Jack Hammer is right. One of the biggest myths about digital connections is the "all or nothing" theory. This is simply untrue. The protocol even may be designed in such a way that a miss bit could cause a major problem, promoting the "nothing" at the other end, but that doesn't happen at the transmission level, and does not occur in most digital protocols, anyway.
I'll give you an example that happened to me. Had two dvd players, both connected to a denon avr 3805, both via optical toslink. Both cables exactly the same (model, brand, length), running the same path, similar bending. Now, inserted an HDCD (van halen-1984) on one, the denon detected HDCD (showed on the screen). Inserted on the other, the denon didn't detected it as HDCD. In fact, it detected it as a regular "PCM digital" and started playing it normally (without hdcd decoding however).
In this case we're not talking about the cables "per se" (they were equal), but about the lossy characteristics of digital connections. The first player had a better electrical-to-optical circuit, a stronger led, or something else, that made the transmission to be less faulty than the second. On the second we all have to agree that bits were loss during the transmission, however not enough to cause a total failure, as the denon still played the pcm stream it was getting.