Accessing only one 896mk3HD from multiple firewire Macs ?

[Klaus]

Can it be done , other than plug / unplug firewire connections ?
> feature request in the driver ?
Until now I have several firewire hubs / Macs / and a 896mk3HD interconnected, and have to
unplug not needed Macs...
Or do firewire hubs exist with selector ?

Best

Klaus

[mikehalloran]

Create an Aggregate / Multiple Output Device in the Audio window of Apple's Audio MIDI Setup.app.

Apple has support bulletins on how to do this and some of the MOTU manuals do, too.

[HCMarkus]

Can it be done , other than plug / unplug firewire connections ?
> feature request in the driver ?
Until now I have several firewire hubs / Macs / and a 896mk3HD interconnected, and have to
unplug not needed Macs...
Or do firewire hubs exist with selector ?

Best

Klaus

I think. Mike got that backwards, Klaus... to my knowledge, it's one Computer per Interface, only. Multiple interfaces to a single computer, as MIke has noted, is very doable.

Firewire anything is getting harder to find; it is essentially outdated at this point. Perfectly fine for older systems but workarounds are pretty much required with the latest Macs. So finding a selector switch for Firewire is not likely to be easy.
.

[mikehalloran]

…
I think. Mike got that backwards, Klaus... to my knowledge, it's one Computer per Interface, only. Multiple interfaces to a single computer, as MIke has noted, is very doable.

Firewire anything is getting harder to find; it is essentially outdated at this point. Perfectly fine for older systems but workarounds are pretty much required with the latest Macs. So finding a selector switch for Firewire is not likely to be easy.
.

You're right. I misread it as three interfaces on one Mac instead of the other way around. Sorry

That said, manual FW selector boxes used to exist and show up on eBay now and then.

https://www.ebay.com/itm/300631326537

I have no experience with these at all.

[Klaus]

Thank you all !
I may try to find one in Europe...

Best regards from Switzerland

Klaus

[dgatwood]

Yeah, your best bet would probably be to do something like Dante Virtual Soundcard or AVB.

https://www.reddit.com/r/audioengineeri ... ?rdt=52115

AVB requires very specific NICs, unfortunately (like all Thunderbolt, as far as I can tell, starting at $150 each), so DVS is a much cheaper solution, unless Apple supports AVB over a simple Thunderbolt or USB-C cable (no idea, haven't tried).

[Klaus]

Yeah, your best bet would probably be to do something like Dante Virtual Soundcard or AVB.

https://www.reddit.com/r/audioengineeri ... ?rdt=52115

AVB requires very specific NICs, unfortunately (like all Thunderbolt, as far as I can tell, starting at $150 each), so DVS is a much cheaper solution, unless Apple supports AVB over a simple Thunderbolt or USB-C cable (no idea, haven't tried).

AVB will allow accessing from multiple Macs / WINs ?

Klaus

[dgatwood]

Yeah, your best bet would probably be to do something like Dante Virtual Soundcard or AVB.

https://www.reddit.com/r/audioengineeri ... ?rdt=52115

AVB requires very specific NICs, unfortunately (like all Thunderbolt, as far as I can tell, starting at $150 each), so DVS is a much cheaper solution, unless Apple supports AVB over a simple Thunderbolt or USB-C cable (no idea, haven't tried).

AVB will allow accessing from multiple Macs / WINs ?

Klaus

Yes, both AVB and Dante allow you to route data to multiple destinations.

See MOTU's AVB FAQ: https://motu.com/avb/avb-faq

[Klaus]

Thanks dgatwood
last AVB question
how is the latency overall i/o ?

Klaus

[dgatwood]

Thanks dgatwood
last AVB question
how is the latency overall i/o ?

No idea. I'll find out if as soon as the backordered LP32 actually gets here.

At least in theory, most of the latency should come from the chunk size (buffer size). AVB itself adds only maybe a couple of milliseconds to that. So, for example, if you used a buffer size of 4096 samples at 96 kHz, latency would be 43 milliseconds plus some additional delay ≤ 2 milliseconds for AVB transport, plus some fudge factor for buffer copying in the OS, etc.

And of course, at the minimum buffer size of 32 samples (for some interfaces, anyway), even a single AVB switch would triple the latency from 0.3 ms to 0.9 ms, so it actually is at least possible for AVB to have a meaningful effect on latency, but only if the latency would otherwise be really small.

But this is all just a bunch of theoretical math spewing on my part. I don't have any idea how small a buffer size is actually practical these days. The point is that typically, the bulk of the latency comes from the amount of data that gets captured into a packet before it sends it out over the wire, i.e. the buffer size / chunk size, and no matter what the transport medium is, it doesn't usually add much to the total system latency because the sample buffer latency is so large that it dominates over the tiny latency that the transport medium adds. And that's unlikely to change no matter what medium you use, though as systems get faster and can handle smaller and smaller buffer sizes before the CPU load gets too high, that penalty will of course get smaller.