Help with the path of our sound

[Ryan Gallagher]

I had thought that switching audio would be a logical thing. I didn't think switching relays were all that good for that directly. I mean do you really want audio inputs and outputs to go from unloaded to loaded (and the reverse) just like that? Wouldn't that blow speakers and amps? I don't know. I don't have any audio experts on staff.

This is definitely true if you are dealing with an analog audio signal... throwing relays, flipping a switch, or repatching in a bay are all things you would want to be muted for.

But in the context of this conversation, it is switching Digital Audio (AES), and much safer relatively speaking. We've had an RJ45 AES input with poor strain relief pop loose before on our AP20... no transients, it just doesn't get the next 1 or 0 bit and stops making audio. But that is a good lesson for needing a system that doesn't involve replugging things in the back of the rack. That example was when a festival person was in there reclaiming their CCAP equipment during the last show that didn't have CCAP features. Don't be in the rack during a show! lol

[Bruce Cloutier]

Once there was a discussion about us providing some audio routing. I actually have a Behringer A AIR XR12 sitting here. It was provided as an example. I am not sure exactly what they wanted to end up with? I will be honest, we've never powered it up. That jumps right into the DSP audio processing (mixing) world. It is not a simple switch which is where I think we can provide some value cost-effectively. Not that we can't handle digital audio processing. We just don't yet use the GHz processor to do that properly. It seems to have several single channel 1/4" phone jacks, a few XLR inputs and outputs, USB, MIDI and Ethernet. The phone jacks and XLR are handling analog signals.

If we had some relays that selected A/B sources for each channel, would you want RCA jacks, XLR, BNC, other? How many channels (A/B switches)? I don't think it would matter if you were switching analog or AES so long as the sources and the destination all were for one or the other. The JNIOR could be programmed to force the amps to MUTE during switching. Or your servers can send those commands. You will have to deal with the level differences between the sources anyway.

There are some nuances. Like isolation. Grounding issues. I wouldn't want to introduce a source of instability or hum given common mode voltage differences from one side of the booth to the other. Sure we are just switching cables but I will probably need to switch all of it including the shields. Then if you don't completely disconnect from one source and connect the other you can end up interconnecting the grounds/shields and potentially cause problems (ground loops). I am just trying to be overly cautious.

[Ryan Gallagher]

Once there was a discussion about us providing some audio routing. I actually have a Behringer A AIR XR12 sitting here. It was provided as an example. I am not sure exactly what they wanted to end up with? I will be honest, we've never powered it up. That jumps right into the DSP audio processing (mixing) world. It is not a simple switch which is where I think we can provide some value cost-effectively. Not that we can't handle digital audio processing. We just don't yet use the GHz processor to do that properly. It seems to have several single channel 1/4" phone jacks, a few XLR inputs and outputs, USB, MIDI and Ethernet. The phone jacks and XLR are handling analog signals.

If we had some relays that selected A/B sources for each channel, would you want RCA jacks, XLR, BNC, other? How many channels (A/B switches)? I don't think it would matter if you were switching analog or AES so long as the sources and the destination all were for one or the other. The JNIOR could be programmed to force the amps to MUTE during switching. Or your servers can send those commands. You will have to deal with the level differences between the sources anyway.

There are some nuances. Like isolation. Grounding issues. I wouldn't want to introduce a source of instability or hum given common mode voltage differences from one side of the booth to the other. Sure we are just switching cables but I will probably need to switch all of it including the shields. Then if you don't completely disconnect from one source and connect the other you can end up interconnecting the grounds/shields and potentially cause problems (ground loops). I am just trying to be overly cautious.

Bruce, I'm not an integrator so I'm really not the one to ask, but my instinct is that in a booth this is very much an outlier situation, if you end up building/adding a rather complicated set of analog projection options and have an existing main late model DCI cinema audio processor (perhaps even in your IMS) that doesn't give you inputs suitable for those extras, but does have AES on either DB25 or RJ45. It seems to be a designer/integrator oversight in Brian's case and probably a pretty rare need? Other likely clients are film clubs trying to add 35mm on a budget without replacing their main processor, or hobby folks.

But if you were after a switching design that would help. I think an economical version would be a JNIOR box that could passively switch between 1 to 1 wired RJ45 connections, bonus points if it has enough ins and outs for 16 channels of AES3 (typically two RJ45 or two DB25, or 8 XLR). So minimum two pairs of two on the input side, one pair of two on the output side. It would not need to be a router or cross-point for the individual channels, just gang switching. It would probably need a "switch" button or remote on it too, as so many 35/70mm booths are still quite manual in analog mode, and don't rely on automation to trigger things, but integration into JNIORs automation would be great (people could make a button if they had to or utilize automation if they wanted to).

