12
Replies
interesting question....hmmm...
I don't think you're actually trying to make CLIPS below one second... you are really trying to illustrate certain FRAMES within a clip as a visual stimulus ( photosensitive epilepsy … with the 'photosensitive' being the key word ).
Don't know particulars but in my mind there's two ways to go about it maybe...
To understand where I'm coming from check out this link ….and the pic below it is an example from that link....
That's using a still camera with a fairly long shutter speed ( say 1/15th of sec ) with a FLASH going off during that exposure... The flash is about 1 1,000th of a second in length ( essentially duration of the light ).
Using flash and strobes on still cameras gets a little tricky. I have strobes for still work, but the ballast takes about 3 seconds to recharge, so it would be useless in a situation like you are looking for.
Sooooo, maybe solution number 1 for you to think about... shoot whatever you want for your clip(s) … do test this stuff out to see if it works first by exporting and looking at small sections of your timeline ( WAB narrow to export and look at small sections )….and painstakingly go to every 4th frame ( 6 in total per 24 fps ) and change it to make it APPEAR like what you see above.
Save time... go to first frame of clip, make selection, sharpen and lighten subject of your choice, invert, add blur and maybe motion blur, whatever... save.
Now select that clip and copy attributes ….select the other frames ( 4,8,12,16,24) and 'copy attributes'....This would get tricky with a mask and movement of your subject in mask, but you'll get the idea... you are just FAKING a strobe.
solution number 2, is the obvious... get a strobe like they use at nightclubs and balance your lighting and put diffusion or ND stuff on strobe to get exposure you want...
Those things strobe about a zillion times a second...
It is it mostly "replicate a real stimulus" i need it so i can train a program to detect if a video could contain a trigger.
But it has been difficult for me to make frashing faster then 1 second so far.
Video is indeed a trigger for many people with epilepsy and is an ongoing problem. Many producers of live shows like to use short clips and fast cuts, use flashing stage lights, strobe lights, etc. Many action movies use the same/similar approach, short clips, flashing frames, fast cuts, etc, etc. Some documentary editors has an odd fetish for flash frames, glitches, glitch transitions, etc, etc. Some music videos as well. The younger the editior is the more flashes and glitches can be seen. None of those flashes, flash frames and glitches add anything to the documentary though...
It´s not the video per se that is the trigger, the seizures are triggered by the light in the video. So fast cuts and flash frames with only dark grey muddy video won´t trigger so much. Using light is the key. The more light in the images, the more likely is it to trigger a seizure. Think strobe lights.
What you need to do is to zoom into the timeline as per the image below. By doing that you can cut the video at frame level more easily and insert 24 *different* frames per second if you use a 24 fps timeline. That should be sufficient enough.
your project is a very cool and worthwhile one... and thank you for working on it
AdChoices