[Gclef]

Okay, so I just purchased a new portable amp (iBasso D6) for my new headphones, and I was reading about the D6 having a very good DAC - capable of 24/96.  Cool.  But I read an article that said the human ear can't distinguish the difference between 16 bit and 24 bit, nor the difference between 44 hz and 96 hz.  Is that true?  If so, what's the point in going beyond 16/44?
 
I rip my music to Apple lossless, which is 16/44.  So, to make a long story short, I'm curious to know if there's a higher quality digital music source that I'm missing out on.
 
I'm a noob with a general understanding of things, or so I thought, but now I'm not so sure?  Can I get a little help here?  I have to be misunderstanding or misinterpreting something (?)

 

[obobskivich]

If you're ripping CDs - 16/44.1 is exactly what should be going on (since that's whats on the CD). DVDs can get into the 24/48 realm, and DVD-Audio (where the entire DVD is used for just audio) into the 24/96 realm (DVD-Audio playback can get tricky on computers though). Blu-ray can go somewhat higher, it's big advantage is lossless encoding. We'll ignore that SACD exists because to my knowledge, it isn't available for PCs (and that iBasso can't do DSD anyways).
 
So, as far as what you can and can't tell apart - 24-bit DACs do and don't exist; devices can process that level of fidelity but in terms of reproducing it "accurately," there isn't anything with minute enough voltage swings (144 dB). It's mostly marketing unless you're dealing with DVD or DVD-A though (and then, it's only because you want to be able to decode the audio, not because it's "better").
 
Sampling rates are affected by the Nyquist limit, I'll just link the wiki:
http://en.wikipedia.org/wiki/Nyquist_limit
 
So basically, 96khz allows us to record higher frequencies - the question then becomes: how's your hearing in the 30-40khz range? Can your speakers, amplifier, etc do that? 
 
I am somewhat oversimplifying. 
 
Summary: higher numbers translate into higher theoretical accuracy, because more data points are generated for conversion to analog. Whether or not this accuracy matters is up for debate (but generally I would say that I don't believe it does).
 
If we're talking about SRC (sample-rate conversion), I'd say: don't. There's no good reason to convert something from 16/44.1 to 24/48 or similar - you can't add data that doesn't exist, and you'll likely introduce more distortion artifacts as the data you already have is stretched and modified to meet the new SR. The only reason you'd want to do SRC (aside from cases where it's forced/not optional, which happens for DSP processing in some cases) is if you're working with a multimedia project and need everything running "the same" for the final output (say if you're making a cartoon and drawing audio from multiple sources).
 
 

 

[Currawong]

Very roughly, higher-resolution files will result in a lower noise floor in the DAC. However, any benefit would be lost on such a product IMO, unless they are making battery-powered DACs that have a seriously low noise floor already. There is arguably some benefit with high-end gear though, possibly more so with recordings that have a high dynamic range or low volume.

 

[TwoTrack]

In my experience, Nyquist while technically correct has lots of implementation issues:

1.  Overtones exist from higher frequencies that are audible in the 20hz-20khz range.  James Boyk at Stanford has done some good work on this.

2.  Nyquist is not perfect for transients in my opinion.

3.  My own listening suggests that higher resolution does have more benefit in terms of transients, dynamic range, and soundstage width and depth.

4.  Some DACs work better at 16/44, some at 24/92 and some at 24/192.  In my experience the improvement in sound from 16/44 to 24/96 is substantial, the improvement from 24/96 to 24/192 or 24/176 less so but still a gap.

5.  Very little music is available above 24/96 so it may not be hugely important to do 24/192.  But the DVD-Audios and Blurays at that rate sound fantastic.

6.  I prefer DSD over PCM.  The music just seems to be more continuous and flow better like a good analog rig.

7.  Computer audio is funny.  Weird stuff happens all the time including digital cables sounding vastly different, music players on my Mac Mini sounding better than "bit perfect" iTunes, and power cords impacting sound quality for better or worse.
 
8.  Jitter reduction is really important.  We can hear it down to the single digit picosecond range.

 
I'm basing this on many recordings I have worked on and my own listening tests using pretty good gear and good sources including digital master files from a Sound Devices 722.

 

[TwoTrack]

Here is a test for you.
 
Get a copy of a redbook (16/44) CD mastering from a good audiophile label like Reference or Chesky and compare it with the corresponding DVD-Audio version.  You can get the DVD-Audio discs from retailers or download the Reference "HRX" hirez PCM file or Chesky 24/96 FLAC file (from hdtracks.com) to your computer.
 
Using the same title from these labels will largely or completely eliminate any mastering differences and you can decide if the difference is big enough for you.
 
Just make sure you keep in mind two things:
 
1.  The resolution of your system will make a difference.
 
2.  Related to that, the quality of your DAC or soundcard will make a big difference.

 

[Head Injury]

Quote:

Here is a test for you.

Here is a test for you.
 
Take one of your 24/96 files and downsample it to 16/44.1 This ensures that it's the same mastering, and the only difference is the downsampling. Then throw both files into Foobar with the ABX Comparator plugin. Do 20 trials and post your results.