If you wanted to make the design more utility/universal, you might make it a DB25 version but offer RJ45 ports that are parallel connections. I think getting into XLR directly starts to require a full rack unit, $$$, and not the JNIOR form factor, but the DB25's could easily take existing analog or AES XLR fan ins/outs if needed. 6 DB25 might push the boundaries of the JNIOR form factor, if so maybe you need two of them, designed as 8 channel units instead (2 inputs and 1 output)? I wouldn't bother with any pin mappings, there are too many variables on the input and output side, 1 to 1 would be a universal box.

It seems like price point would need to be sub 700USD to make it really appealing over other DB25 switching automation capable options? That seems doable considering the 414 is priced at 425$

You could stick with the phenix connectors but it seems to be pretty difficult to get directly to DB25 or RJ45 with them without wiring a bare wire jump to a terminal block adapter. 2x13 header and ribbon cables seem to be an option though. 2x13 header direct to RJ45 seems a bit harder to adapt without going through DB25 adapters first, so I'd stick with RJ45 or DB25 (or both) natively. If sticking with DB25 standards, I think they would all be female on the JNIOR device?

You are likely correct there is a bit of design work to be able to market it for safe analog and digital audio switching and avoiding grounding and signal noise issues. Maybe Brian's "final" solution will warrant such a device, and he can consult with ya? Two 414s is almost there, it's just a connector type conductor count issue as is (and the audio specific design questions).

Really they are usually only feeding 6 channels from their analog processor (maybe 8 if they have the 7.1 upgrades), but the requirement to "unplug" the 9-16 bank coming from their cinema server means it needs to switch all 16 channels.

There are a lot of expensive ways to skin this cat it seems, but a JNIOR way might be a welcome addition? Even if Brian found the idealized rare CAT778 digital output card and could safely ignore all the analog to digital hardware, he would still need the AES switching to complete the solution.

[Bruce Cloutier]

Ryan, thanks for the thinking.

Is there any issue with confusion between Ethernet and AES (or any other use of RJ-45)? I think I have even seen that happen to someone in a thread here?

As for form factor I was thinking of an expansion module that would connect to the JNIOR via the Sensor Port (expansion bus). The JNIOR would command it to switch (or route) as needed. If you hadn't noticed but we have relied on Polycase for these low-volume housings (their SL series). So that gives us a fairly good range of housings to cover all of the fun with connectors. We have shifted to our own (injection molded) housings for the JNIOR models. That let us improve the DIN rail mounting option and save cost. But our 412DMX JNIOR is still in the Polycase enclosure.

Yeah, the rack mount is expensive. I have had a real headache getting metalwork done at any sane cost especially recently. Even finding shops that can actually read drawings is hard. If you can't give them 3D CAD that plugs directly into their machines (i.e. doesn't have some custom tooling requirement) they either turn the job down or quote it extremely high. I ended up with a Canadian company for the LED dimmer housings. That now has its own problems.

As for manual control. We do plan to embrace manual overrides in our new designs. The issue right now is the political and economic environment which has R&D on hold. We are not able to plan. Nor can I get a handle on costs going forward for just about anything. But lets not get people whining about that here (again). I have been fighting suggestions to increase prices. What we need is to find new customers and move more volume. Not milk more money out of our valued customers. So... adding capabilities might help.

So you are thinking an AES3 switch?

A module could have 4 RJ45 input jacks (2 for set A, 2 for set B) and 2 RJ45 output jacks. Those are physically the same but we can do something to make their purpose obvious. But that would be about the size of one of our 10V modules or maybe the 4ROUT (or 8DIN) depending on what I have to do inside there.

[Ryan Gallagher]

Ryan, thanks for the thinking.

Is there any issue with confusion between Ethernet and AES (or any other use of RJ-45)? I think I have even seen that happen to someone in a thread here?

As for form factor I was thinking of an expansion module that would connect to the JNIOR via the Sensor Port (expansion bus). The JNIOR would command it to switch (or route) as needed. If you hadn't noticed but we have relied on Polycase for these low-volume housings (their SL series). So that gives us a fairly good range of housings to cover all of the fun with connectors. We have shifted to our own (injection molded) housings for the JNIOR models. That let us improve the DIN rail mounting option and save cost. But our 412DMX JNIOR is still in the Polycase enclosure.

