Hi,
I have a 2G file (continous) of speeches. I need the following to be performed:

1) Chop them up to Video files (by speeches).
So this is one gigantic 2G file, and the only way of knowing when a speech has ended is by looking at the video and noticing a change of scenery (it's pretty obvious). This file is a bunch of smaller videos mixed together so the quality between each speeches differs pretty significantly.

The last person who did this for me failed completely because of his lack of common sense. If you can't tell the difference between a speech taking place at a school and a CNN interview, then you better not sign up and waste both of our time.

2) After 1) you should have all the speeches separated. I want just the audio of these files too.
These post-processing are needed:
-Background noise reduction (like those background sound) to make main speech clearer,
-volume normalization across all speeches to a certain level so the volume levels of different files wouldn't differ so much.

The output sample will be like this:
2G => 15 video files + 15 audio files (I don't know the actual figure)


Due to the failure of the past project, I would need you to do a sample which would be around 3 speeches. I will pay 50% after the sample is done, and 50% when everything is done. (this is negotiable, but I have to see a sample first because my last winning bid was so bad that I kind of lost faith in humanity).