""I began to make it when I had no money for raw stock and only several rolls of colored leader but nevertheless (had) the need to make or work on a film. As I first conceived the film, I intended it to be a kind of revenge done with the bare hands against - first of all animation - or cell animation in particular and secondly, against abstract film with a capital A as they were practiced in the late 40's and 50's as a kind of engine cooler for the art houses where I first saw serious foreign movies. As I thought about the film, I wanted it to have a very open, resilient kind of structure with the maximum possible amount of rhythmic variety, both in terms of count, beat and variety in the rhythmic changes of shapes and the rate of the rhythmic change. I used a debased form of matrix algebra to make up, in advance, the structure of the film, and tried out several arithmetic models for that structure... with very short film pieces, before I found one that seemed to suit me. As I came to make the film, it consists entirely of 240 feet of black leader into which are welded about 1,000 separate events. Each consists of one frame, and there are 40 kinds of frame, ranging from a frame that consists entirely of red or green or blue to a frame which may consist of red leader with a triangle of blue leader welded into the middle of it. I say welded because the film was put together using three colors of leader and 3 ticket punches - a square, a circle and a triangle - which I felt to be constantly recognizable and also impersonal shapes - and where one color is let into another, or where a color shape is let into black leader, it is literally welded in with acetone. I was doing all of this under a magnifying glass with tweezers and brushes and so forth... they're disposed along the continuous line of film by a scheme roughly the following: in order to avoid a scheme in which certain types of frames would, by rhythmic recurrence, fall at the same spot in the film, or in the same exact frame, I decided to use prime numbers, that is, numbers divisible only by themselves and as a starting-point since they begin to share harmonics extensively only in their very high multiples - I further decided I could use no prime numbers less than 40, because 40 is the number of frames in a foot and didn't want any single type of event to occur any more often than once every one and two/thirds seconds, and then I subjected my list series of tests that involved the sums of their digits-casting out those that didn't meet the tests so that as it turned out the, commonest event, a frame that is entirely red, occurs every 61 frames in absolutely regular repetition throughout the film; and the least common event, a red triangle on a black ground, occurs every 2,311 frames - all of this necessitated an amount of arithmetic which I did over a period of 6 weeks - reduced it to a large stock of 3X5 cards and collated them, and sat down which my rewinds and splicer and simply put the thing together - altogether on the level of personal logistics, it tied up my time and need to be making a film for about three months at the end of which I found myself with a little more money for raw stock and I could go on and make other kinds of films."