Film-Tech Forum ARCHIVE: Atmos playback Issues

This is topic Atmos playback Issues in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=003531

Posted by Monte L Fullmer (Member # 2797) on 12-21-2018, 04:52 PM:

Little help, if possible:

We converted a house to ATMOS

CP850 Processor
GDC SX3000
Barco 23B

Some Atmos content will play normally, but some will cause a crash in the playlist with an error reading of Atmos failed to load and playback stopped.

Thus, I kick out the Atmos content(usually the Atmos VF file) and play the 5.1 content with no issues afterwards.

Would this be a SMPTE issue, or more of an internal issue?

Thx - Monte

Posted by Harold Hallikainen (Member # 5405) on 12-21-2018, 04:59 PM:

I believe Atmos has always used SMPTE packages, so I don't think that's an issue. Reloading the content does not fix the issue? Since you say it happens with certain content, it would be interesting to know if other people are having the same issues with the content. So, is the content, the server, server version, etc.?

Good luck!

Harold

Posted by Steve Guttag (Member # 268) on 12-21-2018, 05:42 PM:

I would reach out to GDC and send them the logs. You might find that there is a software compatibility issue or something along those lines. GDC is normally very fast at responding.

In fact, I'd also reach to Dolby as well in case they know of potential conflicts.

Posted by Leo Enticknap (Member # 534) on 12-21-2018, 10:03 PM:

Interesting. I installed an almost identical system a couple of months ago (SX3000, CP850, Dolby multichannel power amps, Barco 20C). I had only one report of trouble after the installation, that the CP850 suddenly stopped executing IP commands sent by the GDC as playlist cues. Rebooting the CP850 fixed it.

Posted by Steve Guttag (Member # 268) on 12-21-2018, 11:08 PM:

How are you liking the Dolby MCA? Noisy beast but it packs a lot of power in a tiny space.

Posted by Leo Enticknap (Member # 534) on 12-23-2018, 01:45 PM:

I've only done that one install with them, but agree completely. It needs a lot of fan power (and thus noise) to expel the required amount of heat from the box, but that's because it packs into 4 RUs (IIRC) what would need at least double that with any other power amps.

The setup was very pleasant, with the DARDT figuring out the power budgeting down to giving you a diagram showing which speaker connections go where at the back, and an intuitive and well laid out web UI.

I wasn't involved in the purchasing decision or negotiations, so couldn't comment on the cost aspect of them.

Posted by Steve Guttag (Member # 268) on 12-23-2018, 04:54 PM:

In a high channel density situation (e.g. Atmos), the cost/channel is quite competitive. They now have a form of it, 16-channel that will handle a large number of rooms on just 1 amp for a 7.1 system (including the bi/tri or possibly quad amping the stage channels).

An IMS3000 plus MCH could be a very compact sound system indeed with minimal wiring. I'm not sure I like putting all of my eggs in that one basket though. Even with QSC DPA-Q amps, I try to ensure that there are always two for the stage channels (one on left/right and one on Center/Sub) so there is a "backup" of sorts. I know one tech that always specifies a spare "Center" amp in the rack such that all one has to do is move the phoenix connector to the spare amp (it always is fed the signal).

Posted by Leo Enticknap (Member # 534) on 12-24-2018, 12:50 AM:

quote: Steve Guttag
I'm not sure I like putting all of my eggs in that one basket though. Even with QSC DPA-Q amps, I try to ensure that there are always two for the stage channels (one on left/right and one on Center/Sub) so there is a "backup" of sorts.

Agreed. If you did have a hypothetical 7.1 house with an IMS3000, one 16-channel MCH and nothing else, then if anything happened to take the whole amp unit out, that screen would be down, and you'd have little in the way of workarounds. Another single point of failure is that they need a 30-amp power supply with a NEMA L5-30 (twistlock) outlet, meaning that if the supply to that went out, there likely wouldn't be another circuit you could plug into nearby; unlike 15 amp and 20 amp circuits, of which there are likely to be several.

Still, another point in favor for an Atmos house is that you don't need a D to A converter box (e.g. a DAC3202 or a Q-Sys) between the processor and the power amps: that's built into the multichannel amps, which themselves are part of the Atmos Connect loop.

Posted by Steve Guttag (Member # 268) on 12-24-2018, 06:51 AM:

However, with QSYS, you get much more flexibility.

