In the 1920s, the German engineer Curt Stille developed an electromagnetic wire recorder based on earlier attempts by Valdemar Poulsen in 1900. At the end of the decade, Stille sold his patents to Ludwig Blattner, who used them to develop steel tape recording for the BBC, enabling the time-delayed broadcasting of radio programs in the Empire Service. Today, apart from one at the Museum of Communication in Frankfurt, no Stille wire recorder seems to be preserved. Faced with this absence, we decided to build one ourselves.

Redoing: A Hands-On Approach

Starting from our joint practice of tracing the history of sound and media hands-on, Ingolf Haedicke and I embarked on an intensive 18-month exploration of early tape-recording history by “re-doing” one of its apparatuses. Although neither of us had formal training in electrical engineering, we could draw on practical experience. Ingolf Haedicke was a long-standing lecturer in practical construction and media technology at the department of media studies at the Humboldt University, Berlin. I first gained hands-on experience in the department’s Media Technology Workshop under his instruction. Sharing an interest in the insights offered by work with media technologies, our tinkering friendship persisted after I graduated.

As a woman without prior technical expertise, I learned the basics of many sound media by making them again—(un)learning by redoing, if you will—and have since integrated these approaches into my teaching and research. Haedicke’s tinkering skills reach back much further. Born in 1943, he grew up in East Germany, responding to the scarcity of consumer goods by tinkering. After studying musicology in East Berlin, he led the Humboldt University’s phonotheque, set up a recording studio, and equipped seminar rooms with sound systems, many of which he built himself.

In our joint work over the previous ten years, we had focused on electronic musical instruments: we reconstructed obsolete devices (e.g., the Terpsiton), re-soldered and re-listened to historical DIY projects (e.g., a GDR synthesizer), repaired historical apparatuses (e.g., the Telefunken trautonium in Halle), and more. The wire recorder project was the first time we had worked on a sound storage device, but all our projects shared a resonance with practice-based approaches that have recently gained increased attention in the humanities. Documented by scholars including Nick Hall and John Ellis in media studies or Sven Dupré in the history of knowledge, these have given rise to various “re-” terms—whether reconstruction, reenactment, re-experiencing, or reverse engineering—that carry different epistemological implications and conceptual requirements depending on the academic tradition (media, theater, and music studies; history of technology, art, science, and knowledge; anthropology, etc.) and methodology (phenomenological, systematic, epistemological, ontological, aesthetic, etc.) concerned.

Yet none of the wide range of terms available seems to fit our practice exactly. Experimenting and interacting with various historical and contemporary components and materials, we carry out research through and within an object by “redoing” the process of its making and use (Fig. 1). Studying sources with a soldering iron in hand is not only fun, but also opens new research perspectives from within the object of study itself: our literally hands-on approach unfolds a new epistemological horizon, awareness of contingency, and concrete experience. We see this research practice not as simply a way of tinkering with objects or experiencing historical media, but as a critique that sheds new light on previously overlooked aspects of sound media history and technology.

Resources for Redoing

When we started work in early 2022, Ingolf Haedicke immediately suspected that the project could become very complicated, since though the electronics were far less extensive than those of a synthesizer, the mechanics were all the more difficult to get right. Nevertheless, we began as usual with research, using written sources in the broadest sense—historical and present-day books and hobby magazines, circuit diagrams, patents, construction and repair projects, recordings on historical wire devices, and more.

We also looked into building materials, most importantly the recording medium, steel wire, which we were able to acquire online in a 1950s version (Fig. 2). In addition, we procured various electric motors for the drive, considered possible shapes and materials for the “housing,” and assembled and tested different speakers and amplifiers for the sound. We even disassembled an old, defective tape recorder to get an impression of how its functional units had been arranged, also hoping (in vain) to find spare parts for our device. Other material necessities became apparent only during the construction process itself. For example, months later we bought numerous combination switches to find the optimal buttons for play, record, rewind, and stop on our device. We initially tried attaching the sound heads with clothespins, and later prepared them with plastic scraps as edging and glued metal pieces for wire guidance.

