Boygrouping network performance

Hello
I’m boygrouping 8 (eight) pc’s for sync video playback. The videos should be 1024x768@50fps, but after trying many combinations of container (wmv, qt, mp4, mpeg2) codec (h264) and bitrate, I had to fallback on 25 fps, otherwise the system will not sync (it even will not reposition consistently).
The pc’s are commodity Lenovo laptops with Corei5, but each single video plays smooth and fine on them, even in Windows Media Player (win 7 64bit).
Could it be that I need a gigabit switch to do the task? I’m using a 100mb switch right now.
Thank you

use img sequence

have you tried a code which has only keyframes? like mjpg codecs?
the container format should not matter, only the codec is important and how well it reacts to changes of the directshow clock.

Thanks for your answers!
@antokhio
I would like to, but I don’t know how to add lipsync audio… In fact, I don’t know how to do the whole thing: is it enough to boygroup the player(timebased) node, or do I have to add the clock node also?
@tonfilm
It’s just that seeing how good it was on a (25fps) wmv, I didn’t want to pay for a PicVideo MJPEG license, and also I don’t know how MJPEG works on progressive video…
What about 100 Mbps Ethernet: is it enough to synchronize 8 streams (plus the server)?
Thanks again

i am afraid that it’s not an issue of the network performance, but rather a directshow clock problem when it runs with 50fps. what kind of setting do you have on the Wait For Frame at the video texture node? this setting influences the vvvv mainloop.
and what is the setting of your screens, also 50Hz and are you sure that they all run at 50Hz and there is no conversion from 60 to 50 somewhere in the signal chain?

an optimal setting would be if all runs at 50Hz and the wait for frame setting is 1. but you could also try 0 so the vvvv mainloop runs freely.

hey m8, with sequnce is dead easy, send timing from ur sound node to boygroupped pc’s and do seconds to frame number conversion on each