Bid review round two: the large rooms, the touch screens, and the USB switcher

Lightly edited for readability; the issuing organization is left unnamed here.

Hello everybody - round two of the same A/V bid. In the first round I got stuck right where they were dictating the Neat Board Pro in the standard conference rooms, and in one room specifically it wanted a local laptop, an optional laptop, a resident PC, and BYOD. That was just a lot. Oh, and not just that - a USB document camera too.

I got to thinking: one thing I would definitely do, if I had the option and it was greenfield, is switch platforms away from the Neat Board Pro. The way I'd incorporate all those devices is with something like a Maxhub, because that will run Teams Rooms on Windows - or a Yealink. Microsoft Surface Hubs are ridiculously expensive; they start at something like $20,000. Maxhubs have come a long way, and they've got this really unique thing: an NDI video receiver built in. You can add an NDI camera and have it injected into the meeting, and that just comes out of the box. I think that's pretty cool.

On to the large conference rooms. All three have integrated Poly systems that support local presentation, Microsoft Teams Rooms on Android, and multi-platform UC video conferencing via BYOD hosting. Teams calls are the primary video application, but the system as designed will allow for everything else.

Can I just say that this concept of wanting to support everything else really throws wrenches into things? I know you want to - but how often are those other platforms actually used? Is it 50%? 25%? 10%? Do you want to add that complexity and reduce the reliability for something that's used 10% of the time? Come up with a cutoff. If it's only going to be used 10% of the time or less, just say no: we don't support Webex or RingCentral. You can use Microsoft Teams or you can use Zoom - and let me show you the best way to support those.

If you can, always go with either a Zoom Room or a Teams Room, and then bolt on something like Pexip. Pexip is a connector, and there's a lot of value in it. You still get the native Teams Room or Zoom Room experience, especially with dual screens. Teams Rooms and Zoom Rooms have interoperability built in, but it doesn't work very well with multiple cameras, multiple displays, or screen sharing. With Pexip, a Teams Room can join a Zoom call with dual screens, full control of your cameras and room controls, and a native experience from Teams to Zoom. Same thing in the other direction - a Zoom Room connecting to Teams meetings - plus Google Meet and Webex. That's the right way to do interoperability: not bolted on through the hardware, but a native room made compatible with the other services, with a more native experience. That's how you get reliability and simplicity, and it's just easier for the end user to remember how to do this.

Two of the rooms have 98-inch interactive multi-touch monitors. Holy smoke, those are huge. Why do you need touch screen monitors that big? They'll also be used as digital whiteboards via OneNote and PowerPoint - and they'll display content from far-end participants on UC calls. So they're multi-purpose: either they're showing content from the call, or you switch their mode into a whiteboard. How is a user going to think about that? The average non-technical user? Let's say I start sharing content through the Poly, and I know that thing's a touch screen, and I've seen people annotate on there. I'm going to walk up and start drawing on it. That's the expectation. I don't know about that - I'd have to think about it some more.

Let's look at the first of those rooms. A 75-inch monitor over here, the 98-inch touch in the middle, and another 75-inch over there - presumably mirroring the 98. A local HDMI input plate with press feed and duplex power, tightly grouped; the 98-inch touch monitor with a Poly Studio E70 and a PAC-1 wall box.

The room is divisible, with HDMI inputs on the sides as well as the center. When it's divided, you're probably not going to have someone presenting from the side walls - the center is always going to be the centerpiece. So those side inputs are nice to have, but how often will they get used? If it's very rare, eliminate them, and for the odd time someone does need to share from a side angle, do a wireless screen cast into your MTR room.

Back to the spec: the 98-inch interactive monitor is the main display. Here's what I know you can do: plug a Windows-based MTR into a touch screen and it gives you whiteboard capabilities. So I would use something like a Lenovo MTR that has an HDMI ingest right on the codec. Those are really nice because you can choose where the HDMI input lives - at the touch screen or at the codec. A lot of MTR systems put the HDMI input only at the touch screen, which means a room PC has to sit out at the desk, and all the HDMI switching and auto-switching has to happen near the touch screen. This device lets you take HDMI ingest at either end: the touch screen connects over a Cat 6 run - proprietary and length-limited, not IP, so not switchable - to a breakout box near the codec, which exposes USB to the codec for touch and HDMI into the codec's input. Get the 98-inch touch screen, plug its USB into the MTR, and it becomes a touchscreen MTR with whiteboard capabilities.

Anytime a display is this multi-purpose - OneNote whiteboarding and the conferencing display - it's going to cause confusion, and confusion causes these things not to get used. That's not a good use of taxpayers' money. You see it all the time: they start gathering dust. Another way to do this: make the 98-inch a stationary, non-touch Teams Rooms display, and pair it with a touch device using what Microsoft calls coordinated meetings. I haven't used it personally, and I couldn't find many videos of how it actually works, but you should be able to take a Surface Hub on a cart, wheel it into the room, and join the two devices together - so for the infrequent times you'd use the touch screen in that room anyway, you have a dedicated device that does whiteboarding really well.

