The truth about network audio inline filters

[Ben-M]

I've long been interested or intrigued by improving networked audio, but I'm not sure where I sit with "improved" or "engineered" audio ethernet cables. And including audio focused ethernet filters. I haven't bought any cables yet, like a Meicord or Audioquest, but I have bought a couple filters like the Acoustic Revive RLI-1, and then just recently the Acousense (now branding under Artistic Fidelity) GISO GB. 

 

The RLI-1, I never had much of an expectation for it to be any different than either a medical isolator at best, or just a nicer looking patch cable at worst. But the GISO GB, it's in a box, and boxes can hold parts. So I admit to having higher expectations for the GISO GB and always wondering what fancy, ingenious things were in that box and why it seemed to get more better reviews than these other devices that lots of people who'd never tested them said couldn't work. 

 

So, as soon as I received one, I opened it up and took pictures. Now below for all to see. I think I see some standard ethernet isolation transformers like you'd find inside anything (they may not be normal, as they can be custom made), then it also looks like the output ethernet socket is internally shileded, and finally there's some really sticky, black putty across the transformers for some reason. Besides that, there doesn't appear to be any other special parts in there or any "magic". 

 

I don't know enough about this stuff to say otherwise, and my rig isn't up yet for me to pass any personal listening feedback on, so I wanted to post these up for other people's feedback for now. 

 

Outside:

 

Inside:

[One and a half]

Ethernet speeds are quite high speed, read frequency. Each path is critical and distances mean a great deal. Ethernet suffers from common mode noise, and the transformers you see are common mode chokes, series types perhaps also with built in ferrite cores. Attenuation then depends on the magnetic properties of the chokes and the ferrites, without strangling the actual signal, so it's not a trivial task.

 

Like yourself, I'm intrigued by Ethernet and its routing of audio data, but the more I read about the noise propagation with Ethernet, the more I decided to stick with USB and AES3. Far slower speed, not that much less noise, but controllable, purely based on runs on the board with different products that actually do make audible differences. Some for worse, others better, same ball game (shrug).

[Ben-M]

So those are common mode chokes. I've heard that and I guess I can identify them in bigger components, but never really knew what they looked like in general. Thanks. 

 

And yes, that's what I figured in the "custom" part. A manufacturer/designer still applies their personal skills when designing the components, so there's still some degree of "personal touch" happening. 

 

For me, the verdict is still out and preferred format for me, however, I did find it pretty interesting to finally find out what was in that one specific filter. No Bybee Quantum Purifiers, so maybe it's technically alright :-P

 

I'd have loved to have taken the PCB all the way out, but alas it couldn't be pulled out without desoldering. 

 

4 hours ago, One and a half said:

Ethernet speeds are quite high speed, read frequency. Each path is critical and distances mean a great deal. Ethernet suffers from common mode noise, and the transformers you see are common mode chokes, series types perhaps also with built in ferrite cores. Attenuation then depends on the magnetic properties of the chokes and the ferrites, without strangling the actual signal, so it's not a trivial task.

 

Like yourself, I'm intrigued by Ethernet and its routing of audio data, but the more I read about the noise propagation with Ethernet, the more I decided to stick with USB and AES3. Far slower speed, not that much less noise, but controllable, purely based on runs on the board with different products that actually do make audible differences. Some for worse, others better, same ball game (shrug).