Yeah, the rack mount is expensive. I have had a real headache getting metalwork done at any sane cost especially recently. Even finding shops that can actually read drawings is hard. If you can't give them 3D CAD that plugs directly into their machines (i.e. doesn't have some custom tooling requirement) they either turn the job down or quote it extremely high. I ended up with a Canadian company for the LED dimmer housings. That now has its own problems.

As for manual control. We do plan to embrace manual overrides in our new designs. The issue right now is the political and economic environment which has R&D on hold. We are not able to plan. Nor can I get a handle on costs going forward for just about anything. But lets not get people whining about that here (again). I have been fighting suggestions to increase prices. What we need is to find new customers and move more volume. Not milk more money out of our valued customers. So... adding capabilities might help.

So you are thinking an AES3 switch?

A module could have 4 RJ45 input jacks (2 for set A, 2 for set B) and 2 RJ45 output jacks. Those are physically the same but we can do something to make their purpose obvious. But that would be about the size of one of our 10V modules or maybe the 4ROUT (or 8DIN) depending on what I have to do inside there.

For compactness into an expansion module the RJ45 method is definitely the most logical. There are easy ways to get that to DB25 when needed for gear on the other ends.

As for confusion with other RJ45 standards... as a passive switch, it can really do any of them (just don't mix them)... but it may fall short on certain protocols like network transport due to behaving like a dumb barrel connector with extra length in the untwisted pairs... I think the only AES nuances would be if there were particular shielding or noise management enhancements? Someone with a little more expertise would have to weigh in on what makes an AES marketed switch distinct from a passive generic or audio switch device. Sometimes they have other functions that add features like world-clock sync (generation or input) or reclocking etc. But I don't believe any of that is needed in this use case? It may not be a seamless switch as far as AES audio goes, there might be a second while the device acquires sync with the new inputs etc.

EDIT: the other confusion with RJ45 is that often for AES patching it most certainly won't be a standard Cat5 patch cable. The pin arrangement will either be 1to1 or something unique to the equipment. Our doremi cables were bright orange and labeled clearly, but custom longer ones had to be made in our booth. The Doremi ones were also of the variety that actually had a connected shield, although the installer did not use shielded cables in our integration and still works. Most of the common plastic ethernet to DB25 adapters don't do anything with the CAT cable ground/shield anyway other than to carry the DB25 ground/shield on one of the conductors.

[Bruce Cloutier]

A passive switch would use small signal relays. Basically 8 DPDT, the same used in the JNIOR, to gate inputs A or B to the output. That would have a cost advantage since we happen to use a few of them.

That would let the thing be used for any analog signalling or digital communications. Relay contacts have their own characteristics and that varies over time. I don't know if that would become an issue. The circuit design however is simple.

If we narrow it down to only an AES switch then I might look into a solid-state means to switch the signals. Or even consider interposing a processor opening the door to level adjustment, fading and general mixing. At least the easier things that can be done without the hardware DSP.

We can start with the passive switch.

I will think on it. I don't have any AES audio signals handy for testing.

[Ryan Gallagher]

A passive switch would use small signal relays. Basically 8 DPDT, the same used in the JNIOR, to gate inputs A or B to the output. That would have a cost advantage since we happen to use a few of them.

That would let the thing be used for any analog signalling or digital communications. Relay contacts have their own characteristics and that varies over time. I don't know if that would become an issue. The circuit design however is simple.

If we narrow it down to only an AES switch then I might look into a solid-state means to switch the signals. Or even consider interposing a processor opening the door to level adjustment, fading and general mixing. At least the easier things that can be done without the hardware DSP.

We can start with the passive switch.

I will think on it. I don't have any AES audio signals handy for testing.

A fair number of consumer audio interfaces can at least output on AES pairs for testing. Or if you have a DCI cinema server handy can always use test material or unencrypted DCPs (or make your own).

I think once you start adding a way to manipulate the signals you are gonna be reclocking them too I would expect. Another thought would be turning it into a matrix switch per conductor, so people could skip the custom cables and custom adapters and have JNIOR do that. Could have pre-defined software matrix presets for the common brands/schemes on the inputs and outputs, but ships as 1 to 1 on boot up.

Or a matrix switch per AES channel too... but most audio processors and DCI servers permit channel remapping per format, so not really a needed feature in our industry in another box.

