Nick, do you know how people are clocking multiple machines with no interfaces as in Jerome's now famous 8 mac mini's?

dp---

in the Mac Mini's situation, clocking is done throught the s/pdif input.

Jerome

To explain further: at one point I was able to get AUNetsend/Receive to work, but the two machines kept drifting just as magnumpraw described. Max from FX-T explained that I needed to clock the two machines together digitally in order to avoid the problem. He was right, and it solved the problem nicely. Unfortunately it didn't solve the unusable latency problem.

Now, that doesn't mean FX-Teleport needs a separate sync connection - it locks to the ethernet port, as any audio-over-ethernet program absolutely should. I'm biting my tongue to avoid going farther and saying that it's absolutely retarded if it doesn't, because it would be very rude to say something like that and I'm always politically correct.

But if that's what's going on in magnumpraw's system, that's the solution - and of course unless both machines are Macs with built-in optical ports, it's no solution. If you have to add an audio interface to get a digital input...

yeah, that's what I wondering. I didn't know of anyway to clock the computers without an interface and if you've got an interface, what's the point of the Audio over LAN?

@Jerome said:

dp---

in the Mac Mini's situation, clocking is done throught the s/pdif input.

Jerome,

If I come from a mini into my spdif in my MOTU 308 (which is clocked to my master) then the spdif automatically clocks to the 308: I take it this is what you're saying.

Nick,

You are saying that clocking is (or should be) automatic when coming over LAN but latency is still a show-stopper?

Yes, that is correct. For what I know (but I may always be wrong [;)]), a clocked audio interface (either internally or via a master clock) will automatically distribute its clock to its digital outputs.

What you need to make sure is that the other audio interfaces get their clock via the right input, if they have more than one digital or clock input. For example, if you have s/pdif, multiple adat, word clock, etc, on one interface, then the right clock input must be selected in the setup parameters.

Also, in your case, I would rather clock the 308 internally (or via a master clock like a Big Ben), because the Mac Mini doesn't give you too many choices in terms of clocking. "Slaving" the Mac Mini to the 308 seems more logical and reliable than the other way around.

Jerome

Not only was the show stopped, Dave, rotten fruits were being thrown at the stage. But Rob has Wormhole working, so I'm going to try again.

And yeah, I'm saying that the exercise is completely autoerotic if it doesn't clock to the ethernet port!

Well, in all fairness, sync between the two machines isn't an issue if you're just trying to get iTunes to play on a remote computer. But for our purposes there is no purpose.

Thanks Jerome and yes the 308 is indeed clocked to my Aardvark so it sounds like an audio interface can be avoided if one can settle for a single stereo out.

Nick,

Let's get the show on the road. Everyone around here is dealing with trying to make use of the collections they own. Hence all the multiple cpu's/ram limts/dongle+collection-splitting/64 bit-32apps/no interface-clocking/VI interface+controller issues and so forth.

It's almost like there should be some sort of VI Magazine to cover all this stuff and help us all sort it out.

Great minds think alike! That's the title of the first article in our next issue:

"Multiple cpu's/ram limts/dongle+collection-splitting/64 bit-32apps/no interface-clocking/VI interface+controller issues" by Dave C.

Hey All,

Sorry I haven't kept up with this post...

Hey Nick...I used the strings in a show sequence that I was scoring. But that particular segment I was doing at the moment was very short. Vlns, Vlas, Cellos & Basses all seemed to work correctly. Now then, this is just one Mac (the others are coming in in the next couple of days) so we'll see how it goes once I've got 4 online with the G5.

Also, the clock aspect is very interesting (and frightening) ...I hope I don't run into the "drift" thing...not real sure how to deal with that only having Mac Minis?

It'll drift over time if it's going to, so I suspect it won't. If you come back an hour after connecting to the remote computer and the latency is still good, you should be fine.

@Nick Batzdorf said:

Great minds think alike! That's the title of the first article in our next issue:

"Multiple cpu's/ram limts/dongle+collection-splitting/64 bit-32apps/no interface-clocking/VI interface+controller issues" by Dave C.

Uh...okay. I need to organize my materials a bit but I'll get right back to you with a first draft...say around Xmas... 2022.

@Nick Batzdorf said:

Great minds think alike! That's the title of the first article in our next issue:

"Multiple cpu's/ram limts/dongle+collection-splitting/64 bit-32apps/no interface-clocking/VI interface+controller issues" by Dave C.

