192kHz/24bit vs. 96kHz/24bit "debate"- Interesting revelation

Discussion in 'Audio Hardware' started by mindblanking, May 10, 2013.

Thread Status:
Not open for further replies.
  1. mindblanking

    mindblanking The Bourbon King Thread Starter

    Location:
    Baltimore, MD
    I've read quite a few posts here where some are debating whether or not the human ear can really notice the difference between the two sampling rates and I've been purchasing the 192 stuff from HDTracks because I assumed that, for the extra 5 bucks, why not? Today I realized something. Was going through the manual for my Marantz DV9500 and found out that even though there is an indicator light for 192, the player actually "downsamples" to 96. That was frustrating as I realized that I've not been getting the full benefit of either the DVD Audios I've purchased or the ones I've burned using HDtracks files and HD Audio-Solo Ultra. Have loved what I've heard but, still, felt a little duped. Then I cheered myself up by saying "well, hey, maybe those guys who don't think you can hear the difference are right. So don't worry about it." But then I remembered that a couple of months ago I picked up a Meridian Explorer USB DAC to test out. It only costs $300, plays 192/24 bit stuff without downsampling and the guy at the store said it sounds like a $1000 DAC and would beat my Marantz. So I took it home and tried it out. Spent nearly a full week with it and decided to return it because, while it was certainly impressive, my Marantz beat it in every way except one...

    Stuff I had downloaded from HDTracks DID sound better when played directly through the Meridian than it did when burned to disc as a DVD A and played through my Marantz. Not much better, but definitely better. Was true on every file/disc I tested. I even mentioned that to the sales guy when I returned it. My assumption at the time was that maybe it had to do with the burning process. But that didn't really make sense because other discs I've burned (gold cds where I want an extra copy for instance) sound NO different to my ears than the original discs. Therefore it couldn't be the burn process. I think the difference is the fact that my Marantz does not output 192/24 and the $300 Meridian Explorer does. This is, of course, my opinion and is in no way, shape or form meant to be a scientific analysis. Feel free to ask questions but the reason I think my conclusions are valid is because I wasn't even "testing" for this. I assumed both DACs could output 192/24 and, yet, in every instance the one DAC that actually could sounded better.
     
    ianfaith1 likes this.
  2. c-eling

    c-eling Forum Resident

    Thanks for your observations Mindblanking, I have not bought any of the higher than 24/96 yet, Your dac might be coloring the music, what would be neat is if you tried a 24/96 of the same song using the Meridian, I enjoy all formats, vinyl, cd etc. I see HDtracks has Some Yes available, I might jump on one of the 24/192, so far I've been satisfied with a couple 24/96 from them :)
     
  3. Thurenity

    Thurenity Listening to some tunes

    I'm happy with 24/96 max.

    .....until 2017 when I'm sure I'll change my mind. o_O
     
  4. c-eling

    c-eling Forum Resident

    lol, My hearing will be gone by then
     
  5. Given the choice, I always opt for 24/192, unless the original source is a lower resolution digital file and has been upsampled to 24/192. Yeech! In that case, I just go for the original, native resolution file.

    I'm too much of a wuss to purchase the 24/96 and 24/192 versions of the file and A/B them. I'll leave that for the crazies with more of an acumen for spending than sense! :laugh:
     
    IanL and Doug Sclar like this.
  6. Danglerb

    Danglerb Well-Known Member

    Location:
    Orange, CA, USA
    Things would be so simple if "wasn't even "testing" for this" had any effect on bias.

    Of course what would be double funny is if the HD tracks originated from a lower bit or sample rate master. The 24 bit part is moot since no DAC has more than about 20 bit accuracy.
     
    Spek likes this.
  7. TarnishedEars

    TarnishedEars Forum Resident

    Location:
    Seattle, WA
    I'd be willing to bet that the majority of the difference that you heard was because the 192k stuff was being played off of a HD, and the 96K stuff was being played from your transport. To do a fair comparison, both need to be played from the same source. I know that this is still a controversial statement in many circles, but never discount the impact of jitter. Even the very best disk spinners are inferior to a good HD driven system in this regard.

    Try downsampling some of your 192K files to 96k and see if you can hear a difference when they are played from the same source. I'm betting that most of the difference will be gone if you try this. In my experience it is extremely difficult to tell these two formats apart.
     
    Bolero, Albireo and Spek like this.
  8. utenteanonimo64

    utenteanonimo64 Well-Known Member

    I have the Neil Young archive BluRay discs. They have the same audio tracks in 192/24 and 96/24. I do not hear any difference. I have also the Neil Young DVD-As with audio tracks in 176.4/24 and 88.2/24. I do not hear any difference.
     
    Spek likes this.
  9. mindblanking

    mindblanking The Bourbon King Thread Starter

    Location:
    Baltimore, MD
    I think the bias effect does come into play here to some degree. I wasn't looking to prove something I already expect to be true- that 192 DOES sound better. It was a listening test strictly about the two devices.
     
