vog (a BETA videoblog)

I'd forgotten about this one and it was a nice surprise to stumble upon it as I've been rebuilding all the vogs into this new site. The original video is half of this size and I've simply scaled it up here, gone are the days when 194 x 144 was big - well it wasn't in 2000 either but part of my project with these very early works was for the video to be viable on existing networks and there simply was insufficient bandwidth to distribute video at this sort of size back then. It would work into universities, to some early adopters in major cities, but outside of there a video at this sort of resolution would take a silly amount of time for all its data to arrive.

While this vog has no text track, no sprites, it is on the face of it a very ordinary bit of video, I deliberately experimented with filming something that should compress very well and then tried various tricks to generate significant compression artefacts, and a small file size. Work like this compresses well as there is little movement frame to frame, and so is ideal as compression relies on frame differencing - what changes from frame to frame. When you have no cuts, no changes of scene, and a constant background it is easy to get high quality compression with low data rates. So what I did here, and it took quite a bit of experimenting to figure it out, was to export a version at a relatively low frame rate, with a low data rate, and then compress this a second time and doubling the frame rate (for this vog it is 30 fps) while trying to keep a really low data rate. Basically modern codecs (this is compressed with Sorenson Pro which was the codec for several years back then, outstanding quality for low data rates) don't let you just set a really low data rate to make the work look really bad. If it can't do it at the data rate you specify then it just takes more, and it doesn't help that Sorenson is a variable bit rate codec.

So, this was my first effort at exploring compression and its artefacts as a material and formal property of online video, as something not to be hidden behind high data rates but embraced, much like being able to see the brush strokes in a painting. The commentary is, I reckon, still a valuable riff using Deleuze's faciality to think about faces on screens, video within web browses, and the like. Below is what I originally wrote to accompany this vog:

I've been here for 6 weeks and the temperature has, but for one day, been unwaveringly steady between 6 and 10 degrees. Every day. Every night. . .

Video on a web page is making a face with a screen in a page in a screen. This faceness, what Deleuze (and Guattari) have called faciality, is often not liked by people because they can't see the face well enough. artefacts, bandwidth limitations, crappy compressing - whatever, masks the face where no mask ought to be. people don't like it when they can't see a face that ought to be visible like a face.

Then. When they can see the face. They often decide it isn't fast enough.

But faces don't need to be fast.

This applies to all our moving images on web pages. Not just pictures of faces.