Re: [QLab] Qlab issues with aggregate devices (Echo AudioFire 12)

[John Leonard]

There are problems with aggregate devices in 10.7.4. These have been identified by one manufacturer (Metric Halo) and they have re-written their driver accordingly. Other manufacturers may have been caught by the same problem and not yet sorted out their own drivers.

For safety and security, revert to 10.6.8 if possible.

John

On 25 Oct 2012, at 16:21, Benedict Almond <bjr...@googlemail.com> wrote:

> Hello All,
>
> I work at a theatre in the UK and we are experiencing some issues within our system. We have one Echo Audiofire 12 running from the 800mb/s firewire port via an 800-400 cable and originally had the second slaving (via the 400-400 FW port on the back) from that. We aggregated the sound cards together in audio midi settings to create a 24 output device recognised by Qlab where we run our sound samples from. There have been recurring problems of the second sound card (outputs 13-24) dropping out (or distorting the audio) requiring a restart of the program which isn't always possible (working in a theatre where the cues come very quickly).

[John Leonard]

So what was the last stable set-up that you had that didn't have aggregate device problems?

John

On 25 Oct 2012, at 18:21, Dave Norton <dav_n...@hotmail.com> wrote:

> Hi John,
> We tried this, by running off our other mac which runs 10.6.8 (snow leopard). Still having the same issues. It's got to be an issue between the qlab updates and Echo Audiofire driver updates, although we also rolled back the drivers to a supposed 'more stable' version, still with the same results.

[David "luckydave" Memory]

On Thursday, October 25, 2012 at 2:49 PM, John Leonard wrote:

On 25 Oct 2012, at 18:21, Dave Norton <dav_n...@hotmail.com> wrote:

It's got to be an issue between the qlab updates and Echo Audiofire driver updates, although we also rolled back the drivers to a supposed 'more stable' version, still with the same results.

You can always revert to an older version of QLab for your testing, by downloading it here:

http://figure53.com/qlab/downloads

--

luck...@figure53.com 

[Christopher Ashworth]

On Oct 25, 2012, at 11:21 AM, Benedict Almond <bjr...@googlemail.com> wrote:
>
> Do we think this maybe a hardware issue?

Seems possible; if the drivers that were working before aren't now.

> Is there anything else I should try?


Are the devices running off wall power or firewire power?

-C

[Dave Norton]

We're unsure which version of qlab was running and which drivers were used when we had the shows running without problems. We've taken the card out of the system and shall test it running audio for a length of time to see if its a hardware fault. We've had no issues with using 1 of the echo's as audio device 1 and 2 edirol 101s aggregated to give us the other 12 outputs we need. They run off separate USB ports so hopefully no issues with aggregating and syncing. I'll update you with any more info if we find out the drivers, qlab version etc. Thanks for the help

Dave

[Christopher Ashworth]

Hi Dave,

For what it's worth, I do not believe the version of QLab should affect this, as QLab is in general unaware whether a device is an aggregate or not. It treats all devices identically.

I may have missed it, but if you haven't tried running from wall power that's worth a shot.

-Chris

[Dave Norton]

So after 2 weeks of running things with a revised system (without linking Echo audiofire12 FireWire devices), I've had issues again.

Qlab is setup with audio output patch device 1 as the Echo Audiofire12 via FireWire and output patch for audio device 2 as an aggregate device of 2 edirol 101's both connected via separate USB ports on our Mac mini. Operating system is 10.6.8 and spec is 2.53Ghz intel core duo 4 Gb 1067 DDR3 ram. Qlab version is 2.3.8. All devices are powered from a clean earth mains feed.

Over the last week we've had issue with the second device where files delayed when fired (they appear to come a bar later than those coming through the first device, most annoying for our musician) and yesterday's show it dropped out completely again, forcing a restart of the show, which re established them working. Because of the issue we had with the linked audiofires we doubled files up to allow us to the same audio with both devices and it feels to me link the out of sync issue could be to do with that but I'm worried about it dropping out yesterday. There are crucial cues that have been plotted on the second device having set it up with no problems and it worked fine until the end of last week, where we started having the problems.

Anyone have any more advice? I have downloaded the other versions of Qlab and have various versions of device drivers and I need to spend some time (that, lets face it none of us have in theatre) testing combinations until a stable set of results comes up. Other than this and buying a new sound card i don't know what else I could be doing?

For what it's worth, the second echo device that we removed has been running fine on its own for the last two weeks without issue.

[Christopher Ashworth]

Sorry to hear about the continuing troubles; that's not fun.

In terms of saving testing time, I think it is very unlikely that changing QLab versions will affect this, as QLab does not know or care what kind of device it is talking to.

Two things that I may be asking again, since I don't remember if they were already asked:

• Did you send us any console logs to sup...@figure53.com? I didn't see any in our archives when I just searched. If you have logs they may provide more clues.

• Is guaranteed sync set on the cues going to the aggregate device? If so it may help to turn this off.

