SQ to AHM via AR2412 Monitor Port

AHM AHM General Discussions
1

Hi there!

Today I did some fancy stunt but now I’m not sure if that is a good idea.

So this is my test setup:

SQ5 → AR2412 DSnake Port

AR2412 Monitor Port → AHM-16 with SLink card

AR2412 Expander Port → AR84

Turns out, when sending out from the SQ5 on ME/Monitor channel 1, this comes in as I/O-card channel 21 in the AHM-16. Which actually is a very nice feature since all DSnake/SLink outputs on the SQ5 remain available for the outputs on both stage boxes. So I have 16 additional channels I can route into the AHM-16. (Of course, only 8 of those make sense, since the AHM-16 has only 8 analog outputs to got to the PA system.)

But is this really a good idea?

The AHM-16 will likely do sample rate conversion, as it is also set to clock master. I don’t want to set it so slave since I want to also be able to disconnect the SQ5 and run the AHM-16 on its analog inputs only.

Ideally I would want to solve this with Dante, using Dante cards both in the SQ5 and the AHM-16, since I need a second CAT cable either way. One for DSnake/Slink. And another one either for controlling the AHM-16, which sits on stage, or for control+Dante in one cable.

But as it is for now, the budget of the project doesn’t allow for two Dante cards.

Best

Doc

2

I can see why this works, but I am still surprised that it does.

I think you are correct to be concerned about clock rates. The fact that you set the AHM as master clock doesn’t mean the SQ will receive any clocking information based on the way you have it connected. After all the ME system is a SEND ONLY system. I have no idea if the SQ will get clocking information from the AHM. It might, but it might not since it clearly isn’t designed to be connected this way. Even if the audio passes from the SQ to the AHM, I wonder if there won’t be clock drift that occurs over time that will mess things up…..

That being said, as long as you test this over a decent amount of time (to allow any clock drift issues to manifest) and it works, then I would say run it in the short term.

While Dante can certainly be used to connect the two device together, another option is simply to install an SLink card in the SQ and connect it to the AHM’s Slink card. Of course this only works if the SQ’s expansion port isn’t being used already. The benefit is that the SLink card is pretty inexpensive. I think it retails for about $299 USD. That’s certainly a lot less than having to acquire two Dante cards!

3

I think good old DSnake is a simpler than we like to think. Of course, I’m not an A&H dev person, so this is an educated guess: I think the “ME monitor protocol” is actually nothing special, it’s just channels 21 to 60 of DSnake in the outwards direction from the console.

As it says here, DSnake is a protocol supporting 64 channels in both directions: https://www.allen-heath.com/content/uploads/2023/11/DX-gigaACE-white-paper-1.pdf

So, the SQ5 will just send out 60 channels of audio or so. Plus a clock signal. Channels 1 to 20 appear as DSnake outputs (which would be AR2412 + AB168 = 20 outputs), and channels 21 to 60 are shown as ME outputs in the SQ’s routing system. Technically, I believe this distinction is mostly in the user interface.

Why do I think so? Well, the AHM-16 equipped with an SLink card and connected to the monitor port of the AR2412 actually can receive the SQ’s DSnake Output 1, too, on its IO-expansion channel 1.

But, for the question whether or not this is a good idea, it’s probably not even relevant that this trick does work.

The question rather is, if there’s any hidden catch that i am not seeing.

The SLink card in the AHM-16 will do sample rate conversion if necessary (that is, if the AHM-16 is not clock slave). The question would be then, if this has any audible impact at all. (The SLink port will likely convert from/to the 96khz used internal in all newer A&H systems.)

See also https://support.allen-heath.com/hc/en-gb/articles/38290033891985-SLink-Connections – there’s also information on clock leaders in that document. If a Qu is involved, it will be the clock leader, since it cannot be clock slave.

The same thing would happen with the Dante card in the AHM-16. (Who runs their Dante at 96khz? Everywhere I see Dante it’s running at 48khz.) Actually I think a lot of Dante setups do sample rate conversion, people just don’t think about it.

I appreciate the idea of using an SLink card also in the SQ5, but this would mean 3 CAT lines then, 2× DSnake and 1× control.

