PFC: Can I get a couple extra eyeballs on a Dante problem?

Jay Ashworth

Well-Known Member
[ Promoted from comment; I know better. This is a copy of a PSW post; apologies if anyone sees it twicet. ]

My theatre has a small, in-booth Dante net, used primarily to route direct-out audio from our LS9-32 to our Protools rig (PT10 on a Mac Pro cheesegrater, running DVS and Controller). (There's also a Rednet1, but I've never used it.)

That much has been working fine for 6 years or so; we use it 8 or 10 times a year for concerts.

For our next theatre show, we needed to stuff the band in our scene shop, so we pulled out our backup LS9, never used but a couple times, upgraded it from 1.11 to 1.35 firmware, and cabled it up from the shop to the booth, through our stage network, about 90 feet of in-wall ethernet cable.

Controller sees the MY-AUD cards come up just fine, and after adjusting the patching on the B console -- the one I'm trying to use as, essentially, a stagebox -- I got a burst of audio, maybe 5 seconds, a couple of times.

And that was all.

(My audio source was an open mic, listening to Alexa across the room; I could see the level continuously on the console meters, and via SM2 up the separately cabled ethernet run that put that console on our Production LAN. The audio I got through was identifiably the smooth jazz I asked Alexa to play, and as clean as you'd expect across a 30 ft room through a Heil PR-22. Everything is at 48kHz, and I tried a couple permutations of word clocking on the LS9's with no luck.)

A buddy of mine is the FOH mixer at the Berklee Performance Center in Boston, and was kind enough to listen to me yammer and whine about it, but the only suggestion he had that I didn't tell him I'd already done was to loosen the latency knobs for the relevant devices in Controller -- they were at .015 and I opened them to, I think, .05? 4 clicks up instead of 1. Sadly, that didn't help either.

There don't seem to be a lot of diagnostic tools for Dante networks (that aren't hardware costing 3 grand and up), and Controller doesn't tell you much either... so I'm at sort of a loss.

Anyone got any suggestions I can try out? I'll be back in the room 2 or 3 times before tech week starts on Monday, so I'll still have time to apply a bigger hammer, soon's I find out what that is... :)
 
[ Promoted from comment; I know better. This is a copy of a PSW post; apologies if anyone sees it twicet. ]

My theatre has a small, in-booth Dante net, used primarily to route direct-out audio from our LS9-32 to our Protools rig (PT10 on a Mac Pro cheesegrater, running DVS and Controller). (There's also a Rednet1, but I've never used it.)

That much has been working fine for 6 years or so; we use it 8 or 10 times a year for concerts.

For our next theatre show, we needed to stuff the band in our scene shop, so we pulled out our backup LS9, never used but a couple times, upgraded it from 1.11 to 1.35 firmware, and cabled it up from the shop to the booth, through our stage network, about 90 feet of in-wall ethernet cable.

Controller sees the MY-AUD cards come up just fine, and after adjusting the patching on the B console -- the one I'm trying to use as, essentially, a stagebox -- I got a burst of audio, maybe 5 seconds, a couple of times.

And that was all.

(My audio source was an open mic, listening to Alexa across the room; I could see the level continuously on the console meters, and via SM2 up the separately cabled ethernet run that put that console on our Production LAN. The audio I got through was identifiably the smooth jazz I asked Alexa to play, and as clean as you'd expect across a 30 ft room through a Heil PR-22. Everything is at 48kHz, and I tried a couple permutations of word clocking on the LS9's with no luck.)

A buddy of mine is the FOH mixer at the Berklee Performance Center in Boston, and was kind enough to listen to me yammer and whine about it, but the only suggestion he had that I didn't tell him I'd already done was to loosen the latency knobs for the relevant devices in Controller -- they were at .015 and I opened them to, I think, .05? 4 clicks up instead of 1. Sadly, that didn't help either.

There don't seem to be a lot of diagnostic tools for Dante networks (that aren't hardware costing 3 grand and up), and Controller doesn't tell you much either... so I'm at sort of a loss.

Anyone got any suggestions I can try out? I'll be back in the room 2 or 3 times before tech week starts on Monday, so I'll still have time to apply a bigger hammer, soon's I find out what that is... :)

Need more information and screenshots of Dante Controller preferably. Dante Controller can tell you a wealth of information if you're looking in the right places.

