Joat42 wrote:TFLYTSNBN wrote:You actually make an excellent point about tampering with the video evidence. That video is going to be probed along with the guards. Keep in mind that given modern digital video editing techniques, doing a deep fake of a static view of an empty corridor will be extremely easy.
If you actually knew something about video forensics you would be aware that it's not hard to spot manipulated videos and especially if someone used deep fake since it leaves extremely big turds of artifacts in the resulting image/video stream.
isaac_newton wrote:ah-ha didn't you realise that Prince Andrew's personal MI6 agents were using 'extra deeper fakes' - the next gen products right out of GCHQ?
Of course it helps that they were wearing the very latest in full body tin-foil invisibility cloaks...
Joking aside, faking video of an empty corridor is actually very very hard. You may think that you can for example re-use old footage or just copy a few frames over and over but that doesn't cut it.
On the face of it the video may look like an original which will make most people say it wasn't tampered with, until you run it through a forensic tool (see
this tool for example).
But there are other methods where you extract a graph of the light-level over time which can then be compared to another graph taken from a video made by a camera nearby. If the graphs doesn't fit each other one of the videos have been tampered with.
Some may ask how the above could work, but it's very simple. The load on the local grid introduces variations in a light-source's output which the cameras will pick up.
And now someone will say that "why not fake the ambient light level in the video then!" and my answer is that it will introduce detectable artifacts since it also has to match up with the ambient noise introduced from the electrical grid AND the specific camera too. And noise is very very hard to fake because of its entropy.
