Optimal length(s) of a digital aes/bsu cable

Harley wrote in his book that digital cables are optimal at 1 to 1.5 meters.
YRMV

Being balanced doesn’t change anything. Your cable might not be optimal length, but these calculations depend on many factors, like output transition time of the source, signal propagation speed in the cable etc. Also, this cable might be a perfect characteristic impedance match in your system (no reflections at all). The best way is to borrow one and try. Your ears will be the best judge. Mismatched cable induces time jitter of digital signal, that translates as noise added to music (not present without signal/music). This noise will make sound harsher, brighter with possibly worse imaging.

Doesn't matter use the .6 cable short AES3 cables are used in pro studios all the time.

Back when I owned a CAL Delta CD transport and both their Alpha and Sigma DACs; I needed an AES/EBU cable.      Bought a Kimber Orchid in the .5M length, since they were to be parked right beside one another.      I couldn’t believe how bad the combo sounded (with either DAC).      Vaguely remembering something read about digital reflections in transmission lines, I tried the Orchid in a 1.5M length and found Nirvana.           My suggestion would be; audition two lengths of the same cable, for yourself.              https://www.audiosciencereview.com/forum/index.php?threads/technical-article-digital-audio-cable-ref...

yes, to be a bit more specific - 1.25 up to 2 m - so 1.5 m is a safe number

digital reflections can occur (at least in theory) at 1 m length

There’s a lot of good theory on this subject. In practice though, how cable length effects sound quality is difficult to hear. We manufacture cables and have tried various lengths of the same build. For our ENO Digital Cable Cu and Ag, two cables built identically, one at 1M and one at 1.5M sound identical. If you start talking about much longer distances I’m sure things change.

Thanks for the input fellas - all very helpful! BTW - the link from rodman9x is an excellent piece. 

It depends on the output transition time .  Without going into details, it is roughly transition time divided by 20.  For typical 25ns transition it has to be longer than 1.25m.  The problem is that transition time is unknown (and almost impossible to measure) and can vary from 5ns to 50ns.  The main concern is that refection from the end of the cable will distort transition at exactly threshold point.