There are other hardware projects I’ve got cooking, yes, but none so immediately useful or easily implemented. The great thing about this is the hard work (read:software) is essentially done already. At worst I’d need to hack some kind of interface driver, but, really, it’s pretty much functional as is. And, because it’s already been ported to Linux and BSD, it’s really trivial to build a barebones system upon which to base it all. Preserving keyboard + mouse input is a totally necessary design parameter anyway (for reasons of network stream integration, titling(!!), etc.) so hardware can be periodically switched on as it becomes available. I’m tempted to pull apart my languishing Athlon XP, but it feels too powerful for the task (not even kidding… this thing is lightning fast) and I wouldn’t know what to do with the rest of the RAM in it. My biggest concern is tracking down capture hardware that’s Linux or BSD friendly. Ideally there’ll be a security capture card that does PAL at full frame rate and has 4 inputs, because essentially that means it’d be trivial to add a few extra cards and, all of a sudden, it’s quite foreseeable to have a 12 input vision mixer that will key and title away til the cows come home.

One concern I have is that the mixer component only takes two sources… which is much the same as on any hardware mixers I’ve used (two buses: select source on A + B bus, mix buses), but it feels really inflexible. I’d chain them together but think that might necessitate extra genlocking time and increase overall latency. I can’t actually think of a usage scenario for this one, though, so it’s not a big deal. Because keying exists independently of mixing it’s not a concern of 2 sources + keyed source, and that’d be the main situation in which such a thing would be at all necessary.

The other cool thing about this is you can mix digital and analogue sources with impunity. Need SDI? Sure, get an SDI capture card and add an input source. Firewire? Done deal. Same goes for output: because you can output via FFMPEG, your “vision mixer” potentially also encodes an IP-distributable stream simultaneously with realtime output to a monitor.

This is an inestimably cool piece of software, but the most brilliant thing is it isn’t really anything new. I discovered it because I was looking for EffecTV which I’d last used in a production context over 12 months ago… Gephex uses existing open-source filters and processing solutions and just provides an excellent means of chaining them together. You can create some excellent motion artwork with it, but the most exciting thing for me is that it enables use of cheap and disposable x86 hardware in place of hideously expensive and proprietary (read: more expensive, but also inextensible and not particularly flexible) solutions that the ‘pros’ use.

Increasingly I’m disinterested in ‘professionalism’ about this sort of thing, because that’s way out of my price league and, to be honest, the most common place I wish this technology were applied is in church and Christian event contexts, where (even if there is money) no-one is interested in effective communication through applied technology. So we continue to try and push forward with no money and a bunch of innovative and irreverent (to the pros) solutions.

Ultimately, it’s about achieving excellence in the quality and nature of the work done to share the gospel and build up the body of those who follow Jesus — but excellence can be attained without even a smattering of ‘professionalism’.

That said, I’d still love to own an MX-70.

The grain wasn’t really a problem anyway, because most of the vision was intentionally grungy/the input was only a seed for further digital genesis, so that was cool. I’ve got to get the composite input for this computer working before I do another one of these things or get a real capture card, though, because it’s a pain having only one camera. Usually I’d flick it around to a light source bright enough that everything would flare out and then repo, or switch to a heavily distorted effect, or slide my finger over it (it’s a webcam, okay?) to black. The vision mixer didn’t get a workout because I couldn’t get TV-out working properly in time (mostly because effecTV’s fullscreen mode did weird things to sync on composite outputs, but hey, VGA worked).

I’ve really got to learn how to use PureData before next time, too. It looks so incredibly powerful, wow. I don’t understand where it outputs sources/runs, though. Looks great for routing and filtering stuff to the hilt, but where that all ends up is beyond me! Again, hopefully before next time.

It just occurred to me that pretty much every camera with IEEE1394 on the planet can be used as a V4L source. D’oh. Maybe I should finally buy a real camera instead of some boring capture hardware. Capture hardware doesn’t NEED to be boring, if you’re in a Windows environment and need a device that does real-time effects… but I’m not $2500 enthusiastic about all this just yet. And probably won’t be until it starts to have some abstract kind of earning potential… which I’m pretty okay without, but nevertheless, it’s difficult to justify that going into a hole of depreciation.

So… for next time, more cameras mixed (with vision mixer) into capture device, more cameras using FireWire, more effects custom-designed using PureData, and digital vision mixing with FreeJ. I’m happy with VGA output because you’re pushing native resolution to the projector that way. I’m also not too concerned about PC resources/performance. Tonight I’d say it was online for 3.5 hours without a glitch (with the exception of a kick-out 10 minutes in, and I was still partially setting up and wasn’t particularly in the middle of anything), running from a USB source. There used to be a time when that would just not have been possible with consumer gear (the camera cost about $90–110, which means it’s a rather-decent webcam. Probably, at this stage, one of the more expensive components of my setup!)

Oh, and I didn’t get to record any of the performance, sorry. Gem recorded another person’s performance for me earlier in the evening, but I’m too tired to share that right now… probably tomorrow, before Selo comes to my house and kicks my dog… err I mean camera’s SD card memory. I gotta do it before this month ends, because I’m now hilariously over quota. It’s got to be all because uploads count, seriously! I’ve been doing nearly 1GB a month to Flickr alone. Anyway. Stopping rambling and starting sleeping.