I remember seeing this video on YouTube maybe ten years ago (it has since been taken down and re-posted), and it really fascinated me.
The creator started with a VHS tape of the music video for Roxette's 1991 hit, "Fading Like a Flower (Every Time You Leave)." It should be noted that since the starting point of this project was a VHS tape, even the first occurrence of the short passage (0:55-1:14 of the original music video) is not particularly high-definition. There are a few breaks in the sound, for example, and moments of high contrast (like when Marie Fredriksson is standing in front of the open window) are slightly fuzzy because of light-bleed. Next, the creator plays a copy of that original VHS tape. With just a single copy, the quality of both sound and image has decreased markedly - it's something that those of us who grew up in the VHS era are certainly familiar with. Then we see a copy of that copy.
The creator calls each subsequent copy a new "generation," and each generation becomes more and more faded (like a flower, ha) as it becomes further and further removed from the original. Colors start to bleed, images become fuzzy, and the sound becomes softer and muddier on each iteration of this short phrase. During the seventh generation, some of the color begins to disappear, and there is a black-and-white border that appears around the image; the color is entirely gone for the ninth generation. During the tenth generation, the sound has gotten soft enough that an audible white noise hiss starts to become fairly prominent. By the thirteenth generation, the black and white images have become blurry enough that it would be hard to make out what's happening on screen during some scenes if we hadn't seen the video in its original form. During the fourteenth generation, we get the first occurrence of a waver in the sound, similar to what you would hear if you tape player were running out of batteries. Each element continues to become less and less recognizable as the creator continues to copy each copy. By the twentieth generation, there are essentially no recognizable visuals left, and the song itself is just about gone by generation 22. The creator chooses the 23rd generation as the last one.
This project occupies an interesting space with regard to Walter Benjamin's famous essay, "The Work of Art in the Age of Mechanical Reproduction." Benjamin argues that an original art object has an aura that makes it unique, but if an art object can be mechanically reproduced (postcards of the Mona Lisa, for example), it becomes less unique and loses that aura. Many authors have referenced this essay in relation to the invention of the phonograph, and what we're potentially losing and/or gaining when music becomes divorced from the context of a unique live performance. This "generation loss" study with Roxette takes Benjamin's ideas a step further - this isn't mechanical reproduction of an original; this person is reproducing the reproduction. If this had been done with a digital file, it's possible that the original quality could be maintained through an infinite number of reproductions, but the fact that it's done with tape means that reproduction leads to destruction.
A friend of mine mentioned that this is conceptually similar to Alvin Lucier's piece, "I am Sitting in a Room" (1981). Lucier records himself speaking in a room, plays the recording back and records that, then plays THAT recording back and records the new recording. The eventual result is the same as "Fading Like a Flower" - destruction of recognizable speech and meaning, although in this case, they have been replaced by the resonant frequencies of the room. This made me think of other pieces, like "Come Out" by Steve Reich, which also starts with words that are gradually destroyed through repetition and phasing.
These are all works of art on their own, and perhaps what brings them all together, or what makes them interesting to me is that their meaning as art objects is created through the destruction of syntactical meaning. Destruction as creation?