Hi, I’m realatively new to vvvv but am adopting it seriously into my instalation work here in New Zealand. I am planning on building a boy(group) and would like to know what to put in the boxes.
I plan to have 8 PC’s with a total of 16 outputs. I want to be able to run full res video on all 16 channels, or split video from dv/analoge on each pc to all screens, effects etc. The array would really be a medium for future development. Here is the spec for the machines:
Does this sound like overkill, or is this the minimum?
As for the live video side, with boygroups I understand that you need the same media on all pc’s that are going to show it on their respective screens. I assume the same applies for live video input, If I split the output of a video feed to all 8 pc’s, they all have video capture cards, could I span one image across all 16 like a videowall controller?
Also, I chose a raptor because of the speed needed to send two high res video feeds to the dual outputs, is this the best way to go?
Basically this sounds like a nice setup. Could you clarify:
a) what does “high resoolution video” mean? HD? or just PAL? Running 2 HD streams is doable, but it may cost too much CPU to do anything else on the side. Running 2 DV PAL streams is comparqatively easy and straightforward. We are currently researching ways of improving video playback performance especially to make video frame rates very solid.
If you plan to use a boygroup as a vidiwall controller, be aware that the capture-playback route is usually not genlocked. This may or may not be a problem for you. It means that not all images will be in perfect sync - they might be half a frame ahead or behind. Genlocking graphics cards are usually really expensive.
Check out the threads on dual screen display modes to find out more about the joy of graphics cards tweaking.
you’ll notice that playing back 2 plain DV files is not much of a challenge for a contemporary computer. But performance issues arise when you need deinterlacing or more advanced effects. A few hints:
2 Heads in Span Mode (graphics card setting) allows you to create one renderer with video streams overlapping the border between the two screens. This is not possible in Dual Mode.
Latest Graphics cards /drivers from ATI (codenamed DominATIon) promised to do very high quality deinterlacing. But I haven’t tested that, ati drivers usually are a pain, and I’d be intersted to hear what NVidia has to say to that issue.
of course you can do the deinterlacing in preproduction with tools like VirtualDub, Avisynth, or any video editor.
Tonfilm did a very nifty patch that corrects synchronisation problems when playsing back videos on a boygroup. tonfilmsModules
yes, it won’t be hard to feed a Video into 16 machines at once and simply scale the viewport of each client in the boygroup according to its clientID. As i said before, you may experience half-frame differences if your hardware isn’t genlocked.
