Let's help SageTV do multizone sync on the HD200

[MrFusion]

Using a matrix switch (possibly coupled with HDMI over CAT5/6 extenders) is probably the "best" we can get in the near future (certainly the most achievable with sage and hd200s).

Even then, there are going to be problems - differerent receivers / TVs decode & playback with different latencies.

This will hurt where the devices are within earshot/sightlines of each other.

But even if you manage to get two devices perfectly in sync, but - say - at different ends of the house, you'll still perceive a difference (sort of a simple reverb) due simply to the relative distance to each...

So.... how perfect does this have to be?

[sandor]

Originally Posted by MrFusion

So.... how perfect does this have to be?

I would say as perfect as Apple's Airport Express does audio.

http://www.apple.com/airportexpress/.../airtunes.html

I've got three last generation (802.11/g) airport express' that i picked up for $20 a piece, and then iTunes running on my Sage server. All four locations are controlled by the "Remote" app on my iPhone, and the synced audio is close enough that no one notices.

[Narflex]

I did an experiment with this one time and got decent results. It was pretty simple actually. Since you can make SageTV API calls on any extender client from the server; I just sent Seek() calls to multiple HD200s at the same time (same time isn't exactly true; it depends upon the # of cores on the CPU and thread switching and network resource sharing). They were pretty close IIRC; probably <50ms difference between them. And I think with some seek/pause/play logic I might have even gotten them to the point where the difference was imperceptible. (it was awhile ago when I did this test)

I'm not saying that this would work as a solution; but its an interesting experiment to try. With other predictive measure you may actually be able to get it to a point where the synchronization is tight enough to satisfy certain requirements.

There'd probably still be clock drift issues though; but you may be able to account for these as well (although fixing them would definitely require a hiccup in the video to resync things).

[Fuzzy]

Thanks Jeff, for the comment. As mentioned earlier, the difficulty really just comes down to how accurate you want it to be. <100 ms would be enough to keep the crowd reactions down, say at a sports party or something... <50 ms would probably be enough to keep audio from echoing, and be closer to just a bit of reverb... really, if you've got something playing in every room of the house, you probably shouldn't have them turned up so loud that they override the room next door anyways.

[tchapin]

Since only my living room and kitchen could really hear each other, I picked up one of these (http://www.audioengineusa.com/w1_home.php), which has been great. I connected a pair of powered speakers in my kitchen and have been really happy. I also have a pair of their unpowered speakers in my living room, and they sound great.

Hopefully Project Ortus will improve Sage's music flexibility!

[rmac321]

Just as an added illustration of the difficulty in keeping sync, even the HD200 sometimes gets out of sync between its own outputs. I feed the analog audio into a distribution amp and then into the Liv Rm TV and my A-Bus system so we can listen in any room. (I distribute the picture over coax for now) I feed the digital audio into the receiver in the Liv Rm. I frequently get a slight out-of sync between the analog and digital audio outputs after pausing or after a commercial skip. A quick start and stop usually resolves the problem.

[exdirtfarmer]

Originally Posted by Narflex

I did an experiment with this one time and got decent results. It was pretty simple actually. Since you can make SageTV API calls on any extender client from the server; I just sent Seek() calls to multiple HD200s at the same time (same time isn't exactly true; it depends upon the # of cores on the CPU and thread switching and network resource sharing). They were pretty close IIRC; probably <50ms difference between them. And I think with some seek/pause/play logic I might have even gotten them to the point where the difference was imperceptible. (it was awhile ago when I did this test)

I'm not saying that this would work as a solution; but its an interesting experiment to try. With other predictive measure you may actually be able to get it to a point where the synchronization is tight enough to satisfy certain requirements.

There'd probably still be clock drift issues though; but you may be able to account for these as well (although fixing them would definitely require a hiccup in the video to resync things).

I think that some more research into this would be worthwhile. There is NOBODY out there able to sync video to multiple players on IP. Please bring it up at one of your developers meetings and see if some resources could be allocated into further testing. HD video may require a solid LAN, and probably QOS, but it would be super cool if you could get it to work.

Getting this into some type of supportable feature would be worth some serious $$$$$, IMHO.

[mr_lore]

I cant wrap my head around the tech but I have always thought this would be a KILLER APP! I remember way way back in the hay day of winamp someone made a plugin to sync audio across a LAN and when it worked it was like magic, when it worked...