vog (a BETA videoblog)

One of the first reasonably complicated child movie projects I attempted, which back in the day (2003) just did not play very nicely online as bandwidth didn't really cope. Mouse into either video and it toggles through five text tracks with five child movie commentary tracks. There's a bug where once you get to the fifth it does not loop back to the first. Click either video, there are three video tracks for each video pane, starting from highly compressed to high resolution and the cycling back to low res again. The three progress bars embedded are for video and the soundtracks. The ones above the video panes show how much of the track is loaded, and then where you are up to in the current clip. The centre one does the same for the audio commentaries. The original blog post has more:

The work is made up of a parent movie with three child movie tracks. Each of these child movie tracks contain, in turn, multiple movies.

One of the child movie tracks contains five soundtracks, which are loaded on demand. This is determined by counting mouse entries into either of the video panes. Each of the video panes is also a child movie. The one of the left is a view of the Moyne River in Port Fairy, hand held from a small picnic ground. The pane on the right is a view from the ocean wall at the mouth of the river, looking straight down a pier pylon to the sea and kelp below. Each of these video child movies contains three movies, each of identical content, just progressively higher quality. They default to the most heavily compressed versions of each clip when the movie begins.

Mousing into either of the video panes is counted, and determines which of the five soundtracks is played. For each soundtrack there is an accompanying graphic which appears, containing the text of the voiceover soundtrack. Clicking on the video causes another child movie track to be loaded. With the sound and video tracks each cycles through their series continuously, so that once you've viewed each of the video tracks the fourth click will load the first, most highly compressed, copy of the clip. Likewise with the soundtrack.

I added three download progress bars, one for each of the childmovies (the two videos and the soundtrack movies). I haven't used these before as a general rule, thinking that the lag that's involved with childmovies, which is a qualitatively different lag to just downloading a clip, is a condition of the network. But it is clear that users just don't get what it means to use childmovies, that the content resides elsewhere on the network and is not loaded at the same time as the parent movie (the movie that you first 'play'), and so if you click on the video, the movie you are 'watching' requests another movie from a server, and this other movie cannot start playing until it has downloaded. So I've added progress bars so that there is an indication that something is happening. Which is a pretty basic HCI point isn't it?

The progress bars aren't that clear. The blue line indicates where in the current movie (remember it refers to the child movies) you are up to. There is a light grey background to the bar which indicates download progress. A child movie will not start playing unless enough of it has been downloaded for QuickTime to reckon it is can play without having to pause for more data.

I added the text with the soundtrack in response to an observation of Anders'. While what he pointed out about the soundtrack is correct, these works are quite actively not about usability, and explore redundancy, repetition, and noise. So while the use of the soundtrack in the example he mentions doesn't 'add' to the usability of the project, it does require the user to do something, in much the same way that not disclosing all information or narrative at once, and requiring some act for further disclosure, is a reasonable action to have to perform. So, here I've taken it further, by introducing some redundancy and further repetition by narrating and stating the narration as visible text within the movie. What does it achieve? I don't know. Nothing specific, but it overlays a sort of density to the words as spoken and read which is not so much about redundancy (on the way to vacuity or ennui) but rather to foreground the distance, gap and material differences between spoken and written text. It is not an act of erasure, to speak is obviously not the same as to present written text, and by placing both in such proximity is to explore this difference, and to foreground it.

The work is very much about water, a sort of crude Renoirism (Jean, not his dad). The heavily compressed videos, which load first because they'll download much faster than the higher quality videos, have a degree of pixellation that deliberately produces the artefact equivalent of digital impressionism. This is done without filters, just compress it down very hard and let the codecs start to work their magic, largely by working on 16 x 16 blocks of pixels (in the case of Sorenson Pro) which become very visible when the codec has to make do because you've set a data rate too low.

So, the blurring and noise this introduces, particularly when you film things like clouds and water (abstract elements), gives an elegance and beauty to the content that is serendipitous in its relevance. The water has a mutability and play of surface that of course provides quite the reflexive mirror for work exploring digital compression. Light reflections, surface, depth, pattern that is abstract and observed. This is not much different to early narrative cinema, where many films found themselves gazing at light on water or light on clouds, there is a narcassism to this looking which is perhaps less the misread gaze than the machine seeking the mirror to glimpse its face. So, while about water and location, I think this work is much more about video, more so than most of the other works where I have been much more consciously reflexive.

I wonder if this is then a more mature work? Nah. Bugger that.