High-speed camera sequences
...of stuff being destroyed/damaged by insanely large electrical
I was recently lucky enough to get access to a high-speed video camera to record what
was happenning when bending & destroying things with the Destruct-o-tron.
Myself and artist Mark Anderson spent a day filming various things being being warped,
destroyed and generally blown up.
Many thanks to Mark for providing this opportunity (and making the tea), Wysing Arts for
putting up with the noise, and Panavision for supplying the camera and lights.
Notes on video clips
I don't have a lot of experience with doing videos for web download. I used Mpeg format as
(a) I had the software to do it, (b) I didn't need to decide which AVI codec to use, and
(c) it didn't complain about odd frame sizes, which some of the AVI codecs did.
If you have trouble viewing these, I'm sorry but I probably can't help - they play fine in
my Windoze Media Player....!
However if you are an expert on this sort of stuff and have any constructive suggestions
on better ways to produce movies for efficient web download (without spending any money),
I'd appreciate any input. I have uncompressed AVI versions of these files, so I can easily
recompress with different codecs if this would be worthwhile.
These videos were recorded from the composite output of the high-speed camera's framestore
onto a Toshiba RD-XS32 HDD/DVD recorder, copied to DVD, converted to uncompressed AVI
using Max DVD to AVI, chopped and cropped using
Virtualdub and converted to MPEG using AVItoMPEG1.
Media Player is not very good at single-stepping etc. - if you want to look at these
video sequences in frame-by-frame detail I'd suggest using Virtualdub to view them.
On the camera info border, the figure under 'REC' on the right is the recording speed in
frames per second. Most are 1000fps. The camera could do higher rates, but only with
partial frames, so the higher-speed sequences look like very-widescreen movies!
The figure at bottom-left is the frame counter (positive or negative depending on how we
triggered the camera), and the one at bottom-right is actual time in seconds. When you
single-step some of the sequences you will see some blurring of these figures - this is
due to catching the wrong phase of the interlacing at some point in the transfer process,
and the image will also be correspondingly blurred. There are also a few sequences we
forgot to transfer at 25fps (i.e. one video frame per recorded frame) so the frame counts
do slightly odd things. Some of the early sequences were done before we fully understood
the various camera modes, and were using external sync in an attempt to get >1000fps.
I'm pretty sure the sequences with 'Sync' on the right are are 1000fps for full-frame and
4000fps for part-frame.
In many sequences, a dark band is visible right at the start of the action - this is
almost certainly caused by the intense electromagnetic pulse interfering with the camera.
Many of the sequences took place inside a containment vessel consisting of ten-inch
polythene gas pipe. The inside diameter is 215mm, so you can estimate/measure the
distances and speeds involved.
We need a faster camera! Most of the actual shrinking/bending was happenning just too
fast to observe in any detail even at speeds of 6000 frames per second (166 microseconds
per frame!). If anyone in the UK has access to a faster camera & suitable insanely
bright lighting I'd be happy to bring the destruct-o-tron along for further
However we did get some pretty interesting and often spectacular footage - If you look
at nothing else, check out the first vid on the CD disintegrating page.
More high-speed stuff here
|Click image for frame sequences and movies
|Can Crushing - both directly via a coil,
and by smashing with a hard disc platter into it..!
|Warping hard-disc platters at extreme velocity
|Shrinking wire rings