Note that Firewire appears to share the same data-related jitter problem of S/PDIF and AES/EBU. Like them, Firewire multiplexes the clock signals in with the data, which is how the largest form of jitter is created in the first place. A much better solution is to use a separate line for an external clock, as is done in most professional video and digital audio gear. For example, the SDIF (not the same as S/PDIF) digital audio interface format which is used between pieces of pro-audio equipment has a separate clock line.
Solutions based on Philips' I2S (Inter IC Sound) bus have been put forth with varying urgency by different groups. I2S was originally intended to be used between digital audio chips, for example inside a CD player:
Audio hobbyists were perhaps the first implementors of I2S between chassis, taking its data and clock lines from Philips digital chips inside CD players, buffering them and feeding them out to external D/A converters.
Now-defunct Audio Alchemy used I2S between their CD players, D/A converters and jitter boxes on a DIN connector with a pinout of their own choosing. A few other small manufacturers picked up on AA's I2S scheme making it somewhat of a standard.
UltraAnalog, Sonic Frontiers, Kevin Halverson of Muse and others devised a more complex but robust version of I2S which they call I2S enhanced or I2Se. I2S enhanced allows the master clock to be in either the DAC (best performance) or transport (next best) and also includes logic to choose the best scheme automatically based on signals present on the bus.
I2S enhanced uses the same 13W3 connector used on computer workstation monitors. It includes 3 coaxial channels (used for red, green, blue on computer monitors) and several twisted pairs. The coax is used for clock signals and the data and other control signals run on their own twisted pairs. One nice thing about this standard is that it's a properly designed interface, unlike either S/PDIF or AES which used physically compromised connectors. The better design of I2S enhanced makes it much less sensitive to cabling differences than other standards. The spec also includes well-designed input and output buffering.
I2S enhanced is currently in limbo. Sonic Frontiers and UltraAnalog, makers of high performance DAC and Interface modules, have a nicely defined specification for I2S enhanced, but Kevin Halverson of Muse has split off and is moving in slightly different (and unfortunately incompatible) directions with it. For example Halverson is working with DVD-Audio planners to include encryption of data on his version of the interface, the idea being to make piracy of high resolution content it carries less likely. This is foolish as piracy is not a sufficient problem to stall and fragment technical progress.
As it stands now I2S enhanced is essentially a proprietary interface format used between a given manufacturer's digital audio products. Since most manufacturers are implementing their own version of I2S enhanced, they're preventing exactly the kind of plug and play compatibility (i.e. standards) they designed it for in the first place. They could learn a thing or two from computer networking (such as the Internet) where openly engineered and freely adopted standards are a large reason for their success.
Probably the biggest problem for I2S enhanced is the lack of commercial adoption. High performance audio simply doesn't have the same strength of market forces to drive it as the computer or mass market consumer video fields. Another major factor contributing to the lack of adoption is the general uncertainty over future digital audio formats. If Firewire, which may be inferior for jitter, is highly likely to be the next interface format, why bother with I2S enhanced? The quick answer is because it may sound better.
From http://www.jeffchan.com/audio/