TNMOC Blog VII – Making music from the sounds of the world’s first computers

Originally published on the British Library Sound and Vision Blog

Between January and March 2015, Matt Parker was artist in residence at The National Museum of Computing, Bletchley Park. The residency, which was supported by the Arts Council England’s Grants for the Arts Scheme, was an audio archiving project that resulted in the production of 116 unique audio recordings of some of the world’s most historically significant computer technologies. Within the collection are sounds of the world’s oldest original functioning digital computer, the Harwell Dekatron (also known as WITCH), a faithful replica model of the world’s first digital computer, Colossus and a replica model of the electromechanical decryption device the ‘Bombe’, designed by Alan Turing and Gordon Welchman. In the first of two posts, Matt wrote about developing The Imitation Archive and his experiences during the residency period. His final post looks at producing a catalogue to accompany the archive and using the raw recordings of historic machines to create musical compositions.

Producing the Archive

Having recorded for many days, fighting against the elements of heavy rain, snow and school trip groups, I set upon the task of cataloguing the material. During the recording period, I was careful to speak into each microphone before recording a particular device, announcing what I was recording. I kept the file names constantly updated with time and date so I would be able to follow what happened, when and where, and I tried to record everything as a 96 kHz 24 bit .wav. I took photos with my smartphone for reference of each item and tried to take higher quality pictures with a camera where possible. A thorough reference of microphone placement and signal path was important to the accuracy of the recordings.

I went through each recording, cropping out the noise of setting up and spending as much attention as possible on listening to the main activity or process that the recording was set up for. I tried to record as cleanly as possible, but in some cases it was necessary to clean up the recordings with a bit of EQ and multiband compression; nothing that would alter the core character of the recording at this stage, just faithful archival replication.

In some cases, I recorded with a few options, for example a stereo, mono, and transducer setup all in one multitrack recording. I was wary in the recording process to be careful with phase alignment so I was effectively able to fade between different microphones. I think there is an interesting question to be asked with the notion of subjectivity with the recording process here. How important is it to capture the object as one hears it? How clinical should a recording be? Is there any point in capturing recordings of things that can’t really be heard naturally by the human ear? Do we want to shut out the architecture or environment that an object exists within?

In the case of The Imitation Archive, I felt that it was important to capture ambient recordings of the objects within the space they occupy in order to demonstrate their presence within a particular environment. It seemed like a pertinent decision, and one worth making, given that I was to record Colossus which is set up in a room where the machine was actually used during the Second World War. In the studio, it felt like perhaps I could play with these sounds to find the most interesting combinations sonically.

Heath Robinson codebreaking machine

Composing the Archive

As a composer, I wanted to find an interesting way to work with this new ‘sample library’ of material. More than just working with the archive in this way, I wanted to draw on the themes and experiences of the recording process; the museum, the objects, the themes around the very concept of producing The Imitation Archive.

One of the key things that struck me was the constant durational aspect of these machines. Many of them were designed to run 24/7 without fault or interruption, performing repetitive cycles. I felt that this would be an interesting idea to explore so I chose to focus on the machines in operation as much as possible; the work cycle, the operational cycle. I also decided to make the composition seamlessly flow between sections, a never ending cycle of computing.

I was also very much drawn into the historical narratives of machines at Bletchley and found myself wanting to reflect the architectural relationship with the sounds as much as possible by playing with impulse responses of the rooms (made using a balloon pop so not an exact science!) and convolving the sounds of the recordings with the space impulse response itself. I used impulse response as a filtering method, locating fundamental frequencies that peaked within the recordings. I would push and emphasise these frequencies to turn the machines into instruments playing their own unique keys. I thought it was an interesting discovery to find how some of the fundamental frequencies tended to harmonise with themselves. Some of the machines such as WITCH, Bombe and Colossus have very distinctive mechanical rhythms, and it seemed to make sense to play with this as much as possible. Their repetitive rhythms would occasionally break from the cycle and create a surprise extra half beat or other micro-beat. Overall, I hope I have given a sense of what it might be like to work with these machines day in day out, in different environments, as well as draw on their relationships with the space in the museum as it is today. Similarities and differences all punctuated through a musical composition.

Conclusions and Future Plans

My experience of producing The Imitation Archive has given me a sense that there is much more to explore in the world of computer and technological sounds. I have been working on a further project that is specifically looking at the relationship between modern IT infrastructure, the latest, cutting edge technologies in computing and their architectural habitats.

As I begin to explore the sites of our contemporary internet landscape from a technological infrastructure perspective, new questions are beginning to emerge; how do we reflect the shift from desktop PCs being the locations of our digital content to placing everything in a mobile networked ‘cloud’ system? What are the environmental relationships between these new palaces of a digital world and their local habitat? As computers become increasingly prevalent in our day to day activities, smart devices, the internet of things, connectivity to remote machines, have all changed our relationship with digital technology. Can sound illuminate for us anything about this somewhat abstract and increasingly estranged relationship? My work in this area can be seen on my project website www.thepeoplescloud.org. As I continue to develop this research, I will be starting to produce a PhD at The London College of Communication in September within the Creative Research into Sonic Arts Practice department (CRiSAP).

