what digital cables for my dac?

Feel free to try shorter lengths of digital cable. While some cables may sound better in longer lengths, the general rules about all cables still apply to digital, shorter is always less damaging to the signal being passed in the cable. I'd suggest using the most revealing digital cable you can get, and letting your analog gear/cables tweak your sound. I try to never give up resolution in order to gain pleasant tonality, and any tweaking I do, I do as late in the signal chain as possible.
JMO.

It seems that most of the analysis of digital cable length is based on theoretical levels of reflections in cables. These usually assume some horrible mismatch between transport and dac, or a digital cable improperly terminated. If we are talking about a properly designed and terminated cable, then other factors besides reflections come into play; such as bandwidth attenuation caused by inductance and capacitance values that increase with length.
If we're comparing apples to apples, quality cable in different lengths; then I suspect that much of the praise for longer lengths is due to personal sonic preferences.
I used to use the standard 1m lengths in my system, but then I discovered that my setup prefers .5m. It's more open, and less smeared and colored in the midrange and treble.
Maybe the best bet would be to buy some decent cable at various lengths, and see what your system 'likes' best? Then you could use that length in further listening tests with other models of cable.
Okay, so that's maybe not the most practical option, but what about this hobby is 'practical'? :)

Drubin
Sterevox. This is the designer who created the Illuminati digital cable, which was something of an industry standard for many years in the 90's. He's back and with a digital cable for well under $200. This is an easy decision.

Agreed, it's an outstanding cable for the money. I think Galen Carol audio has them at a decent price too. Try gcaudio.com.

Tobias
Jack Dotson, there doesn't seem to be very much at that site. The nav buttons don't load anything--no cables, no systems, no speakers... maybe it's my browser.

I think it's your browser or its' settings, it worked fine for me.
It's interesting that their digital cable is cryo'ed. I found cryoing of digital cables to make a very positive improvement as well.

Tobias, yes, I've tried 1.5 meters in digital cables. I think it has much to do with what cable you're using. I read a statement by Ray Kimber that their digital cables tend to have more upper-frequency emphasis in shorter lengths, which is my experience listening to a number of different cables. I found the shorter length cable to present a more realistic sound on percussion.
I think people should just trust their own ears, if 1.5 meters sounds better to some, and .5 better to others, who am I to say they're wrong? I suspect that the improved rise time in shorter lengths may offset any possible increases in reflectivity. But that's just speculation on my part.

Mrtennis, I agree, the xstream digital from ps audio was an overlooked gem. I suspect alot of people didn't care for it because of it's extremely revealing character.
The digi-cable from Discovery cable is another standout, but too inexpensive to be taken seriously by some :)

I know the use of a longer length seems counterintuitive to those of us who are used to "shorter is better". Shorter _is_ better, for analog cables.

Go to http://www.empiricalaudio.com. Scroll down to the link "Paper on S/PDIF cable length in PSOnline".

In a nutshell, this paper explains that signals are reflected internally in digital interconnects. Good DAC design can minimize these reflections but this is usually not done. In shorter cables, the reflected signal returns to the output very close to the rising edge of a wave, and confuses the DAC about the timing of the signal. The result is jitter. The longer length allows the reflection to return at a time when it is less likely to be confused with signal. The result is more coherent music.

(I apologize for any inaccuracies which may be found in the very brief summary above.)

You can also refer to Chris Ven Haus' Web site, where he specifies that 1.5 meters is optimum length for digital interconnects, and to UHF Magazine number 74. UHF tested different cable lengths and now do not recommend any length shorter than 1.5 meter.

UHF Magazine

I did my own tests and came to the same conclusion. I used 1-meter and 1.5-meter lengths of 75-ohm Apogee Wyde Eye (still a contender for best at its price point). We listened using my TEAC VRDS T1 and Apogee DA-1000E-20, and also a friend's Theta Universal and Kora Hermes. The results were unquestionably in favour of the longer length.

You could try a test yourself and see.

First of all, let me say that I'm not trying to prove anyone here wrong for believing that 1.5 meters is the 'best' length for digital cables, but I am trying to show that there certainly is not a consensus on the subject. According to Dan Lavry of Lavry Engineering, that Empirical Audio SPDIF article is "nonsense": http://recforums.prosoundweb.com/index.php/mv/msg/11678/0//0/

Tobias, sorry, I think you're over-thinking my use of the word 'belief'. In this case, it was used in it's broader sense to mean something a person accepts as truth, regardless of source or experience. You believe it because you've experienced it, which is the preferred way IMO. I also believe shorter to be better because I've experienced it. Which really leaves us nowhere, (as usual in this hobby).
I was mainly trying to rebutt the 'technical' support for the longer-is-better argument. Not because I need to be right, but because I want people to trust their ears on a case-by-case basis, and not defer to anyone's 'technical' expertise.
Also, your friend with the 1 meter digital cable could 'swing either way' as it were, and go up or down a 1/2 meter for greater satisfaction.
Who knows? ;)