I feel like adding such things moves it into a fairly narrow AES audio product, and away from a dumb, perhaps more generally desirable RJ45 switch, maybe both exist someday?

[Brian Mendelssohn]

Hi all. I wanted to give a quick late night update. After much struggle and lots of grounding issues we feel are finally resolved, the show must go on.

First, this board's knowledge is incredible and the community are have joined as we launch our film program has been incredible. We are always happy to share what we are learning, or help with any training if anyone needs.

Second, we are running Inglorious bastards tomorrow. So what is our solution............. After trying to get like 20 pieces of equipment in on time, we found this...Odyssey Products D16-3-KIT/THX Drive-In Mixer Kit. This device is built for drive-ins which inputs a DB-25 and mixes down to unbalanced stereo, aka RCA. We fed the RCAs into our non-sync, and upmixed it with our Dolby CP950 to 5.1. Its a solution, not too bad.

Third, based on discussions from this board, we felt we had three routes.

1. Separate parellel paths to our amps with a DB25 switcher (like Ryan shared). We would have to add a crossover after the CP650 to match our setup with the CP950. This would be around $1,500 in total we think.

2. CP650 -> CP950. As you can see from the discussion, its mind blowing how complicated this seemly easy task is. This requires a AD converter for 8 channels (really difficult) and go into the CP950 with AEG3 RK45 port. Our thought was setting up a 35mm or 70mm is pretty manual and requires someone to do work in the both, switching the RJ45 ports manully would be no problem and take a complicated equipment out of the process. A draw back of this one is pre-show. This one is around $2,500 in total

3. Everyones first instinct, either purchase a Datasat Ap25 or Ovation2. After going down Path 1 and 2 for the last week. I'm sucking it up and going with the Ovation2. Everyone says its a great piece of kit and we won't be disappointed. This will be around $5k in total plus install (Equalizing the room is beyond our skills)

TLDR
We decided to go with option 3 and purchase the Ovation2

Anyone want to purchase a Dolby CP950? Still has the plastic on the display and everything.

Brian
Row House Cinemas

[Ryan Gallagher]

Hi all. I wanted to give a quick late night update. After much struggle and lots of grounding issues we feel are finally resolved, the show must go on.

First, this board's knowledge is incredible and the community are have joined as we launch our film program has been incredible. We are always happy to share what we are learning, or help with any training if anyone needs.

Second, we are running Inglorious bastards tomorrow. So what is our solution............. After trying to get like 20 pieces of equipment in on time, we found this...Odyssey Products D16-3-KIT/THX Drive-In Mixer Kit. This device is built for drive-ins which inputs a DB-25 and mixes down to unbalanced stereo, aka RCA. We fed the RCAs into our non-sync, and upmixed it with our Dolby CP950 to 5.1. Its a solution, not too bad.

Third, based on discussions from this board, we felt we had three routes.

1. Separate parellel paths to our amps with a DB25 switcher (like Ryan shared). We would have to add a crossover after the CP650 to match our setup with the CP950. This would be around $1,500 in total we think.

2. CP650 -> CP950. As you can see from the discussion, its mind blowing how complicated this seemly easy task is. This requires a AD converter for 8 channels (really difficult) and go into the CP950 with AEG3 RK45 port. Our thought was setting up a 35mm or 70mm is pretty manual and requires someone to do work in the both, switching the RJ45 ports manully would be no problem and take a complicated equipment out of the process. A draw back of this one is pre-show. This one is around $2,500 in total

3. Everyones first instinct, either purchase a Datasat Ap25 or Ovation2. After going down Path 1 and 2 for the last week. I'm sucking it up and going with the Ovation2. Everyone says its a great piece of kit and we won't be disappointed. This will be around $5k in total plus install (Equalizing the room is beyond our skills)

TLDR
We decided to go with option 3 and purchase the Ovation2

Anyone want to purchase a Dolby CP950? Still has the plastic on the display and everything.

Brian
Row House Cinemas

Great to hear you managed something in the short time frame. I don't know what a mint 950 goes for, but maybe it can fund it's own replacement. Though having a spare DCI processor around is never a bad idea (especially if you do high profile events), this one even has a room tune already on it.

Option 3 seems like a perfectly reasonable choice, despite the headache. I do think the 650 -> 950 method is doable for much less than $2,500 if you are selective and are okay with 2nd hand equipment, but the Ovation2 will probably serve you better in the long run and be a much cleaner solution.

