Random usefulness

• Dapper: http://www.dapper.net/index.php
  Make API’s for any website
• JungleDisk: http://www.jungledisk.com/
  Unlimited online storage. Pay only for data stored and transferred ($0.15 & $0.20 per GB).

Some leads to follow up

Regarding slitscan things and rectilinear/other projections from curved lenses and mirrors.

• Kurt Ralske: www.retnull.com
• Also see: auv-i.com/ for Kurt’s Max/Jitter objects
• Ghetto 360 video experiement: blog post
• Daniel Shiffman: www.shiffman.net not sure if anything much useful will be here though.
• 360 degree movies: www.360movies.de
• Timescape: http://www.phantasian.com/timescape/timescape.htm
• Daniel Crooks: www.dlab.com.au/
  www.acmi.net.au/train_no6.jsp
  

Video cube save now works

I have managed to write code to save (actually it’s ’serialize’ but whatever) a videocube object, so now the loading frames procedure which can be rather time consuming only has to happen with new movies. But the tradoff is the size of the serialized objects is quite large because it is essentially the uncompressed video size. Not sure is serialization supports any kind of compression scheme – it may be worth having a quick look into that.

A zip of the sketch is here (~580KB). Hopefully there’s enough stuff and notes to allow it to be run elsewhere. Sorry I cannot make an applet of this one – it’s pretty much gotta be run from processing itself.

I’ll look at a couple more things with these before I move on:

• interpolation of the borders between frames in the affect area
• persistence of the affect area with a decay back to the stable image (the kronos projector does this – it makes the affect area seem to flow more like deformation in a viscous liquid).

Processing video library mod

I posted about this at the processing forum here.

The zip of the library is available here

A few more ideas

In addition to some potential directions for study this semester, I’ve thought of a few more things:

• supershapes using the superformula. I think this could be applied to the videocube perhaps. I’ll probably code up a quick class or library for processing too to do these.
• a ‘display’ using electromagnets and iron filings. Building it will suck, but I guess a proof (or denial) of concept could be built on a small matrix of say 16 x 16. Still looking for some materials for the moment.

I’ll also have to consider looking at the dynamic composition engine idea that I looked into last year. Some of Jared Tarbel’s processing and flash experiments are relevant there. I think the supershapes may be useful there too. Rectilinear projections from video of fish-eye lenses or spherical mirrors may also tie in with that kind of stuff. Haven’t done enough research there yet though.

Wii remote hacking

I was thinking of maybe getting a Wii remote and looking at using for a controller either for music or as an input device for controlling processing. So far I have found the following info on hacks to use the controller from the Wii game console for other systems and purposes:

• A short blog post:
  http://createdigitalmusic.com/2006/12/09/update-wiili-wii-controller-hacks-mirror/
• A longer log post and some video:
  http://createdigitalmusic.com/2006/12/07/midi-control-for-music-with-wii-remote-teaser/
• Some tech info:
  http://homepage.mac.com/ianrickard/wiimote/wiili_wimote.html

Video cube in processing v4.4

Sorry for the lack of activity on this blog – I’ve been working on stuff for SSiC as well as some other heinous group project…

So the video cube has come along quite nicely but I have now run up against some more limitations. Basically the code below uses the processing.video library to read a quicktime .mov file into an array of PImages. Pixels are then picked out of the cube to create a view image (also a PImage object). This works fine for smaller movies, but larger ones run a bit jumpily on my aging computer (which is why the pic that follows is so tiny). The code processes input from 2 mice to proved an x-y center of affect, and a depth-radius (z-q) degree of affect. Mice are switchable by pressing the space key just in case you have a preferred mouse for x-y (or z-q) and it’s not the right device number for the ‘default’ way the code is set up. The proControll library is required to use this. See the processing website or the link in a previous post here.