Optical Thunderbolt 3 Cable

1. I don't disagree on the price being ridiculous, but that's what optical TB3 cables seem to go for. If you need to eliminate the ground loop via a non-metallic means of connection, optical seems to be the only way. What are you gonna do? I don't see any other option.

On a positive note, this would afford me a 33 foot cable versus the 6 foot cable I currently have. A 60 foot cable is only another $20 or so. The conversion to and from optical must be what drives up the price, not the length. I would honestly prefer a 6 or 8 foot optical TB for a cheaper price, but these optical cables seem to be marketed to people looking for longer lengths which are apparently not possible with copper cables, and not to people who are looking to eliminate ground loops, even if elimination of a metallic connection is an ancillary benefit, and the primary reason I became interested in the first place.

That is unless you know of cheaper optical TB options or some other means of eliminating the ground loop through my copper TB cable? Or are the cables you use at work copper and not optical?

2. Data buffering and things of that sort are the kinds of things I was wondering about. I have no idea if these optical cables introduce additional latency due to that sort of thing.

Sadly, data connections between a computer and external audio interface (whether it be USB, Firewire or Thunderbolt) are quite often responsible for completing a ground loop, and it's always reassuring to track down a cable that completes the ground loop and causes problems.

However, you don't necessarily have to break the loop at that point - as long as the loop is broken the problems should still go away. Sometimes it can be significantly cheaper to break the loop(s) via balanced audio connections where available, or failing that, to introduce transformer isolators in your audio chain, such as this one:

www.soundonsound.com/reviews/art-dti

..or the ART Cleanbox II artproaudio.com/product/cleanbox-ii-hum-eliminator/

All of my audio interconnections already have the shields on the input end lifted. Also, everything in my racks is isolated from one another, so there's no path for ground loops there either. I have no idea about the specificities of shielding on my TB cable. I just know that, when I unplug it, the ground loop stops.

Unfortunately, it just comes down to the TB connection. So I'm really just focused on trying to determine if there are any pitfalls with optical TB cable, such as added latency or things of that nature.

I don't disagree. I just wanted to make sure I won't be trading one problem (ground loop) for another (latency, etc.).

Based on the reply from Svart, it sounds like the answer is that it shouldn't add latency, though he did add a potential caveat which still makes me hesitant without gaining more info.

There will be no perceptible latency. The manufacturer says "low latency", but you need to understand the difference between actual latency and latency that can be perceived.

Now, I don't know these cables, but but my education is in electrical engineering, so I'm making pretty reasonable assumptions:

The conversion to the optical signal will have some minuscule latency, then the "light" might also have addition rise time, and the detector likely has a threshold—it something is rising "slowly", and the thing watching it has a threshold, there will be latency for the detector to determine a rise. But how much? We don't know, but consider this: the digital signal being transported over the cable is changing at a rate FAR faster that audio samples change. After all, that thunderbolt cable can send hundred of audio channels at the same time, each of which is 24-bit (or whatever), so the bit rate of transceivers is extremely high compared with the sample rate. If that latency is slower than a fraction of the bit rate, this whole concept will not work. (And think about it—a FIFO is not going to help, since the cables are supposed to work at a continuous rate.)

So, "low latency" is only meaningful for digital transfers. You might as well say "no latency" when talking about audio.

Those are good points. I had already suspected that latency was a non-issue, but I figured it was still worth asking about, since the optical cables do involve additional things in the signal path beyond that found in a copper cable. In any case, your explanation makes a lot of sense and that's probably good enough for me to not worry about it anymore.

In that case, apologies for my previous post suggesting audio isolating transformers - if you've got shields lifted at the input end of your balanced audio cables and rack metalwork from each unit isolated then only the Optical TB cable will do the trick. Good luck Quint!

Can't borrow a lesser priced cable?
Cheers, Ross