[Tj Hopland]

Glad to hear you were able to locate something for hopefully cheap that will get you through a few shows. It pays to know people how have or know where there is 'stuff' laying around.

Only feature I/we are aware of that the Ovation2 is lacking is the HDMI de embedding. It sounds like you already have a bunch of external gear related to HDMI so maybe you already have something to do the de embedding? I think as an option you can get Dolby and DTS decoding in the Ovation but I'm not sure which flavors and which ports that's used with. if you have an external decoder you could be somewhat back in the same situation again with how to elegantly get a multi channel signal in.

[Ryan Gallagher]

Only feature I/we are aware of that the Ovation2 is lacking is the HDMI de embedding. It sounds like you already have a bunch of external gear related to HDMI so maybe you already have something to do the de embedding? I think as an option you can get Dolby and DTS decoding in the Ovation but I'm not sure which flavors and which ports that's used with. if you have an external decoder you could be somewhat back in the same situation again with how to elegantly get a multi channel signal in.

The main thing the Ovation2 offers is a set of 8 balanced analog inputs, which the 950 did not, that gets the 650 talking to it... unless there is an option card for direct 35mm? It also has AES on both DB25 and RJ45, so things like a DTS processor are relatively easy to add.

But yeah no HDMI port directly. It has plenty of digital aux inputs though, SPDIF on RCA and COAX, as well as a TosLink... so BluRay is easy etc. If you want to feed it HDMI from elsewhere in the venue then yeah a disembedder might be handy, but 5.1 disembedding options to consumer connections like TosLink or SPDIF are numerous and cheap, neither TosLink or SPDIF support PCM audio, but they are good enough for the occasional blu-ray show. (although those consumer decoders appear to be an optional software license).

[Brian Mendelssohn]

My extron hdmi matrix switcher does have a S/PDIF digital audio outputs. Something new to learn and figure out. That should solve that issue for the Ovation2. I assume a jnior works with the Ovation2?

Bruce, I have a 412DMX Jnior. Will that easily talk with the Ovation2?

[Steve Guttag]

Extron's S/PDIF outputs are, typically, for 2-channel sources to provide a de-embedding. Since, as I recall, I was talking about the 4x2 when you were on the hunt, and it does not have a S/PDIF output, I don't know which one you have.

Note, taking the CP650's outputs and doing a passive mix down to 2-channels does not create an Lt/Rt mix again so I would presume that the surrounds are completely lost but L/C/R should get back to something resembling their old selves. If you are feeling really hacky, you could flip down the CP650's cover on the front and pick up the Lt/Rt mix there.

image.png

There wouldn't be any noise reduction there but you'd have the 4 channels matrix encoded.

Using the normal CP650 outputs, you could use an external mixer to also re-recreate the Lt/Rt mix by summing Ls/Rs and then mixing it back into L/R by inverting the signal going to Right.

But you might leave well enough alone to get through the opening event (tonight).

I think you'll not regret you decision to switch to the OV2. As for tuning it, one of its hallmarks (in fact, one of its biggest selling points) is that it tunes itself and does so at a level well beyond what a CP950 can do.

After you have lived with it for a month or two, let us know what you think. I don't think you'll have a problem off-loading the CP950. It is perfectly suitable to replace the 10s of thousands of CP750s that are dropping right now (suitable once one has the adapter boards/cables because Dolby didn't match the inputs/outputs...why would they?).

[Brian Mendelssohn]

Good memory Steve and nice to hear from you. I ended up getting a 4x4 matrix, and I'm glad I did, I ended up using three outputs. If the Extron de-imbed is only stereo, then I'm on the hunt for an external device. I know Extron makes one (according to their website it is no longer supported of course), the HAE 100 4K Plus.

[Steve Guttag]

If you are looking for an HDMI audio decoder, the Extron SSP200 is a professional decoder.

https://www.extron.com/product/ssp200

It is brand agnostic so it can handle PCM, DTS or Dolby and all of the way up to immersive. As with most Extron, it's not cheap but nothing for commercial really is. If all you want is to knock HDMI down to 2-channel, regardless of its encode, Atlona makes a piece I use for preview monitors. AT-HD-M2C.

https://atlona.com/product/at-hdr-m2c/

I don't think it makes an Lt/Rt mix...just 2-channel so a conventional 2-channel TV can play it. But, truthfully, I've never paid attention to it. I just wanted 2-channel for a preview monitor (e.g. small TV). It handles from DTS-HD/Dolby True HD on down (so not the immersives).