Uh...okay. I need to organize my materials a bit but I'll get right back to you with a first draft...say around Xmas... 2022.

lol. Then you'll need to begin again, Dave, as VSL tops 5 TB over 25 instruments, and that lowly 10 multicore processor computer with a modest 560GB of ram can't keep up!

Regards,

Alex.

[[:|]]

Just an update...have just hooked up 5 MAc Mini's, Lots-o-Instances and everything is working great!

The one thing that I discovered that seemed to make the difference from an audio side is having DP installed on each computer. Even though I'm not using DP to host (I'm using Rax), I did set the "Core Audio" inputs and outputs via DP and that seemed to make all the difference in the world. I really don't have a specific reason for this but I just know that when I was initially experimenting with this system (before the Mini's arrived) I was using my MacBook. It happened to have DP5 on it and everything worked great. I then got the 1st Mini and the audio was extremely eratic. I then installed DP (merely to have a different playback engine) and immedialtely it asked to assign the outputs. Once that was done, I've never had another issue as far as latency, eratic audio etc.

I have no idea why this has made a difference. Maybe someone out there smarter than I can give me a reason. But hopefully this will help others that are trying to use WH2!

Ah, here's the thread. Well, I started a new one - sorry about that.

It's not working for me at all between two G5s. Half a second of latency.

@Nick Batzdorf said:

It's not working for me at all between two G5s. Half a second of latency.

I second that. Wormhole 2 gives me unacceptable latency - 3 Mac G5s, no intel Macs. Perhaps its an intel thing? And I really really wanted it to work (because after trying it endlessly I ended up buying two of M-Audio's new 4ADAT lightpipe boxes instead.)

Well, after a couple of weeks of battleing....I finally have to "raise the white flag" and call it a day. Wormhole2 does not work. I have not had a latencey issue but a "clicks and pops" issue when syncing up machines. I constanly had to "check and uncheck" the sync box on my host machine to what ever Mac Mini that lost "audio sync" to get it to stop "clicking and Popping".

I felt like I was so close!

I emailed Mark at Plasq and this was the reply:

"Hi there Rob,

Sorry for the delay in getting back to you; I had to ask Adrian (the creator of WH)for some help on this one. This is what he had to say:

"I think this is not going to work, and I coded WH2.

Sync is a function that lock multiple channels together when they are transfered from machine A to machine B. It's like having one big multi-channel connection. However there is no way sync could keep channels on multiple machines locked.
As the machines couldn't know about sample positions on other machines.

Sync can only work if one machine/app is the sender and one machine/app the receiver.

If you use WH2 with that many machines, you'll get tons of network travel negotiating connections, and you won't be able to sustain a low latency connection.

The only way to deal with latency in such a scenario is to use loops and have all audio leave the main host. In that case there won't be sync and you'll probably have to run the loop with quite some latency to get things to work."

Hope this is helpful? Or at least makes more sense.

All the best,
Mark."

So there you have it....unfortunately this is not a viable solution anymore. Guess us Mac guys have to wait or FX-Teleport (which I got an email form Andy at FX in which he wrote: Hi Rob,

We really hope to ship it early next year. Be sure to sub to our news update
at www.fx-max.com (if you've not done so already [:)] ).

Best,

Andy).

So there you have it. But hey, nothing ventured, nothing gained!

Rob

Thanks Rob. I'd be happy if I could get it to work with one machine.

Mark responded to an email I sent him by saying I should lower the buffers in Wormhole, which I may or may not have done when I tried it. That's my next test when I get a chance.

Hey Nick,

What worked for me was setting the buffer on the host (G5) at 256 and the latency on the slave (Mac Mini) at 0. And believe it or not, no latency what so ever.

The thing that makes me wonder abour this program is that I had all 5 machines running at one time...no latency...it was just an issue of the pops and clicks after a moment or two.

So close, and yet so far a way! A real bummer! C'mon FX-Teleport!!! It can't be ready soon enough! But at least it sounds like they are right around the corner with it.

Hope this helps!

Well, as you say, it turns out that it does work pretty well on one slave machine if you set Wormhole's buffer to 0 on the sending machine and - realistically - about 600 samples on the receiving one (I was getting crackles at lower settings and lots of voices when I played a harp). That gets the latency down to a reasonable level - still noticeable compared to an audio interface but not out of the question.

I'm going to have to try it with four of them (for eight outputs) and see how that does. It's good to see that it might work with for one slave machine.

This test was with...dare I say EWQLSO Platinum Pro on the slave, which is the toughest test of all since it has all those release samples ringing and eating up voices.