  10. mindblanking

    mindblanking The Bourbon King Thread Starter

    Location:
    Baltimore, MD
    I'd like to try this and, if I'm right, it's easy to do with HD audio solo simply by changing the output sampling rate to 96/24 and then just burning, right?
     
  11. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
    That can be influenced by many factors: stereo playback, DAC performance at each resolution, and critical listening skills.
     
    Vidiot and mindblanking like this.
  12. mindblanking

    mindblanking The Bourbon King Thread Starter

    Location:
    Baltimore, MD
    And to your first point, yes that was my original thought. However I really don't hear a difference between CDs I rip using db poweramp, then burned using hd ultra vs the original disc. There may be a difference but I can't hear it. I definitely did hear the difference I talk about above.
     
  13. utenteanonimo64

    utenteanonimo64 Well-Known Member

    Of course, my system is poor and my critical listening skills too. On the other hand I haven't seen any hard factual evidence that anyone (even with highly developed critical listening skills) can distinguish between 192/24 and 96/24.
     
    Vidiot, Dan C and Spek like this.
  14. Spek

    Spek Well-Known Member

    Location:
    DFW, TX
    I think this may be responsible for a lot of the differences people hear. Try upsampling the 24/96 to 24/192 and then you can actually compare with all things being equal (since the converter should perform the same on both).

    As I've stated before, I've seen most theories supporting 24/96 have to do with avoiding the 44.1 kHz brickwall filter. At 24/96 this isn't remotely an issue to begin with, so the only thing added is high frequency content above 48 kHz. I'd love to know how this affects what people hear.
     
  15. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
    This represents a fundamental misunderstanding of how hirez audio works.
     
  16. utenteanonimo64

    utenteanonimo64 Well-Known Member

    Why?
     
  17. Danglerb

    Danglerb Well-Known Member

    Location:
    Orange, CA, USA
    Is it your contention that a higher sampling rate in any way changes the amount of information collected within the audio band? If so that represents a fundamental misunderstanding of how ADC/DAC works.
     
    QBNCGAR, drSeehas, Dan C and 2 others like this.
  18. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
    Friends, hirez benefits occur from 20hz to 20khz as well. Anyone who has experience listening to live events recorded 24/96 and 16/44 knows this. It's the solving of the pre-ringing and resulting imaging problems that Stuart discusses but its also the fact that ultrasonic information can impact the audible band. Here is a paper James Boyk wrote that discusses it.

    http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
     
  19. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
  20. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
  21. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
  22. Spek

    Spek Well-Known Member

    Location:
    DFW, TX
    Again, we're talking about 24/96 to 24/192. There are no "pre-ringing" or filtering issues with 24/96 that will be solved by 24/192. 24/96 is already well out of the audible band.

    Even your buddy Stuart says 24/96 and 24/192 are excessive. Please read the Wikipedia article on sampling because there is obviously a misunderstanding of the fundamental principles of how it works. When going from 96k to 192k sampling, it only extends frequent range from 48 k to 96k. It's a fact.
     
    Vidiot and utenteanonimo64 like this.
  23. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
    Spek, I believe there are some improvements in terms of filtering at 24/192 because I hear better imaging. I record small classical and jazz ensembles in Atlanta at 24/176 but we started at 88.2 and worked our way up because we kept getting closer to the sound of the live event.
     
  24. LeeS

    LeeS Blue Note Fan

    Location:
    Atlanta
    I would add we tried this across a variety of ADCs and DACs.
     
  25. Spek

    Spek Well-Known Member

    Location:
    DFW, TX
    I would say something was wrong with the filtering if it couldn't be achieved transparently at even 96 kHz. Someone like Black Elk can fill in much better details than I (if he even bothers to read the hardware section anymore), but filtering with modern designs is hardly an issue anymore at 44.1 kHz with oversampling, or at least not the issue a lot of audiophiles make it into. It was a problem back in the 80s with the earliest converter designs but I think much better ones are being used now in general.

    What I can't imagine is why the filter would be a problem at 88 or 96 kHz ... that awful brickwall filter that mangles and destroys music at 44.1 kHz isn't needed at double that rate, is it? A *much* simpler, gentler, and easier to build filter can be used at that rate. Doesn't Stuart even say that you don't even need 96 kHz to achieve a proper filter design? That's how he arrived at 58 kHz. To quote him, "Sampling at 88.2kHz or 96kHz is too high, and therefore wasteful of data," and, "The use of sampling rates above 96kHz to convey a wider audio bandwidth cannot currently be justified."

    (Notice how it's understood from his last sentence that the purpose of higher sampling rates would be to "convey a wider audio bandwidth." Stuart knows that, provided satisfactory filter design, this is the one and only effect of increasing the sampling rate. )
     
Thread Status:
Not open for further replies.

Share This Page