My first questions are... what else is on this network? What control do you have over that network? What is your clock leader in Dante? I have a suspicion that your two consoles are fighting for clock priority...


aaaaaand now I'm picturing sock em bopper robots, but with LS9's Battle Bot style instead of the red and blue robots...
 
MY-AUD 1 and 2 from the permanent A console
MY-AUD 1 and 2 from the stagebox B console
Rednet1, permanent
MacPro, permanent.

I manually patched the B box outs to the A box ins in Controller, and got green checks on all of them. I can change anything I want to. I do see occasional tradeout messages about master clocking in the Controller log, thought not like it's flapping.

I've tried -- on the smaller version of this network, just one LS9 and Protools -- setting clock preference to the MY-AUD (where it usually lives anyway) and had everything come apart for no obvious reason...

I *think* I've been through all of Controller and didn't see anything else that looked likely -- IT is my day job; I'm used to trying to extract state from software into my head.
 
Last edited:
My limited Dante experience has been with X32/M32, but I think it still applies. Update Dante card firmware on both consoles. Make sure one of the Dante cards is designated as the clock master, not the console internal clock. Try putting both consoles together in the booth, to eliminate possible network issues. Do away with the secondary network until you get the primary network working.
 
MY-AUD 1 and 2 from the permanent A console
MY-AUD 1 and 2 from the stagebox B console
Rednet1, permanent
MacPro, permanent.

I manually patched the B box outs to the A box ins in Controller, and got green checks on all of them. I can change anything I want to. I do see occasional tradeout messages about master clocking in the Controller log, thought not like it's flapping.

I *think* I've been through all of Controller and didn't see anything else that looked likely -- IT is my day job; I'm used to trying to extract state from software into my head.

So both ports of your my16 cards are on the same network? Are the consoles set to redundant or switched?

I assume by the number of things in the list that there is a switch somewhere? Is there anything else in the switch that my be competing for P2P priority? I'm never a fan of the go to forum response you'll see for Dante questions about QOS and multicast whatnot, but if you're dealing with a converged network it can cause sort of the symptoms you've described... which can be solved that way or by segmenting traffic in VLans.

I have a suspicion we're looking at clocking here. Let's grab a screenshot of the clock status page with everything powered up and connected the next time you're around Dante Controller.
 
My limited Dante experience has been with X32/M32, but I think it still applies. Update Dante card firmware on both consoles. Make sure one of the Dante cards is designated as the clock master, not the console internal clock. Try putting both consoles together in the booth, to eliminate possible network issues. Do away with the secondary network until you get the primary network working.
Moving them together isn't practical at this point, but I did reset each to clock off its Dante Lo card, and that didn't help, alas.

No secondary networks, only one switch at each end:

Core is a Netgear GS724T, down at the B console is an 8 port TPlink GigE; neither switch has anything on it but Dante traffic.

Photos to follow.
 
Here are screen photos of the controller clocking page, and the word clock selection pages on each board.
20220720_155446.jpg


20220720_155500.jpg

20220720_155652-jpg.23244
 

Attachments

  • 20220720_155652.jpg
    20220720_155652.jpg
    264.3 KB · Views: 404
A couple of things right off the get go... that is an older version of Dante Controller, so I would start with an update there. I think you'll notice a significant difference in the amount of information you're able to gather. Similarly as mentioned before let's check the firmware of your Dante cards after the controller update.

Next, let's unplug the ethernet from the "secondary" or second port on the my16 card. I suspect you may be running those cards switched in which case you've created a network loopback and you don't want that.

Then let's see if the issue is still there and what new information we have from Dante Controller.
 
Controller is 3.2.4.1, I'll go look for an update.

As controller notes there, the Interface is Down on all secondaries, consonant with my not having any cables plugged into those. I did do at least a little of my homework before diving in. :)

[ Why does *everyone* got to be like this. Download controller requires an account on Audi's website, which will almost certainly require us to find a piece of paper shipped with these desks 15 years ago...

Anybody got a direct link to the 4.2 Mac Controller installer?

And will it run on 10.8.4, a hard fence I cannot practically work around? ] ]
 
It will not. Controller 4.6 requires 10.15 or newer, which is pretty thoughtless; *none* of my Macs are newer than 10.13, nor - for Not Breaking Other Stuff reasons - likely to go past that; the PT machine can't, it's a Pro5,1.

And I can't even test-install it, cause I don't think I have the 3.2 installer, and if I break it, I'm schrod.

