i suggest you contact kevin halverson: [email protected] he has an unmatched ability, in my experience, to answer questions like yours in an erudite but understandable manner.
7 responses Add your response
First of all, a couple of corrections to minor inaccuracies in your post. First, both zero's and one's are represented within transitions (there are always transitions around the actual data). So a solid run of zero's (or one's) are not represented by a DC voltage. Jitter is caused by an inaccuracy in the generating clock, in the transport, among other things. If it is tied in any way to the amount of electricity and/or heat, etc. It can also be caused by a sufficiently corrupt digital signal (reflections, etc., cause inability to detect transitions accurately). The only way to fix this is to have extremely accurate and insensitive clock chips in both transport and DAC, have some sort of word clock control over data "collection" on the transport (not getting enough, faster; too much, slow down; etc.). I know of only several ways of doing this: the Wadia system; both DAC and transport have word-clock on them, and master them to one that is extremely accurate, either internally to one or the other, or externally; or "hack" a connection from the DAC back to the transport. I found a web page on this last one once, have it bookmarked at work, will try to remember to look it up on Monday.
Kacz, that is a hard one. Sometimes it makes the sound soft and mushy, sometimes it makes it harsh and edgy. But it always reduces the focus and resolution. I guess ther must be different forms of jitter which interact with DACs in different ways. The best way I can think to describe it is that it screws up the harmonics. I have never seen or heard anyone explain why, but it just seems that a data stream with jitter entering a DAC chip causes harmonic distortion in the output of that DAC chip.
Redkiwi is pretty much correct -- lost focus and soundstaging. It's pretty hard to describe, until you hear it go away. Also, the explanation I was talking about is at this address: http://www.lcaudio.com/dacdrev.htm -- this is a company that sells upgrade clock chips for CD players, but there is lots of good information on jitter on their site.
Ask your self a question. Does jittering of digital signal in time domain influence the characteristics of the analog output of your DAC? which is frequency and phase critical for the localization of sounds reproduction. The answer is yes. The stereo signals store on the CD is multiplexed into a single stream of digital signals. The error in demultiplexing the stereo in time domain (timing jitter) will introduce mismatch in phase (not to mention IM distortion) for the analog signals generate by the DAC. Most of the modern CD players are incorporated some forms of clock recovery circuit so the jittering problem I think does not exist anymore.