16 bit to 24 bit.....What’s the other 8 bits?

Discussion in 'Audio Hardware' started by timztunz, Oct 15, 2017.

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

    timztunz Audioista Thread Starter

    Location:
    Texas
    I have ripped Redbook CD’s to 16/44.1 AIFF fed via USB from a Mac Mini to a Bricasti M1 DAC. I use Audirvana + for playback which reports that it’s sending out 24 bits. What are the extra 8 bits? Where did they come from?

    I’ll admit to being lazy and trying to avoid wading through a sea of information about this. I’m hoping for something of a simple explanation, if there is one. Thanks to all who chose to participate.
     
  2. Claude Benshaul

    Claude Benshaul Forum Resident

    Most likely it's 16bit padded with 0 to 24bit.
     
  3. TimB

    TimB Pop, Rock and Blues for me!

    Location:
    Colorado
    Just added bits. If they were actually used, it would lower the noise floor to about -140dB.. Red book 16 bit is about -95dB noise floor. Adding the extra bits will not do anything to the playback.
     
  4. Stone Turntable

    Stone Turntable Independent Head

    Location:
    New Mexico USA
    If you listen very carefully during the quiet passages you may hear something that sounds like this:



    It's called science! :idea:
     
    goodiesguy, Keith V, crispi and 7 others like this.
  5. timztunz

    timztunz Audioista Thread Starter

    Location:
    Texas
    Well that’s somewhat maddening.
     
    Leroyd and SandAndGlass like this.
  6. Thouston

    Thouston Forum Resident

    Location:
    Mattoon, IL
  7. MrRom92

    MrRom92 Forum Supermodel

    Location:
    Long Island, NY
    If you’re configured to send a bit perfect output then the 16 bits are just being padded with 8 0s. The chip is expecting 24 bit data, true 16 bit converters are a bit of a rarity these days, or even true 44.1khz converters.


    If your computer is performing any DSP whatsoever... those 8 extra bits will count and you’ll be glad you had them. But they won’t be 0
     
  8. basie-fan

    basie-fan Forum Resident

  9. stax o' wax

    stax o' wax Forum Resident

    Location:
    The West
    What did you find misleading?
     
  10. SandAndGlass

    SandAndGlass Twilight Forum Resident

    Bit depth equates to only one thing in the digital realm.

    Dynamic Range.

    From the softest passage to the loudest, which, as @TimB has noted, is a range of 95-dB.

    I don't think I have any music that remotely reaches this range.

    The 24-bit is commonly used in studio's to offer a wider capture range, then it is all reduced to the 16-bit Red book standard for CD release.

    Which means, as long as the original material's dynamic range is captured intact and is able to be accurately reproduced at 16-bit resolution, then having 16-bits or 24-bits, does not add any further quality to the recording, which is determined primarily by the sample rate.
     
  11. darkmass

    darkmass Forum Resident

    I suspect that's not the complete story...

    While digital audio samples are signed binary numbers (that is, can take on both positive and negative values), for clarity of discussion assume only positive values are permitted. Also, certainly in my experience, "full volume" for 16 bit audio and for 24 bit audio seems pretty much the same for both.


    Here's an arbitrary volume level sample in a 16 bit word, most significant bit on the left:
    0000 1010 0000 0000 xxxx xxxx

    Here's the exact same volume level in a 24 bit word:
    0000 1010 0000 0000 0000 0000


    Here's the next higher level for the 16 bit word:
    0000 1010 0000 0001 xxxx xxxx

    A 24 bit word can certainly take on that same volume level:
    0000 1010 0000 0001 0000 0000

    However, that's not the next higher volume level the 24 bit word can take on...this is:
    0000 1010 0000 0000 0000 0001

    And here's the next 24 bit step upward:
    0000 1010 0000 0000 0000 0010

    24 bit audio has more refined volume level increments than 16 bit audio.

    You might argue that the 24 bit volume steps are too small to hear, and that could be a valid point, but nevertheless there's more to a 24 bit word length than "Dynamic Range"...maybe it could be called "Dynamic Smoothness" or "Dynamic Resolution".
     
  12. TimB

    TimB Pop, Rock and Blues for me!

    Location:
    Colorado
    I would agree with the more refined volume level.
     
  13. Doug Sclar

    Doug Sclar Forum Legend

    Location:
    The OC
    Imagine a sound stage from left to right. When you mix, you can pan a sound anywhere in that sound stage. There are more positions available with 24 bits than with 16 bits.
     
  14. Higlander

    Higlander Well-Known Member

    Location:
    Florida, Central
    I think we all realize they are not only too small to hear, but Far too small to hear. It for sure is a valid point.
    The reason that only dynamic range is ever mentioned, is that a lot of those bits are creating the volume gradients or levels all at levels that are very low and usually even below the inherent noise of the source or the equipment.

