USB to SPDIF interfaces - how to evaluate?

I've been reading up on USB/SPDIF interfaces and having a hard time determining what are the right measurements to evaluate and what audibility threshold exists for these measurements.

The linked site from Audiophilleo has some interesting comparisons:

The comparison highlights Phase jitter (in two different frequency bands) and period jitter. Should I be worried about both? Is period jitter below 5 picsoseconds (ps) really audible? Are there other specs that are even more important to the performance of the USB/SPDIF interface?

Why would additional power improve the performance of these interfaces, which is something that is observed regularly in listening evaluations?

Finally, is the price of this jitter reduction about to decline to a very low level as software and hardware experiences the performance improvements regularly seen in the IT industry? (So does this type of equipment fall outside of standard audio equipment economics?)
I just came across this thread, interesting no one has commented yet. In general, I have found in my research the following things:
try to use a converter that utilizes two clocks, one for each "family" of sampling frequencies, 44.1kHz and 48kHz. This avoids "clock synthesis," which can cause jitter.
galvanic isolation from the computer's power source is good, this is why so many converters are battery driven or come with wall warts.
There are several techniques to measure jitter, some of which actually cause jitter, so going by specs is risky. There are also several different types of jitter, as you saw on the audiophilleo site, and these can be measured at different points along the signal path, producing different results. Jitter measurements posted by manufacturers can be misleading.
Jitter sounds like high frequency noise, so I've read. I use a Vlink into a PS Audio DL III, and I think the music sounds too bright, so I'm either going to upgrade the speakers or the converter. Haven't decided.
Looking at your system I found reason,why your system sounds bright-Monitor audio bx speakers.Some years ago I had MA rs6 wich sounded bright too.They need some warm sounding amp in order get balanced sound from these speakers.
V-link doesn't sound bright.It has nice separated highs,a little bit extented,a that's what I hear with Dynaudio and Harbeth speakers.
People evaluating audio gear usually by listening to it and expectations which will confirmed or not.This is influenced by personal experience.For expample many people raving abaut vlink interfaces.I am not,because I like masive ,mascular midrange which vibranting into the room from speakers.And v-link doesn't do that,do nice highs and bass,separating everything fine,but midrange is too smooth and too thin for me.Stereophile noticed that ,but they did it in very diplomatic way:)
Actually I don't care abaut jitter if the system sounds good to me
Yes Extravaganza, I am looking at speakers with soft dome tweets. These BX5s were intended to be temporary, and will be relegated to 2nd system in the bedroom eventually. I was coming from B&W 685s, and noticed a distinct change in the mids. At the time, I thought this was a change from the B&W "house sound" to a more neutral BX5 sound, but after upgrading my head unit, upstream weaknesses are being revealed.
Even though jitter is the #1 problem with digital audio, using specs to determine the best component is fraught with problems, particularly jitter specs. These measurements are not standardized, can be measured several different ways and there is still no correlation to audibility, not to mention the cheating and outright lying in the measurements.

You are much better off to read the reviews that are shootouts to narrow it down, and then purchase one to actually hear the effect in your system. Most of them have money-back guarantees. You can also read the major DAC reviewers reviews to see what converter they use to compare DACs with.

Steve N.
Empirical Audio