You know how it goes. I let it slip to a friend that I could make CDs from old phonograph records, and well, word got around and next thing I knew I had stacks of old records cluttering my studio. So it should come as no surprise that I can’t remember if the bizarre recording that I found among the many piles originally belonged to me or to one of my friends. When I played the record on the turntable, I immediately realized something was very wrong. At first I thought my record player was broken because all I could hear was a shrieking buzz. The noise produced from my speakers was at the same time both jarring and monotonous. The piercing drone buzzed relentlessly, reminiscent of a mind control ray gun or a malfunctioning machine. I quickly turned the volume down and changed the speeds on the turntable in a pathetic attempt to create a pleasing tone. But, to no avail. At any speed it sounded as if the very air was being torn to shreds. To listen to this strange sound click-here.

I quickly turned off the player and inspected the record. Its pale blue label was very fragile and read, “Recordio Disc” by the “Wilcox-Gay Corporation,” which was a company that manufactured home recording systems, allowing amateurs and enthusiasts to “cut” their own records. I did a double-take when I read the title, “My T.V. Program,” which was written in pencil. It was so strange, why would anyone record ten minutes of noise and then title it “My T.V. Program”? Listening to the record straight through, I discovered that it was badly damaged, riddled with skips, clicks, scratches, and pops.

            I imported the recording into Pro Tools, an audio editing program in which I was able to examine the sound graphically and carefully edit out the clicks and pops. Inspecting the visual display of an audio waveform often reveals underlying patterns. For example, the waveform of the pure tone of a violin has smooth flowing curves, while the waveform of the human voice is a chaotic cacophony of interwoven frequencies.

            This mysterious waveform was no exception and I instantly uncovered a fundamental pattern. Examining a very short section of the audio signal, I quickly saw a pattern of twelve and a half repetitions per second, or 12.5 hertz. In other words, the waveform repeated every .08 seconds, resembling the sickening sound of a skipping CD.

            Expanding the waveform to view .08 seconds, which I will call a major section, I perceived a secondary pattern. Each of the major sections was divided into 48 minor sections, and while each of these sub-sections was different, there was a sort of evolution in them. It was as if the minor sections transformed into each other, with the seemingly random patterns repeating themselves with slight variations in each major section.

            Listening to the recording over and over again I racked my brain to find the solution. Then it hit me, maybe this is not a recording of the audio track of a TV program, but a recording of the image transmission itself. Thinking I was on to something, I Googled “TV 12.5 hertz.” On the very first page, I discovered mechanical television, a primitive precursor to tube television.

            After this wonderful discovery, I spent many weeks researching the pre-history of television. It became clear to me that this phonograph album was definitely a recording of a mechanical television which means that somewhere in all this crazy noise there was a series of images; but, images of what? And, can I somehow reveal the images that are hidden in this dreadful noise? 

Mechanical television had a brief life, spanning a little more than a decade, as it was quickly replaced in 1939 with modern tube television. Although not perfected until the 1920’s, mechanical television was the product of the mind of a 23 year old engineering student, Paul Nipkow, who on Christmas Eve, 1883, thought of a way to scan an image using a perforated, spinning disk. Nipkow proposed drilling a series of spiraling holes through the disk in such a way that, when an image is projected onto the spinning disk, only one hole at a time will let light through and so that a single rotation of the disk will scan the entire image. Strangely, Nipkow never made a spinning disk, but his sketches and patents led to the very first televisions.

Technology finally caught up with Nipkow’s idea in 1924, when independent inventor, John Baird succeeded in transmitting images of simple shapes across a rented room in London. His television camera was made of a spinning disk with 30 holes, giving him a mere 30 lines of resolution, and used a selenium cell photoreceptor to transform sequential streams of light into an electrical signal. His image was only the size of a postage stamp and flickered pretty badly, but his intense focus and single minded commitment led him to go on to invent both the first color television in 1928 and the very first video recording on, get this, a phonograph record.

Meanwhile, in the United States, experimental broadcast stations like W2XB out of Schenectady, New York began expanding on Baird’s success; specifically, by adding lines of resolution and improving image quality. By the 1930’s, there were a dozen mechanical television broadcast stations on the East Coast. But unfortunately the turmoil of the Great Depression meant that there was very little commercial interest or investment in mechanical television, which remained mostly experimental until its death.

Armed with the knowledge gained from weeks of research into this amazing invention, I turned back to the strange recording. If you’ll recall my analysis showed that the signal cycled 12.5 times per second, which corresponds to 12.5 frames per second. And also, that the 48 minor sections suggest 48 lines of resolution within each frame. This meant that within each second of noise there are twelve and a half, delicate 48-line black and white photographs. My research showed that this signal was consistent with the signal from K4XV3, a station which broadcasted mechanical television signals from Philadelphia, Pennsylvania for a brief time in the mid-1930s. So then I knew where the record came from and about when, but the question of why this recording was made would only be revealed when I was able to reconstruct the mysterious images. This, at first, seemed an almost impossible task, but the more I thought about it, the more I began to think up ways to transcribe the audio waveform into a visual image.

The signal is created as the light coming in through the lens of the camera passes through the spinning disk. Due to the specific arrangement of the holes in the disk, only one line of resolution is scanned at a time. The light detector inside the camera translates the stream of light into an electrical signal, which when run through a speaker, produces an audio waveform. Because the signal is derived from varying amounts of light, a peak (a high value on the waveform) corresponds to a bright spot on the image, while a trough (a low value on the waveform) corresponded to a darker spot, I thought that I would just measure the value at each point along the waveform and then reconstruct the value as a level of brightness for each line of resolution. Well, quick calculations placed 3,456 points along each cycle of .08 seconds, the equivalent of a single image. With almost ten minutes of signal it would have taken many decades to transcribe the whole recording by hand. But, despite the overwhelming tedium, I spent a few weeks doing just that and excitedly uncovered my first images. They were small, crude, blocky, and distorted, but they were real. Yet, even with this early success, I knew there had to be a better way!

The solution came as an epiphany while engrossed in Photoshop. I reasoned that since the waveform is a graphic representation of the amount of light at each point in the image, it can be translated into brightness values by filling the space above the waveform with black and filling the space below the waveform white. Then, flattening the waveform into a straight line, Photoshop averages the value according to the amount of light and dark at each point. With higher values remaining whiter, and lower values turning darker, a perfect representation of each line of resolution is quickly created. Splicing and stacking the lines together into the final image is then all that’s left to do, a task which can be automated within Photoshop.

Although still a tedious task, the images finally started to be revealed at a decent rate. A very strange show started to take shape. It appeared to be a collection of short moving subjects. Twelve subjects in all. It is not altogether clear if there is even a plot, because at first it seemed to be just a random collection of objects that move, like a series of animated test patterns. But the selection of objects sets up a strange narrative, one which suggests the experimental nature of this early form of television, and its eventual end in obscurity.

"The Stopwatch", from the Little Black Box of Sunshine by Jeffrey Moser c2008 10" x 1.5"

  "The Cat", from the Little Black Box of Sunshine by Jeffrey Moser c2008 10" x 1.5"

"The Bellows", from the Little Black Box of Sunshine by Jeffrey Moser c2008 10" x 1.5"

"The Hero", from the Little Black Box of Sunshine by Jeffrey Moser c2008 10" x 1.5"

"The Electrode", from the Little Black Box of Sunshine by Jeffrey Moser c2008 10" x 1.5"

"The Spiral", from the Little Black Box of Sunshine by Jeffrey Moser c2008 10" x 1.5"