I don't know a lot about sampling rates and have been unable to work out what to me is a puzzle.

I understand that the standard for audio CDs is the Red Book and it calls for 16 bit/44 kHz sample rates to convert analog to digital. I interpret this to mean that the information is recorded to the CD based on the datastream arising from that sample rate and size.

Does a 24 bit/192 kHz DAC just convert that stream to analog at the 16bit/44kHz rate or does it use some algorithm to upsample it to the higher rate and then convert the limited datastream to a more refined analog sound?

Or do I have it muddled?

Upsampling DACs use an algorithm, which places zeroes in between the real data. Here's a good primer:

http://www.audioholics.com/techtips/specsformats/upsamplingvsoversampling1.php

Upsampling DACs do, but mostly, they have the ability to play at lower rates and resolutions. I don't know if they just use the "bottom" 16 bits, the "top" 16 bits, or something else, but basically it's just getting the 24/192 DAC to decode the 16/44.1 as if it were a 16/44.1 DAC. Most DACs are able to do that.

Thanks for that link, though I find the use of terminology I haven't quite got a grip on makes me unsure I'm really understanding it.

As I interpret it (and DC's comment), the 16bit/44k sampling reduces the analog signal to a datastream of those particular dimensions, much like a digital camera reduces an "analog" scene to a dataset of colour depth and resolution. Just as that scene can be captured at low depth/resolution, the analog sound stream is captured at 16/44. Enhancing that low resolution dataset to a higher resolution set requires interpolation of the intermediate data points since they don't actually exist in the low resolution data stream.

That either means the higher resoution DAC either reduces it's resolution to that of the datastream or it "enhances" the datastream to it's native resolution, a function I assume is carried out by the filters. That would suggest the 24bit/192kHz DAC in my Toshiba sounds better than the 16bit/44kHz in my Phillips unit more as a result of the filters associated with the DAC than any superior use the higher resolution DAC is able to make of the datastream.

Have I got that right?

Have I got that right?

Sort of, yes. The higher rate enables gentler filtering, which sounds better.

Thanks!

That helps.