Quote:
Originally Posted by ZenErik /img/forum/go_quote.gif
Do you know if the OMZ upsamples or oversamples at all? I'm just curious.
|
Two different terms. "Oversampling" is a technique of converting digital to analog in small chunks. That's a "feature" of any delta-sigma DAC chip, which is 99% of the market today. It has (usually negative) implications on sound quality, but these are very subtle.
The term "upsampling" refers to a technique which translates the incoming samples to a higher sample rate and/or bit per sample. Sometime you'll find the term "re-clocking" in that context.
There is no free lunch. The data contained in the incoming stream is all you have, so do not fool yourself to think that merely upsampling 44.1/16 to 96/24 would "improve the resolution". There is no more data in the upsampled stream than what the unit received from the transport.
There are advantages to upsampling. Moving the sample frequency higher means that the digital noise after conversion is shifted up in frequency and is easier to filter out with less impact on the treble.
On the down side, computing "new" samples inevitably adds error, although quite minimal.
But the real advantage of any sample rate conversion is in the potential improvement of the bit and word clocks, which could result in less jitter. The implications of jitter range from subtle so very obvious, depending on the unit in question.
If you were thinking "well here is a real free lunch" - then don't!
It's easy to find a unit with sample rate conversion. Not that easy to find one with a decent clock...