    So yes, in theory your point is absolutely true, but in practice 16/44.1 has volume gradients, that are beyond what any human can discern.
    I never hear anyone say, "Hey turn that up 1/20th of a decibel!

    Normal human hearing at low levels can barely discern a one decibel change in loudness, let alone tiny fractions of decibels.
     
    Robert C likes this.
  15. Higlander

    Higlander Well-Known Member

    Location:
    Florida, Central
    Except again, 16 bits provides far more than can be discerned by human hearing.
    In theory you are correct, but if in practice it makes no difference, why not mention that?
     
    Tim Müller, anorak2 and basie-fan like this.
  16. SandAndGlass

    SandAndGlass Twilight Forum Resident

    With more bits, yes you could have more gradients, meaning finer volume adjustments.

    I would suppose, that would be a reason to up-sample something from 16-bits to 24-bits.

    I would doubt if that is why music is recorded in a studio in 24-bits.

    In the recording studio, the objective is to be able to have a greater dynamic range to capture all transients that may be present in the original live performance.

    This might better be addressed by a member who works in a digital recording studio.
     
  17. basie-fan

    basie-fan Forum Resident

    There are no "stair steps" in digital audio, either before or after conversion to analogue. By using this false graphic it appears as though hi-res digital offers greater resolution or accuracy than Redbook (16/44) yet it does not. See post #160 here with video:

    MQA bails on Rocky Mountain Audio Fest*

    See also my posts #118 and #50 in the same thread.
     
  18. SandAndGlass

    SandAndGlass Twilight Forum Resident

    That doesn't seem quite right to me. It looks like it would be the same volume level if we were counting from the left, which is different than what I remember.

    xxxx xxxx 0000 1010 0000 0000 16-bits

    0000 0000 0000 1010 0000 0000 24-bits

    This would represent the same volume level in either 16-bits or 24-bits.
     
  19. Higlander

    Higlander Well-Known Member

    Location:
    Florida, Central
    I have nothing against 24 bit, and in the studio maybe even it is essential.
    I just feel that the benefits in real world home listening have been so grossly exaggerated, that some find 16/44.1 lacking or not truly high fidelity.

    But in reality we are talking somewhat big differences, for sure,as some pointed out, but they are beyond the realm of human perception.

    Maybe similar to the difference between 4K TV and 8K TV.
    8K has far more data, but if they human eye can not perceive the difference, that data goes wasted in practice.
     
    Tim Müller and SandAndGlass like this.
  20. Higlander

    Higlander Well-Known Member

    Location:
    Florida, Central
    In all fairness though, there are "Gradients" in level, but extremely minute even with 16/44.1

    True about "Stair Steps", Sine waves can be represent completely with a couple Samples.
     
  21. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    I googled "Audirvana", so it's an audio player software. Apparently it's upsampling the original audio, that means it interpolates. There is no useful information in the added bits, it's a pointless exercise.
     
    Tim Müller and SandAndGlass like this.
  22. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    The whole argument about audible bit depths is pointless here. The original source is 16 bits as described by the OP. You cannot add information later that wasn't captured in the beginning. Meaning, the noise that is already contained in the recording because of its original bit depth cannot be removed at a later stage by adding random bits. It will still be 96 dB even after upsampling.
     
  23. ggergm

    ggergm another spring another baseball season

    Location:
    Minnesota
    Eight bits are a buck.
     
  24. Thouston

    Thouston Forum Resident

    Location:
    Mattoon, IL
    There is nothing misleading about this article.
     
  25. SandAndGlass

    SandAndGlass Twilight Forum Resident

    I have nothing against higher resolution, be it audio or my new 4k TV.

    I had an initial interest in hi-res because I never particularly cared for the sound of CD's. But after listening to CD's that have been well recorded and mastered, I realized that I didn't have any beef with the CD's Red book resolution or, for that matter, digital audio in general.

    I don't bother with hi-res or DSD, never listened to either of them and therefore have no opinions concerning them, one way or another.

    If they come up with new and better formats, that is great, but I'm still not going to go out and re-buy my collection.

    I now have a 4k TV, but still my Oppo BDP-93, which is compatible with Blu-Ray and my aging Emotive UMC-1 processor, which is not 4k compatible.

    Even though the Oppo is capable of playing SACD's, I choose to buy an inexpensive player, because I only use the audio as a CD transport. The only way I can play a SACD is through the less expensive DAC that is in the 93, so I don't see the point of me doing it.

    There are still so many records and CD's that were not recorded as well as they might or should be.

    I operate a small motel, so I live where I work and listen to $4/mo. paid Pandora, which I understand can stream as high as 192-kbps.

    The streaming box goes into one of the digital inputs on the Peachtree iNova, where it is decoded by an ESS Saber DAC, that was their flagship offering at that point in time. Material that was produced with audio quality in mind, sounds fine.
     
    LarryP and Catcher10 like this.
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine