Quote:
Originally Posted by jimshowalter 
Jitter is audible, if bad enough, although it takes surprisingly large amounts of jitter to show up in an A/B test. (Looking for that article, but haven't found it. Read it just a couple of weeks ago, and should have bookmarked it.)
To be clear, if a player sent a signal over HDMI or any other connection, and the receiving end depended on the player for its clock, then yes, they receiver would be at the mercy of the mastering studio, player, and cable.
But that's not what receivers do, because it would be stupid.
The only thing the D/A converter needs to know is the value of the original samples. It knows that the samples were supposed to be taken by an infinitely accurate clock at the recording rate (for example, 44.1 kHz). So it takes those samples and converts them to analog at the recording rate.
It doesn't matter if the samples arrive from the player separated by wildly varying time gaps. Those gaps are ignored. Only the values of the samples matter. That, and the recording rate.
If you think your way through this, you'll eventually conclude, correctly, that the only possible source of timing errors is when the recording is made, and when the samples are converted in the D/A converter at the end of the playback chain.
If a playback chain is susceptible to any other timing errors, its designer is an idiot.

Jitter is audible, if bad enough, although it takes surprisingly large amounts of jitter to show up in an A/B test. (Looking for that article, but haven't found it. Read it just a couple of weeks ago, and should have bookmarked it.)
To be clear, if a player sent a signal over HDMI or any other connection, and the receiving end depended on the player for its clock, then yes, they receiver would be at the mercy of the mastering studio, player, and cable.
But that's not what receivers do, because it would be stupid.
The only thing the D/A converter needs to know is the value of the original samples. It knows that the samples were supposed to be taken by an infinitely accurate clock at the recording rate (for example, 44.1 kHz). So it takes those samples and converts them to analog at the recording rate.
It doesn't matter if the samples arrive from the player separated by wildly varying time gaps. Those gaps are ignored. Only the values of the samples matter. That, and the recording rate.
If you think your way through this, you'll eventually conclude, correctly, that the only possible source of timing errors is when the recording is made, and when the samples are converted in the D/A converter at the end of the playback chain.
If a playback chain is susceptible to any other timing errors, its designer is an idiot.
It's not that simple. How you describe it is how I thought of it for a long time, but recently I've revisited jitter issues, simply because I became convinced I could hear it. Here's an article that explains how it's not enough just to capture the 'original' samples and re-clock.
http://www.madronadigital.com/Library/DigitalAudioJitter.html























)