So if there is no hidden in my solution, I’d rather stick to it… the additional outgoing channels are there, not being used.

(Of course, this renders the use of an actual ME1 system difficult, if not useless. But I don’t have one anyways.)

4

I’d be cautious as this is not a supported or tested setup. SLink sample rate conversion is synchronous and doesn’t eliminate the need for correct clock sync. Since the SQ doesn’t have an option to ‘sync from dSNAKE’, your only option here is to sync the AHM from the dSNAKE stream. Again this would be an unsupported setup, but failing to set a clock leader and follower will likely result in audio glitches as the clocks drift. This is best tested with a sine wave over a period of time.

5

Thanks for your reply, @NicB . So, if I put the AHM-16 to clock slave (sync to I/O port), what will it do when the DSnake system goes down? Will it fall back to its internal clock?

(Part of this whole thing is: I want to tear down the FOH, while some DJ continues on stage. So the DJ feeds into the analog ports of the AHM-16 and then goes on forever standalone, while the SQ5 and I go to bed.)

Further question: So, when going Dante, the Audinate chip in the expansion card would do full sample rate conversion for me, without the issue?

I know it’s more like a dirty trick what I am trying to do, but it’s nothing to the trouble I ran into when trying dirty tricks with AES67 in another setup, eventually going all-Dante.

SQ Mixes 12 stereo available but only 14 physical sockets
6

No, Dante will also require a clock leader / follower setup (in the mixer settings AND in Dante Controller). But the advantage would be the option to set the AHM as leader (internal clock) and the SQ5 as follower (sync from Dante).

The AHM will indeed revert to its internal clock when a valid clock is not present on the SLink port, however this will likely generate an error or warning every time you disconnect the mixer. Just to reiterate, it’s an unsupported setup so not something we tested for reliability or performance.

7

Thanks again.

I did a little more testing today. However, I cannot yet provide any reliable result. So I fed the 500Hz Sine from the SQ5 generator into this dubious system and I wasn’t able to hear any clicks on the speaker I had connected to an analog output of the AHM-5. Now one thing is, I really can’t listen to a 500Hz for much longer than 2 minutes without going nuts, so I didn’t do that.

The other thing is, in this test session it seemed irrelevant what the clock sync was set to in the AHM-16. The unit log remained silent, and disconnecting the DSnake cable in both scenarios yielded a seamless playback of audio coming from an analog input of the AHM-16. So it remains unclear whether any clock sync did happen at all in this test.

I’d feel much better doing this with Dante, however the clocking would remain a hot topic still. I installed an AHM-32 with a Dante card earlier this year, all good with that, but thinking about it now I should have set the clock source to IO/card on that one and leave it to the Dante network to sort out the master. I know I can set a preferred clock master in Dante Controller. (Currently that installation does not use Dante, but the budget allowed for it, so I put the card in for the future, since their analog MixWizard hooked up to that AHM looked like it was way over the zenith and the replacement will most likely be digital).

I know there is the “sync to external” or something option in Dante Controller for the SQ-Dante cards… which basically means the Dante card will sync to the internal clock of its host… what happens if this is switched on for two Dante cards in the network?

So, while my experiment in the end might not be successful, it turns out there is a lot to learn about this clocking thing here, and there might be a slight chance that lurking readers benefit from the weird discussion.

Best

Doc

8

Today I failed again at breaking this setup. :wink: Introducing to the game: Good old Qu-16.

So, when I do this:

Qu-16 → SQ-5

then I can get the ME/monitor from the Qu signals also from SLink/Dsnake channel 21 on. And when I do not set the SQ’s clock to SLink, it properly fails. With nice pops and clicks as to be expected.

However, when I replace the SQ with the AHM, so going:

Qu-16 → AHM-16 / SLink

Whatever clock sync I set on the AHM-16, I do not manage to get audible clicks and pops. Even though in theory it should clearly fail the same way as the SQ-5 would as receiver.

It would be lovely if this failed properly on the internal clock setting of the AHM, because then I’d feel like it works properly when set to I/O card clock.

Best

Doc