CD player, giving rise to fewer uncorrectable errors (uncorrectable errors cause the software to "fake it" by throwing the data out and interpolating across the gap).

    Anybody know of comparisons in raw error rate in playback on CD player vs playback on a computer CD-ROM drive?

    --Steve
    Steve,

    When a normal redbook CD drive reads a disc it gets "one bite at the cherry" for each sector as the disc spins (yes I know there are a few that go beyond this but they are exceptional), whereas EAC or whatever take as many goes as they need to get the right bits from the disc without having to resort to any form of interpolation. However, I don't think this is the big difference...it is however what passes as the "science" behind the many treatments (cryogenic treatment of discs, mats, green pens etc) that claim to improve the readability of discs. This is despite the fact that a CD-ROM mech can reliably read 99.999% of discs without error! If these treatments are doing anything, I don't believe it is to do with "the bits" - I think it is that they may make it easier for the drive to read "the bits". Easier in this context means that the transport has to work less hard, the various servos and motors do less work and therefore...


    The above can easily be proven. None of the available commercial transport/disk treatments produce any difference in the bits ripped by a CD/DVD ROM mech via EAC etc. On the other hand, many people claim to hear improvements in replay quality from such treatments - why?


    I believe that the big difference is that the transport mechs of CD players inject electrical noise into the circuitry that then finds its way into the SPDIF stream or internal DAC and hinders the ability of the DAC to recover the clock with total accuracy (aka "transport-induced jitter"). This is a non-issue when ripping to hard disk as there is no D/A conversion, no clock and hence no TIJ.

    In summary:
    1) looking at recovered bits is relatively pointless - the bits should always be perfect, either because they have been ripped with 100% accuracy or because the CD player is reading them in real-time with equivalent accuracy. The reality is that ripping has a slight advantage here in that the bits will ALWAYS be the same when read from the hard disk, whereas each time a CD player reads the disk it MAY get slightly different results.

    For hard disk replay, assuming an accurate rip, we should stop thinking about "the bits" in the same way that we don't worry about what our computers are doing with their "bits" - if something is wrong we will find out pretty quickly!

    2) Differences in transport-related audio replay quality are not primarily connected to the bit-recovery process, but to the D/A process. With an external DAC connected via SPDIF, the SPDIF mechanism itself is susceptible to noise that interferes with the ability of the DAC to do its job. We can look at reducing this noise at source, or at making the DAC immune to the noise... or both.


    3) Comparing hard disk replay to CD replay is in some ways about as far from comparing "eggs with eggs" as you can get. Up to the point where the SPDIF stream is sent (externally) or the bitstream is presented to the DAC (internally) the HD replay process is entirely deterministic and repeatable - just like a computer is. After those two points we enter the analogue domain and all bets are off!. CD replay has the added (random) complications of possible real-time read errors/corrections and transport-induced jitter...

    The only conceivable way that TIJ could exist in an SB-type product would be if the ethernet/wi-fi interface and/or the CPU generated "noise" that affected the DAC processing or the SPDIF signal generation...

    This noise could be directly induced or indirectly introduced via modulation of the supply lines...

    Jitter on the SPDIF stream can be dealt with to some extent by the downstream DAC circuitry. However, it would be better if this jitter was not there inthe first place. Unfortunately, SPDIF is inherently compromised in this respect and the termination/cabling etc means thats complete elimination of jitter entering an external DAC is almost certainly not possible.

    What we need is a better digital transport interface that doesn't embed the clock...

    If all transports were fitted with a clock-in and all DACS with a clock-out life would be so much easier!
    You want to see the signal path BEFORE it gets onto a CD/vinyl...it ain't what you'd call minimal...
    Touch(wired/W7)+Teddy Pardo PSU - Audiolense 3.3/2.0+INGUZ DRC - MF M1 DAC - Linn 5103 - full Aktiv 5.1 system (6x LK140's, ESPEK/TRIKAN/KATAN/SEIZMIK 10.5), Pekin Tuner, Townsend Supertweeters,VdH Toslink,Kimber 8TC Speaker & Chord Signature Plus Interconnect cables
    Stax4070+SRM7/II phones
    Kitchen Boom, Outdoors: SB Radio, Harmony One remote for everything.

  10. #10
    Senior Member
    Join Date
    Nov 2008
    Posts
    310
    But that explanation can't account for the differences heard by the O/P since both signal paths between which he hears a difference are sourced from an SPDIF stream going into the Tact processor and then the DAC.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •