Originally Posted by Joe Fernand
All of your ideas/suggestion would be great if we were still in the analog non-copy protected world.
Yeah, don't I know it! In the analog world, doing what I was originally describing would have been relatively straightforward.
You can’t have a unit which is sometimes a Switch, sometimes a Splitter, sometimes a Matrix and sometimes a Preview device – you either have a single Output Switch or you open up a whole world of HDCP and EDID complexity and create a Matrix with Passthru ports, /Mirrored ports, preview port etc. HDCP management, EDID management,
I realize that. It's probably why there seem to be so very few HDMI matrix switches on the market. Almost all of them seem to be maxtrix splitter/switches. They allow sources to handshake with all the devices attached to the outputs, regardless of whether they have that source selected. (I know. . .it's by design, probably to keep all the active devices on the HDCP network properly registered with one another.)
From the discussion in this thread, it seems to be an issue with all
HDMI matrix switch/splitters (except maybe for some extraordinarily pricey ones I see that have builtin scalers and such). I just figured that since many people already find matrix switch/splitters to be acceptably functional (as evidenced by the fact that both amateurs and installers continue to buy boatloads of non-top-end models), that their capabilities would be adequate for my suggested specifications (in which playback was generally intended for only a single zone at a time [the playback systems in other zones would be logically disconnected and not pulling signal]; but, when mirrored in multiple zones for temporarily monitoring during set up, high-def audio or video would not be required).
I came to understand that even that is asking too much of a matrix splitter/switch. (Now I'm not really even sure what they can
do, if apparently they can't even be relied on to facilitate good handshaking between a single
source and single
display and subpar handshaking when more than one display is involved.) That is why I eliminated splitters and mirroring functionality from the description. That, and to avoid the issues of bridging two or more HD networks. (You may have missed my post from a few days ago.)
In what I last described, there is a static 1:1 correspondence between the inputs and the passthough outputs. In that respect, under normal operation (that is, without anything going to an external zone) it's just like having single cables between the sources and the inputs on the AVR. However, I want it to operate as a selector switch that can, when desired, override any one of those connections to re-route the source signal instead
) to an external zone. The idea still being that, even if someone in the main zone is using a source (for example, the game console), someone in zone 2 will remotely be able to access one of the other sources (for example, a BD jukebox) physically located in the main room. (Of course, if Zone 2 accidentally selects the game console [provided that it is one of the sources that goes through the switch], there will be screaming in the household. Such is life.) Under these conditions, HDCP and EDID should not be an issue because the selected source device will no longer be logically connected to the main zone's HDCP network, nor would any other main-room devices be connected to the external zone's network. Everything handshakes with what it's connected to, and doesn't with what it isn't.
Sure, it could be done with 4 A/B switches and a 4x1 splitter, but that wouldn't prevent the user from selecting output B for more than one source. (In this scenario, the average user might have difficulty troubleshooting connection issues unless she has a good picture in her mind of just how it operates. I could troubleshoot it, but I wouldn't expect my kids to understand it clearly enough to know that a switch is set wrong. [Well, actually, my kids seem to figure out such things pretty quickly. But maybe not Grandma.])
Alternatively, it could also be done with 4 A/B switches and a 4-to-1 selector switch, but that would have many of the same problems and also require correctly setting two
switches each time. (I suppose that operation could be simplified if the A/B switches were autoswitches that gave priority to output B and the selector switch was a 5-to-1 so that there could be a "none" choice. But, apart from being more costly, autoswitches are often finicky, and can be especially
so if they must be relied upon to detect, maintain, and properly release an HDMI connection.)
For that matter, it could be done by plugging & unplugging cables, but HDMI connectors are not intended for that (and it requires physical access to the tangle of cords). Or it could be done by hauling the source device into the other room. IMO, home theater owners would not wish to sacrifice that much convenience in order to gain versatility.
Having a switch that can selectively override/redirect any one (or none) of the connections with a single button or command is much cleaner. [Grandma could handle that.] As long as there's no mirroring, I'm not seeing a technical challenge as far as HDCP goes. It seems to me to be just an issue of integrating the switching into a single unit.
As I mentioned in my previous post, though, one unwanted side effect is that, when Zone 2 is done with a source and releases it back to the Main Zone, most AVRs will probably try to switch to it when the source tries to re-handshake (if still powered on, that is). . .causing some interruption [again, potentially eliciting screams from an unhappy gamer] and the annoying need to change the AVR's input selection back to what it was on. Thus, it would probably be best in most cases when done using a source remotely, to turn it off before switching away from it.
Control system integration…
That is where I see the most potential for complication. If the switch is just being used to redirect a source to, for instance, a device to feed the signal over IP, all control could be local (press a button on the switch itself or an IR remote). But that would be unsuitable if what you want to do is control both the switch and a source device from another zone. For that kind of control, such a switch would likely have to support RS-232 control [as much as I hate long-run serial], some sort of smart control via HDCP, IP commands over LAN (or over HDbaseT or HDMI with Ethernet), or RF control, etc. Additionally, it would probably need an IR blaster/repeater with emitter ports. That doesn't necessarily mean great expense; you can find examples of all those control methods on even some really cheap
But, just because such control issues have already been solved many times before, does not mean that it can be modified cheaply for integration into a new product. (However, given the number of quick-to-market, back-alley-manufactured switches out there with such capabilities, I strongly suspect that development costs are not that high. The chips and code are apparently relatively easy to come by.) Another option, of course, would be to have separate control gear, rather than integrated into the switch itself; such equipment is currently available off the shelf. The complication with control is that there are so many options and no single "most preferred" method among them.