Support camera video recordings as input

[dechamps]

Currently, videojitter takes as input a WAV file representing light intensity over time, which is expected to be generated from the signal recorded from a light sensor.

It may be useful to also support video camera recordings as input, so that the user can literally just point a camera at the display under test and then feed the video to videojitter.

This is unlikely to provide more accurate results than using a light sensor (in fact, it will likely be much worse unless a high-speed camera is used), and would involve much bigger input sizes and processing time. Despite this, it might still be more convenient for users who already have a suitable camera at hand.

In practice, this could take the form of an additional videojitter tool that takes a video file as input and converts it to the usual recording WAV file format based on the total luminance in each video frame.