[ I've been having some pretty bad tunnel vision working on this; don't know why. As I note below, that Mac Pro need not be the only thing I run Controller on... ]
 
Last edited:
You don’t need any additional information to download Dante Controller other than an email, but without updating your version you will have a harder time updating Dante Firmware or probably shouldn’t, I can’t say for certain that’s your issue, but it could be depending on the firmware version of the consoles, stage boxes, computers and so on…. You Mr Ashworth may be at an impasse.
 
You don’t need any additional information to download Dante Controller other than an email, but without updating your version you will have a harder time updating Dante Firmware or probably shouldn’t, I can’t say for certain that’s your issue, but it could be depending on the firmware version of the consoles, stage boxes, computers and so on…. You Mr Ashworth may be at an impasse.
Well, it was pointed out to me that I can run Controller on ... other things. Like my X1 Carbon laptop with Win 10. So I'm gonna try that tomorrow. :)
 
Your main console has two MY16 cards. Do know for sure that the console is getting its sync from the same MY16 that is designated as the master for Dante? Any chance the direct outs are set for multicast? That would be bad.

I'm not confident that the switches you have are up to Dante. I've gotten away with plain old switches once or twice, but I wouldn't bet the farm on it, and Audinate generally says not to do that. The Dante net that I built uses a pair of Cisco SG series switches, for which I have a complete how-to for Dante configuration. I chose to spend the money for the Ciscos instead wasting hours on frustration. I don't know what half the settings do exactly, but they are extensive.
 
I'm sure nothing's multicast; I Preferred slot 1, and slot 1-1 is the master now on both desks, and while there's presently a temporary second switch, the 724 was the only switch for years, and worked fine. And indeed, this problem started in this setup while I only had the Slot 1 of the B desk connected directly to the cable. So I think it unlikely to be my root cause. More on this story later today.

My local reseller tech guy, whom I talked to about it in the afternoon, will hopefully be swinging by today or tomorrow, to help me suss it out, and says he's built on lots of switches, including Ubiquiti's, with little to no problems...
 
My MY-16AUD cards are firmware 3.3 in the A desk, and 3.7 in the B desk (it got its cards about 6 years ago, instead of 12).


Of course, Dante Updater says "Contact Manufacturer". :)

Was there anything else people thought I should look at in the more current Controller?
 
SORTED.

It was in fact the clocking, though which combination of bad settings isn't clear.

We are presently clocking both desks from Dante, with no Preferred master and no sync to card forced in C Controller for either desk.

Thanks for all the suggestions, and to the local megachurch hotshot who actually flipped the right switches for me. :)
 
As was stated, Dante cards need to be set as the clock source for it to work. We ran into this networking a bunch of x/m32s together while also dealing with the death pop static discharge from the aes50 lines. Nothing was happy with anything being clock source til we fixed the cable issue there, but that's another thread.

One other thing to note is that the Dante cards will figure out on their own which one is the clock master. The controller is not really a controller in the sense that it acts as any sort of arbiter between devices on the network. It really just monitors the clock status and can command subscriptions and reconfigure devices. When you select a clock master you only can indicate a preferred master. If the cards figure there's a better clock source by an unspecified margin, it will ignore your preferred master and select that one instead.

The other thing you might want to check is that IGMP snooping is enabled on the switches. Dante makes a lot of use of multicast traffic for discovery and will also switch audio streams to multi as you fan out subscriptions.
 
also dealing with the death pop static discharge from the aes50 lines. Nothing was happy with anything being clock source til we fixed the cable issue there, but that's another thread.


Is there a thread for this somewhere? I just had a regularly occurring daily pop of static that was as loud as a gunshot that I couldn't figure out but everyone consulted said it was an "aes50 cable issue" and not much beyond that.
Would love to discuss. Thankfully it never happened during show and the last couple runs have been flawless so got distracted and forgot about it until now.
 
If you search the forum, you might find more. It boils down to always use shielded cat 5e/6, with ethercon connectors, for AES50. The shielding and grounding through the ethercon shell prevents static electricity from being induced into the wire pairs, causing a sync unlock and nasty noises. Mechanical disruption (bending, twisting, kinking, crushing) to the pairs can also cause issues, so rugged cable and protecting it from being stepped on or rolled across is also a good idea. It's a hardware issue that Behringer did not know about until the X32 was in the field, so the manual does not mention it. They later acknowledged the issue and came out with a rugged cable.
 

Users who are viewing this thread

Back