We tend to think of sl in terms of the inworld experience, but there are many things that we can do, based around sl, that expand the virtual world beyond the boundaries and constraints that ‘inworld’ imposes – effectively expanding sl beyond the confines of the viewer. These range from the simple – like this blog – to perhaps the most complex and demanding boundary-crossing activity of them all: machinima.
There can’t be many of us who, at some time or other, haven’t dabbled with at least the idea of making an sl video. Similarly, i’m guessing that a large number of us who have given it a whirl have discovered there’s far more to even a simple screen-captured short, than we ever imagined.
It seems a straightforward enough concept – login to sl, fire up Fraps, or your screen capture software of choice, film your blockbuster footage, then publish it online for the world to marvel over. Only it’s not quite that simple.
For a start, decent machinima makes some pretty hefty demands on your hardware. Using video capture software is processor and memory-intensive, particularly when you happen to be running it in conjunction with the sl client, (or multiple clients, if you’re filming your own alts concurrently), which can also be extremely demanding. You’re only going to make matters worse if you want sl to look its best – cranking up your graphics settings is just going to pile on the pressure! That’s not to mention a whole bunch of other hardware considerations we need to consider if we’re going to get serious about the business – video capture is notorious for both texture and disk thrashing, both of which can result in dro ped fram s, st-st-st-stuttering, and jumps – not to mention the joys of prematurely ageing your hardware, through sheer hard work.
Experienced machinimists, plan every scene carefully, fine-tuning draw distances, visible textures, HUD attachments and debug settings; they shut down anything that steals precious memory, and have a dedicated, empty and constantly de-fragged hard disk, specifically for video.
Technical aspects aside, decent machinima requires a high degree of manual dexterity, hand-eye co-ordination and an eye for detail and composition. Those seemingly simple zooms and pans are anything but, when you’re trying to achieve seamless transitions and capture a scene so that it looks just perfect – and you need oodles of patience. Many of the top-notch machinimists forego the mouse and keyboard entirely, in favour of a 3D device, such as a Space Navigator… but they’re not cheap and they take an entirely different approach towards navigating sl, one that doesn’t come naturally at first. You can, of course. opt for a cheaper joystick – but, in my experience, they’re a complete nightmare to use effectively and setting them up requires a degree in applied maths – on the plus side, you can experience sl from some rather interesting angles… a familiar place viewed upside-down is always fun!
Making a movie is never as simple as just grabbing a screenshot and publishing it to Youtube – there are plenty of those attempts out there, and they’re never particularly satisfying. Films in rl require a significant investment in post-production and a huge amount of time, (i know: i’ve made a good few): sl movies are no different, as anyone who has ever made a proper one will tell you. They take over your life, deprive you of sleep and drive you completely round the bend! Typically, you are camera-operator, sound recordist, director, costumier, gaffer, lighting cameraman, foley artist, choreographer, Grip, actor/s, editor, special effects supervisor and producer, all rolled into one – and it can be damn hard work.
Despite all this, most of us will give it a go – some will continue to do so, even though the results are, at best, pretty poor – a few will achieve machinimatic greatness, but most of us will realise our own limitations and those of our equipment, and quietly give up gracefully.
However, times are changing – Oculus Rift offers a radically different angle to machinima-production, since it quite literally allows a scene to be seen from a director’s point of view, and without any fiddling about with Space Navigators or flycam. There are also some pretty powerful scripted tools now available to virtual film-makers – such as ReScene – (although definitely not for the faint-hearted!) One of the biggest things to happen in the machinima world of late, however, happens to be entirely thanks to Linden Lab, with some technical changes mentioned – albeit it tantalisingly briefly – by Oz Linden in the recent Firestorm Q&A about the future of sl.
Almost buried away at around 40 minutes into the video is some very exciting news indeed: sl has never adequately employed the capabilities of OpenGL – a bug that LL have always known about. The solution, for years, has been to set a cap on VRAM usage – essentially, you could have the most powerful graphics card on the market, with enough video memory to blow any graphics-intensive software out of the water, but you’ll have got no better performance than you would from a cheap and cheerful card from your local catalogue retailer. SL would never permit access to all that additional graphics memory – which is why you may have seen blistering performance with WoW and other online games, but sl would perform like your granny on a Sunday afternoon when Antiques Roadshow is on. All that is about to change though – the Lab have finally got around to fixing the problem, which means we’ll finally be able to get to all that latent memory and, at last, we can unleash the full power of our graphic cards inworld.
This is great news for all of us, but in particular, to anyone with an interest in making virtual films… in fact, it’s such good news, i may even give it another bash myself!
Now she walks
through her sunken dream
To the seat with the clearest view
And she’s hooked to the silver screen
David Bowie – Life on Mars