In addition to the Poly E70 main camera, a second Poly Studio E60 PTZ camera is provided as a presenter camera - rear of room, facing forward. The rooms get both USB host and device switching and HDMI content matrix switching, supporting the resident mini PC and BYOD laptops. The USB switchers drop the VBUS signal as part of the design, and they're controllable via the Extron or the Poly TC10 control system. The spec says these features let the Poly G62 codec operate within its designed behavior to support multiple hosts - it runs appliance mode with native software like Microsoft Teams Rooms or Zoom Rooms. Are they saying they want to switch between Zoom and Teams? I'm pretty sure that's a factory reset and a long process, five or ten minutes, to switch modes. Am I wrong? Tell me in the comments.

Let's look at the schematic. The E70; the E60 for the back of the room; the TC10 - good. Local HDMI inputs; the portable laptop; the Extron USB switcher. The G62 takes both outputs to an HDMI matrix switcher - good, as it should. USB out from the DSP to the G62 - good. And here's the interesting part: USB touch one and two from the 98-inch touchscreen to the G62, with a secondary output to the Extron. I'm not saying this isn't going to work. But what I see is some challenges. If the Extron - the USB switcher - isn't abstracting the USB signals, and it's just doing switching, that's going to be a problem. USB switching should work all the time, but it doesn't.

If you can abstract a signal, you can effectively split USB. Inogeni does this - I think they do it pretty well. Say you have a mouse. The computer has to negotiate with it: what kind of mouse, how many buttons - there's two-way communication, and as the mouse moves it sends signals back. Now say you want that mouse across multiple devices. The abstraction is a splitter - a multiplexer - that acts as the computer to the mouse: it negotiates the buttons and the scroll wheel and the features. Then when you plug laptop one into the splitter, the splitter presents itself as that mouse, and it does the same for device two, replicating the mouse's signals out to both. It's an intermediary - a virtual mouse. You can't do that for every USB device, but you can for some: mice, speakerphones, touch screens would be relatively easy. That's the most reliable way to share a single touch screen with multiple devices.

If you're doing USB switching instead, every switch means each device's USB software stack has to reset. As I've said in previous videos, that just isn't reliable. You end up doing workarounds - shutting devices off and back on. Say this 98-inch touchscreen goes through a fast USB cycle and its internal software stack gets stuck in the middle. The only way to unstick it is to power cycle the device. Sometimes unplugging the USB and waiting a while does it - the device detects the electrical disconnect and resets the stack - but otherwise the thing sits in limbo. So you could build a workaround that power cycles it - and now you have to think about where that power is. Is it on a PoE switch? Is it locally powered? Do you need a PDU for that? How do you power cycle this thing? I'm not saying it's wrong. I'm not saying it won't work. I'm skeptical of the reliability.

And why am I so skeptical, and why does that matter - when you can just call, create a service ticket, and someone will come fix it? Because typical non-technical users fear technology, and they don't like using something that isn't reliable. I think of technology and features in this priority order. It has to be reliable: it has to work within a few seconds, 99% of the time. Second, it has to be easy to use. Third, it has to have the right features. Ease of use means the absolute minimum the end user has to remember - no different modes to configure, no decision trees, no being locked into certain states. That's too much for the average non-technical user. I can troubleshoot a lockup just fine, but I've worked with users who, in the heat of the moment, can't comprehend the first step of troubleshooting. They literally lock up. And that's not their fault - it's just human behavior. They're non-technical; they're probably great at other things. But troubleshooting? No. So they build up trauma, or they just stop using the system. I think that's the majority of users. What you find is very technical people designing these systems for all these use cases - and a technical person could figure it out, put instructions on the wall, a reset procedure, a support number. But the usage you actually get out of these rooms is a fraction of what they offer, unless you eliminate some features, streamline the workflow, reduce how much people have to think - and just don't say yes to everything.

The fourth priority is meeting transparency: cameras at eye level; really clear audio with no background sound, where you can hear a whisper; cameras picking up facial expressions and non-verbals; remote audio that's natural rather than booming, coming from the display - because the faces are around the display, not in the ceiling. And the fifth is room health: CO2 levels - humans put out a lot of CO2, it's easy to raise it in a room, and then you get drowsy, memory suffers, creativity suffers - and acoustics, because a reverberant room puts extra cognitive load on your brain. Between high CO2 and listening through reverberation, you can come out of an hour-long meeting drained and not know why.

So that's my why. Again - I'm not saying this won't work. I'm critical of it because of how I know non-technical users will likely react to this type of system. Lots of great features, lots of bells and whistles. But if it's not reliable, the best feature just doesn't exist. And if it's not easy to use, the best feature just doesn't exist. Thanks for watching.