As the probability analysis will show...the more things you have in a system, the greater the likelihood of failure somewhere in the system (each device having its own probability and adding into the entire chance of failure). However, in the event of failure, the severity of the problem is reduced by an amount (having 5 amplifiers instead of one makes your chance of a A failure is 5x more than 1 amplifier...but the severity of impact is better than 1/5 since a failure in the single amplifier system is 100%).

Another thing I've toyed in a QSYS multiplex is how many screens to put on a single core (with a redundant core, of course) versus 1-core per screen yet having them all on the same Q-LANs. In the event of a CORE failure, quickly dump the design of the bad CORE into another CORE (combine designs) such that any remaining COREs can back but a failed one). A CORE 110c can, theoretically, run up to 8-7.1 systems.

Posted by Leo Enticknap (Member # 534) on 12-24-2018, 10:16 AM:

It's the same probability risk assessment analysis I was taught in a fire safety class once: balancing the likelihood of a failure happening against the consequences if it does happen. The instructor gave both a 1 through 5 number, and used the example of catching a common cold being a 5/1 (very likely to happen, but only minor inconvenience if it does), and being in an airliner at 30,000 feet when a bomb goes off as a 1/5 (very unlikely to happen, but if it does you're buggered).

The Q-Sys example you gave shows just how flexible the thing is: arguably, too flexible for its own good. Do you put everything through one core that is redundant, or individual cores and buy a spare to put on the shelf? You could spend days chewing through the "what ifs" before coming up with a decision on that one.

Posted by Steve Guttag (Member # 268) on 12-24-2018, 11:01 AM:

Leo, with respect to QSYS, there is the financial one. If money isn't an object, the answer is simple, individual cores. If you aren't trying to save money, what was the advantage of the single core? Common UCI IP address?

Where QSYS can save money is you could run most any (7.1) complex on a pair of CORE 510c (one is redundant). There is plenty of horsepower there. You could have the power of QSYS for less than the cost of a typical DCinema sound processor.

The risk? 100% of the sound going out in the complex until the redundant core comes on line (10-15-seconds typical...providing it fails in a way that the redundant core "sees the primary is in a bad state"). Cores have been in use for about a decade now and in some high-profile applications (not just theatres) so they are pretty robust. But if man made it, it will break! (and if nature made it, it will break...but naturally). So it is a matter of "when" not "if." How many exhibitors would put up with such a failure every say, 5-years?

And it would be less than that since there is about a 50% probability that it would fail during down time (if running 24-hours). So you'd merely see that you were on your redundant core.

I'm going through the exercise with a customer now on individual cores per screen, one pair of cores for 9 screens and splitting the complex up so never would a core ever take down the whole complex.

I probably wouldn't have a "core on the floor" as a spare whereas you would burn away its warranty there and never know if you had a defect in the spare. Each running core can handle running more than one screen for the duration of what it would take to get a replacement or repair.

If they could get the CORE 110c to have 3 NICs (with a USB to Ethernet dongle), it would be a natural for say 6-screens or less (could handle 8 screens max). That would make QSYS quite price attractive.

Posted by Adam Martin (Member # 641) on 12-24-2018, 11:50 AM:

quote: Steve Guttag
So you'd merely see that you were on your redundant core.

Assuming that was noticed before the redundant core also failed. [evil]

Posted by Steve Guttag (Member # 268) on 12-24-2018, 04:05 PM:

Design your UCIs (or auto-mailers) right and you should know that you've moved to your spare tire or that the spare tire is flat.

Posted by Marco Giustini (Member # 4544) on 12-27-2018, 05:31 AM:

Something I don't like of those "cores" is the cooling. They feature - if not mistaken - standard small fans on the CPU and one on the chassis.

Time will come when those fans will fail. Wondering if the core just overheat or can survive without it? I believe the board inside is a standard intel-based board. Such a critical device should have - in my opinion - a number of redundant fans and PSU's.

Posted by Steve Guttag (Member # 268) on 12-27-2018, 07:23 AM:

It monitors its fan and temps so if there is a problem, you should know about it. Probably a second fan wouldn't be a bad idea (I honestly don't recall if there are multiple fans). This is on the baby core...the 110. The larger ones would have more cooling (heck, the largest ones are actual Dell servers).

I've used DSP processors in cinemas for a LONG time now. They seem to be no more likely to give grief than many analog things and never get "noisy" controls.