Description

This project was a study on the speaking style of different politicians and orators. It was an attempt to capture a sort of "portrait" through analysis of clips of them delivering a speech or public address. It can be seen as a way to distill or summarize an entire persona into a single static image, which in turn helps us to compare different speakers in a quick and efficient manner.

Basically, my program takes a video clip, strips the audio, analyzes and intelligently segments this audio into small sections marking their point in time, and then gives each of these sections a rating in terms of frequency and amplitude. These segments are then sorted, usually from loudest to softest. The program then returns to the original video clip, grabbing "snapshots" at each of the loudest points, which it then combines into a single image, giving us a still portrait of a speaker at his most emphatic moments.

Throughout the process I experimented with many different variations on this basic formula. Sometimes I would create an overlapping portrait, sometimes I would plot each moment in a panoramic timeline. Sometimes I would create an even larger panoramic timeline, allowing the frequency to correspond to Y-axis position. Overall, I feel that the most effective version largely depends on the way in which the clip is filmed and the speaker's use of movement.

The graphs were a visual representation of this sorting process. The top portion shows the clip segmented but still in the correct order. Color corresponds to volume, with dark red being the loudest and dark blue being the most quiet. Compare the graphs of Obama vs. MLK. You'll notice that King's is a clear linear arch, starting soft and leading up to a climax, while Obama consistently leaves various quiet moments throughout, perhaps for reflection.

The sound clips are sorted in the same way as the portraits, going from the loudest point, to the quietest. Very monotone speakers like H.W. Bush will sound almost the same from beginning to end, while more emphatic speakers, such as Mussolini will have a wide range of emotion and will therefore move from extremely loud to almost silent. An interesting coincidence is that Ronald Reagan's loudest moment happened to be when he's saying "the wall" referring to the Berlin Wall, his main topic in this particular speech.

I'm particularly fond of how Nixon, Mussolini, Hitler, MLK, and Reagan turned out. I think it gives a really interesting view into the way in which they use their voice and body to create a political presence. I feel that it's unique way to reflect on these characters and I hope that it gives deeper understanding into their style and personalities in general.