Best,
Chris

[Dave Norton]

Hi Chris,

I think Ben may have sent console logs last time we had an issue. So far in today's show there are no problems but I will send you a console log between shows this afternoon. I'll also make sure guarantee sync is off. Hopefully that's the issue. Cheers for you help

Dave

[ra byn taylor]

If your rig is stable with only one or the other audio devices but not stable with both, the logical decision is to stop using the rig in aggregate device mode...

I've never trusted the aggregate device concept. Too many chances for drivers to want to work at different sample rates. No true "sync".

FYI, the aggregate device function appears to be currently broken in 10.8 at the OS level as per lots of discussion on the MIO list...

ra byn (robin)

[Mark Valenzuela]

Agreed. When I have tried to put together aggregate devices, they have never been stable. If you are in a position where you are able to do more testing on your rig, I'd start by cutting that out of the equation. 

Mark

--
Change your preferences or unsubscribe here:
http://groups.google.com/group/qlab
 
Follow Figure 53 on Twitter: http://twitter.com/Figure53

[Eric Lott]

I've had my fair share of issues aggregated devices especially on newer OS's. Recently downgraded a new machine to 10.6.8 just so we didn't lose the ability to go aggregated when needed. Once you get it set up it works well but can cause plenty of aggravation along the way.

How do you have them cues programed? I haven't really tested it in depth, but I find a fire all children simultaneously group to be closer then an autofollow or other group options when it comes to matching up two files. Like I said can't prove this may just be a placebo effect of knowing I changed something, who knows.

Is there anyway you could get away with 22 outputs instead of 24. I know with the MOTU Urtlalights I've used the main/monitor outs to get two extra outputs out of the unit a couple of times.

Hope that helps and best of luck.

Eric

[Dave Norton]

I have read the console log for the last couple of days. I'll send you guys (qlab team) a copy. We don't have any other options in terms of using multiple devices without aggregating them, unless you're suggesting setting up qlab to run though 3 separate devices and patching accordingly! Surely that's counter intuitive.

[Rabyn]

I'll take anything over instability...reduced function, reduced design, different gear, etc...

If the issue is truly using two devices in aggregate mode, a solution is available and simple. It's a matter of compromise or money. No need to keep wasting time trying. Time is money too. My solution would be to get a device that has adat ports and add extra I/o that way. RME ff800 and some stand alone 8x8 adat boxes. In my case I can do up to 18 I/o with (2) metric halo 2882 without breaking a sweat. Another 2+ if I count the headphone output on the main box. And another (2) if I use the AES or SPDIF out to another 2 channel box.

Focusrite pro40 with adat I/o should be good for 16. Presonus audiobox 1818vsl and adat I/o for 16. Cheap compared to a house refund or bad review.

Ra Byn

[ra byn taylor]

I end up with various board options depending on what has been rented or the house owns. Generally it's a M7CL or PM5D but my house venue has two VI6 consoles. I also use the smaller Yamaha 0/1v96 but I think it only has 1 mini MY slot. Is there a mini MY dante option?

Most of what I do with QLab is less than 8 outputs so I don't have a problem in general. But if DANTE is rock solid & makes a 32x32 rig simple, affordable, etc... it sounds like a  great option. Lets say you only need 8 total, do you still have to kill 32 on the board by using it? Does the Dante card i/o just show up on the patchbay via slots?

Is it possible to have (2) Dante cards to add redundancy? 1 thru 8 on Dante A / 1 thru 8 on Dante B?

Has anyone had problems with Dante & Qlab in a show? If so, did you figure out the problem& what was the cause? For those of you who are using Dante, what are the downsides in your mind? Pointers to doing it right?

Thanks in advance,

ra byn

On Fri, Nov 23, 2012 at 7:00 AM, Macthoast <thor...@gmx.de> wrote:

Hi Ra Byn,

what kind of mixing desk do you use? There are several digital desk where you can use Dante Cards, i.e. Yamaha, Allen&Heath, Digico etc. With the Dante virtual sound card by Audinate, you can run easily 32x32 channels of audio I/O on your Mac Mini. No need for a stack of audio interfaces, driver issues etc.

If you go directly to your amps, you can use an Dante Breakout box as Fouraudio DB01 for example. 

[Macthoast]

Hi Ra byn,

The Dante card works like every other ydai card. you can patch your channal inputs to the dante channels. Since the Yamaha desks just support 16ch on a slot, you just have 16 i/o. For more you need a second card. so if you need 10ch input only, you just patch 10.

The Dante cards have a primary and a secondary network connection. They can be daisy chained or two port for redundancy. In an redundant network you need two switches. But because your computer has just one RJ45 connector, redundancy makes no sense at this point.

In Germany a Dante card for Yamaha desks is 600€. Until today i found Dante very reliable and cool. And you can record directly to your favorite DAW on your computer, even with 32 inputs. This, you can set up at in the virtual sound card driver by Audinate, 16x16, 32x32 or 64x64 if your computer is fast enough. No problems so far. If there is a Dante possibility for the SI6, i don´t know.

Regards 

Thorsten