Videogrep

Videogrep is not a video essay but a tool for making such essays. For making supercuts, to be precise. It is a piece of software written by Sam Lavigne (a command line tool written in python, for those in the know) that generates video supercuts based on dialogue. It goes through the subtitle track for a video, searching for a given word, phrase, or even a specific grammatical structure. All instances it finds are then spliced together into an automatically generated supercut. The software is now also available as a standalone Mac desktop app.

 

Videogrep is the technological equivalent of automatic writing: an inanimate computer, possessed by a single human command handed down in code, unthinkingly creates a videographic list. The results can be amusing (such as this supercut of the White House press secretary weighing his words) or as monotonous as a shopping list (this supercut of all mentions of the word “time” in the movie In Time is, well, not really worth your time).

 

Sam Lavigne‘s tool, as he admits himself, is a crude one: since it uses the subtitle track (.srt file) as its guide, the quality of the resulting video is very much dependent on the accuracy of that track. And yet, that rough-around-the-edges quality also adds some mystique and artistic serendipity to the otherwise automated process. Consider the supercut above, compiling all silences in the movie Total Recall. It flows with an eerie rhythm, silent yet staccato. (It also demonstrates that silences are used most often by villains and seductresses). One might be inclined to see Videogrep as a useful tool to quickly edit together a rough cut, only to refine that automatic version further afterwards. But the poetic strangeness of this Total Recall piece shows there is something to be said for just letting the computer work its circuital magic.