 |
Author |
Topic: DVD into Processing (Read 2611 times) |
|
bren
 
|
DVD into Processing
« on: Jan 14th, 2004, 3:17pm » |
|
just a little example of taking DVD footage and screwing around with it in Processing:
http://www.brendandawes.com/sketches/dontlooknow/
I used a video camera to sort of act as a bridge from the DVD (AV lead) player into the Mac via firwire.
|
|
|
|
|
TomC
|
Re: DVD into Processing
« Reply #1 on: Jan 14th, 2004, 3:41pm » |
|
Neat. Are you scaling each image to be one pixel wide, or taking a 1 pixel wide sample from each image?
See today's thingsmagazine for a pointer to "slit photography" images of trains http://www.thingsmagazine.net/2004_01_01_oldthings.htm#10740676959677096 6 (and in the comments, my recommendation to look at http://www.serratedimage.com - as discussed elsewhere on this forum - I still don't know how he does the movie version).
Thanks for sharing.
|
|
|
|
|
bren
|
Re: DVD into Processing
« Reply #2 on: Jan 14th, 2004, 4:29pm » |
|
that slit photo stuff is very cool.
I'm setting the video capture to 1 pixel by 300 rather than scale it or take a sample.
|
|
|
|
|
TomC
|
Re: DVD into Processing
« Reply #3 on: Jan 14th, 2004, 5:16pm » |
|
I see. Anybody know for sure what capturing at one pixel wide does?
|
|
|
|
|
REAS
|
Re: DVD into Processing
« Reply #5 on: Jan 15th, 2004, 8:24am » |
|
brendan,
i suspect that capturing the video at 1 pixel wide is identical in practice to taking a 1 pixel wide sample from the left edge of the image. it doesn't look like the software is averaging the frames.
would you like to expand on the taxidriver series (for example doing a comparison study within one director and also across directors) for the processing exhibition?
+ casey
|
|
|
|
|
bren
|
Re: DVD into Processing
« Reply #6 on: Jan 15th, 2004, 10:54am » |
|
Casey
yeah would love to expand the idea as you already mentioned. The original movie I had in mind was Hitchcock's Vertigo as there's very definate red and green parts to the film which would look interesting. I'll get to work this weekend on a selection of movies.
Cheers.
|
|
|
|
|
bsr
|
Re: DVD into Processing
« Reply #7 on: Jan 18th, 2004, 4:47pm » |
|
i really liked this project so i had to give it a go myself: http://hippocamp.net/misc/snowglobe/
yeah, my understanding is that setting the video size to 1 pixel only results in rendering the left most pixel. a simple way to over come this would be to cut the frame in half and render the right half to left incrementing each time. if this makes sense... it'd be easier than reading all the vertical pixel values from one slice into an array and take a whole lot less code.
i'm working on that now, reading the centre slice should give much better results and detail.
in the meantime here's the simple code reading the left most slice:
int numPixels;
BVideo vid;
void setup()
{
numPixels = width-width;
size(320, 172);
vid = loadVideo("snowglobe.mov");
repeat(vid);
}
void loop()
{
image(vid, numPixels, 0);
numPixels++;
if(numPixels > width) {
saveFrame("sshot-####.tif");
numPixels = 0;
}
}
i'm struggling with trying to split the image in two, any pointers would be much appreciated. cheers to brendan for being a genius.
am working on a timescale overlay as well which is added to each image showing the time in seconds together with a nice scale done by reading the fps of the movie.
|
| « Last Edit: Jan 18th, 2004, 4:53pm by bsr » |
|
http://hippocamp.net
|
|
|
bsr
|
Re: DVD into Processing
« Reply #8 on: Jan 19th, 2004, 12:01am » |
|
been thinking about the timescale overlay.
this could be programmed to be automatic by reading the resolution width and the fps of the movie then applying some simple math.
is there a way to read the resolution and fps from a quicktime movie? not sure, but will look into it.
and one more thing... i've got a feeling the results of the code snippet depend on the performance of the machine as well as the compression used in the quicktime source. a faster machine would result in much sharper output. i think.
|
| « Last Edit: Jan 19th, 2004, 12:30am by bsr » |
|
http://hippocamp.net
|
|
|
TomC
|
Re: DVD into Processing
« Reply #9 on: Jan 19th, 2004, 3:01pm » |
|
on Jan 19th, 2004, 12:01am, bsr wrote:
and one more thing... i've got a feeling the results of the code snippet depend on the performance of the machine as well as the compression used in the quicktime source. a faster machine would result in much sharper output. i think. |
|
Right. You can use the videoEvent method to let you know when it's safe to redraw. See my first example here for code which does just that... http://processing.org/discourse/yabb/board_Contributions_Audiov_isual_action_display_num_1071658656.html
(it's also copying the middle column of the video source, if that's of interest).
|
|
|
|
|
bsr
|
Re: DVD into Processing
« Reply #10 on: Jan 19th, 2004, 3:32pm » |
|
fanatastic, cheers tom. it was giving me headaches.
will try and decipher your code tonight and see if i can get my head round it.
|
http://hippocamp.net
|
|
|
synch
|
Re: DVD into Processing
« Reply #11 on: Feb 6th, 2004, 11:20pm » |
|
Hey guys -
I'm working on a school project about the new high speed train connections in Europe. Wanted to do one long photo of the view out of the train from Amsterdam to Rotterdam. Saw the slit photo technique at http://www.asahi-net.or.jp/~js4k-nmk/egindex.htm.. Tried emulating the technique it in flash, my processor disagreed  But your code made it more then possible so thanks!
Here's a little preview ( a couple of minutes of video): http://synch.dev.capcave.com/trainview.jpg
It's probably going to end up as a cd booklet accompanied with an audio cd with interviews with people that live in villages that the train passes. The villages will be marked on the slit photo..
|
|
|
|
|
bren
|
Re: DVD into Processing
« Reply #12 on: Feb 10th, 2004, 3:59pm » |
|
cool
|
|
|
|
|
shooter
|
Re: DVD into Processing
« Reply #13 on: Mar 22nd, 2004, 4:14am » |
|
Hi everyone,
Brendan, I saw your project and I am intrigued. I have a very limited knowledge of java, but i am eager to learn. I had a thought about your project (with taxi driver and vertigo). Maybe you or anyone else here could help me out.
What I was wondering is if it is possible to duplicate this process with a video already on the computer, or does it have to be streaming through firewire? If that will work, then would there be any way to break the linear format of the film, and choose a random second from the video and essentially duplicate the process? I know there is an operator for random integers, but is that only used for numbers, or could it be applied to the quicktime file's counter?
The result would hopefully look very similar to your project, only the film would not progress in a linear fashion. I don't know if it all makes sense, but it would be great if any of you could help me out with this. This is a great little programe by the way.
Thanks!
|
|
|
|
|
|