Clearly, then, the project proceeded in a more recursive manner than the coherent form presented in the video documentation. While historical sources provide guidance, actual implementation often demands trial and error. This can be due to missing materials (finding a suitable sound head), mechanical and structural challenges (e.g., creating the winding device and tape supports), soldering errors, and the fragility of the electronics (e.g., wiring the key set, where connections often came loose or unwanted solder bridges silenced the device). In addition to the actual device, we also built various test setups (Fig. 3). One has since become a permanent part of my teaching as a demonstration object: a small handheld recorder in which the sound wire is freely suspended in the room and the sound heads are manually guided along it.

A Noisy “Tape Recorder”

Our project of redoing had three essential elements, which implicitly echo the three elements of a wire sound device: storage medium, electronics, and mechanics. With historical photographs of Curt Stille’s inventions as our guide, we first considered the mechanics, especially the shape of the device. For magnetic storage technologies, this is not just about aesthetics or usability; it also requires a secure mechanism for guiding the steel wire and using it to record and play back sound. Electromagnetic interactions are crucial, affecting recording and erasing heads and making them susceptible to interference from power supply transformers or feedback from the microphone and built-in loudspeaker (video, 1:56–2:01 min).

We modeled our device on Stille’s prototype of a sound film apparatus, which anticipated the design of the later Blattnerphone (Fig. 4). We mechanically reduced the design in a T-shape and enlarged it compared to Stille’s device.

The second phase involved finding suitable electronics. Stille’s wire recorder already employed many elements of later tape technology, such as amplifiers for recording and reproducing sound, premagnetization of the steel wire, and electrically operated forward and reverse motion of the wire. We opted for a simple transistorized preamplifier based on the circuitry of a 1950s battery tape recorder and an IC module for final amplification. The sound head required the most attention. After extensive testing, we chose a 1950s studio sound head from VEB Goldpfeil Magnetkopfwerk for its superior recording and playback quality (Fig. 5).

When the wire was mounted on our frame and connected to the amplifiers, we heard heavily distorted speech, overlaid with noise (video 2:04–3:00 min). This prompted the third and most challenging phase: grappling with the steel wire sound carrier. Here, the quality of the wire material proved to be key. We acquired 1950s Recordophone wire, which had a slightly smaller diameter and different alloy but offered enough length and stability. The ultrathin wire had to be handled very carefully to prevent it from unwinding or breaking—it tangles easily, necessitating an optimized environment with smooth surfaces and precise synchronization of wire spools to avoid loops, burrs, and knots (Fig. 6, video 00:00–00:41, 2:34–2:44 min).

In our version of Stille’s wire sound device, the interplay of wire length, quality, and speed was audible. Irregularities in the wire and inconsistent movement affected the sound, producing distortions, noise, and volume fluctuations. Although the principle of sound recording is electromagnetic, the mechanics are always present: when recording, one’s ears and eyes constantly check the sound head to ensure the wire is working properly. The moving mechanical parts create background noise, especially the rotating tone wheels, which implicitly set the rhythm to which one’s speech melody adapts. When all components work well together, the sound quality is surprisingly good for spoken language (video 3:16–3:37 min), but for music, we were not able to produce a clearly recognizable recording.

Redoing Stille’s wire recorder revealed that even the seemingly simple act of recording sound with steel wire is a delicate negotiation between material, mechanics, and electronics. The tiniest irregularities in wire tension, movement, or quality translated directly into audible distortions; clearly, the theory of “noiseless” electromagnetic recordings rarely coincides with practice. Methodologically, our redoing demonstrates that working with historical objects does more than reproduce past technologies: it can shift the focus of inquiry within the field itself. The material and technical contingencies of the wire recorder foreground aspects of sound media history that textual sources alone cannot reveal. Redoing is thus a critical practice that generates new questions and methods directly from engagement with the object.

Today, our device is housed at the media studies department of the Humboldt University. It serves as a constant reminder that the early use of steel wire as a sound storage medium was neither natural nor obvious, as is evidenced by its long development period from the first demonstrations around 1900 to marketable recording devices in the 1920s. As we found out, it required steel wire to be literally bent into shape to make it sound. Redoing that process of bending and sounding allowed us to listen to the noisy episodes of media history and set new tones.

By Christina Dörfling, Humboldt-Universität zu Berlin