Please Educate Me


If I can’t find the answer here, I won’t find it anywhere. 

Something I’ve wondered about for a long time: The whole world is digital. Some huge percentage of our lives consists of ones and zeros. 

And with the exception of hi-fi, I don’t know of a single instance in which all of this digitalia isn’t yes/no, black/white, it works or it doesn’t. No one says, “Man, Microsoft Word works great on this machine,” or “The reds in that copy of Grand Theft Auto are a tad bright.” The very nature of digital information precludes such questions. 

Not so when it comes to hi-fi. I’m extremely skeptical about much that goes on in high end audio but I’ve obviously heard the difference among digital sources. Just because something is on CD or 92/156 FLAC doesn’t mean that it’s going to sound the same on different players or streamers. 

Conceptually, logically, I don’t know why it doesn’t. I know about audiophile-type concerns like timing and flutter. But those don’t get to the underlying science of my question. 

I feel like I’m asking about ABCs but I was held back in kindergarten and the computerized world isn’t doing me any favors. Now, if you’ll excuse me, I have some work to do. I’ll be using Photoshop and I’ve got it dialed in just right. 
paul6001

Showing 1 response by erik_squires

I'm not even sure if there was a question here.

One of the biggest sources of variance over time have been:

- Clock quality
- Type of DAC
- Upsampling or oversampling and algorithms
- DAC output stage
- Noise sources, like power supply and digital ground loops

The quality of the clocks and jitter elimination circuits has vastly improved since about 2010.  For the most part it's almost a moot point.

DAC's and how they are configured matters quite a bit.  For instance, some DAC's use multiple parallel DAC's simultaneously per channel to achieve the lowest noise, distortion and highest output current capabilities.

Next, how about those output filters?  Up or oversampling can affect the response through the top octave.  Upsampling or asynchronous sample rate conversion rely on lots of math to interpolate between the original samples.  They don't come out with the same results. Older upsampling could clip the signal in the interpolated samples. Hopefully everyone is now aware of this.

Lastly is the output buffer. A preamp stage, even if it has no volume control, which ensures uniform output regardless of the downstream device.

And... none of this matters if you can't hear it.  Buy the cheapest device which looks good and you can't hear better from.