Working on The Imitation Archive has been a fascinating opportunity for me to consider the historical impact of computing on culture and society. In the future I hope to find out more about the impact of computing on culture and society within the contemporary moment.

The Imitation Archive was supported by Arts Council England’s Grants for the Arts scheme and is available at the British Library (collection C1679)

TNMOC Blog VI – Recording the sounds of the world’s first computers

(Originally Posted on The British Library Sound and Vision Archive Blog)

Between January and March 2015, Matt Parker was artist in residence at The National Museum of Computing, Bletchley Park. The residency, which was supported by the Arts Council England’s Grants for the Arts Scheme, was an audio archiving project that resulted in the production of 116 unique audio recordings of some of the world’s most historically significant computer technologies. Within the collection are sounds of the world’s oldest original functioning digital computer, the Harwell Dekatron (also known as WITCH), a faithful replica model of the world’s first digital computer, Colossus and a replica model of the electromechanical decryption device the ‘Bombe’, designed by Alan Turing and Gordon Welchman. In the first of two posts, Matt talks about developing The Imitation Archive and his experiences during the residency period.

Project foundation

In August 2014 I produced an audiovisual installation titled The Cloud is more than Air and Water. The installation was concerned with the claustrophobic sound of the IT Data Centre. Through original footage and field recordings, the piece was composed to demonstrate some of the key issues related to the dislocation between user and machine in the ‘Cloud’ age of computing. Having focused my work on the modern environments of computing for almost a year, I began to think increasingly about the historical narrative behind computing; how the environments and relationships between users and computers have shifted over what is a relatively short period of time. This led me to the discovery that there isn’t a comprehensive archive of sounds of computers in existence (at least not in the UK).

It has been argued that perhaps computer sounds are not really that distinct from any other electromechanical machine. There is whirring, clicking, buzzing. Much of the noise is really made by the cooling units rather than the processors themselves. I had begun to wonder myself whether the sounds I was looking for might not really be all that distinct, however when I arrived at Bletchley, the sounds of computing turned out to be wide ranging and unique in ways I couldn’t have imagined.

Harwell Dekatron Computer (WITCH)

Planning the recordings

After having some initial discussions with Stephen Fleming, the PR Manager for the National Museum of Computing and a couple of the museums volunteers, I prepared a list and rough plan for a full week’s worth of recording at the museum, creating slots for each of the major devices and grouped slots for some of the smaller featured devices.

One of the best things about the museum is that there isn’t anything in the public collection that’s there just to be looked at from behind a Perspex screen. It’s one of the really important aspects of the museum’s value as all items are either fully restored or in the process of restoration. Functionality is critically important to the museum; after all why would you just look at a computer (some of the early machines do look phenomenal though)? The very purpose of a computer is that it performs a task, which they all do with accuracy and reliability that makes the reliability of my middle of the road smartphone look pretty embarrassing.

The volunteer pool is large and made up of hardware and software computing engineers, many of which are retired. They are all enthusiasts. They speak in a language that understands computing; they refer to Boolean logic as if it’s common vernacular and discuss the virtues of machine code over software defined interfaces. It’s a fascinating environment where each volunteer contributes their expertise towards the restoration of unique machines. Some are fascinated by the wartime codebreaking machines, the pioneering devices that are said to have reduced the length of the Second World War by six years. Others are absorbed by the mechanics of early era networked systems, giant tape reels and hard drive disks the size of a tractor’s back tyre. Others meanwhile find themselves fascinated by the early PC era, when the initial boom of Silicon Valley hit, when the UK was a powerhouse of the computing industry, battled out between Sinclair, Amstrad, Acorn and their US competitors at Apple and IBM.

Colossus-The Imitation Archive

The residency recording period

From the first day of arrival, it was clear that my initial schedule was not going to run fully to plan. A challenge with programming this kind of activity is disseminating this information to a broad volunteer group and to gain their understanding about what I would be trying to achieve.

On my first day of recording, I setup extensively within the Tunny Gallery, home to a replica Tunny Machine, a series of analogue radio receivers, teletype machines and the restoration project for a Heath Robinson machine (a precursor to the Colossus). The volunteers were very happy to assist with my recordings, patiently waiting for me to setup various microphone configurations and placements. I’m very happy with what I captured. When a microphone is in a room it picks up sound from all corners. As is often the case with recording in spaces with people in them, it doesn’t occur to mind that perhaps any sound you make, any movement, heavy breath, *whispering*! will be picked up by the microphone! Achieving silence in the space for more than two minutes would be an almost impossible task!

I started to schedule sessions at night, after the museum had closed, which was the only way to have guaranteed quiet! One of the unique intricacies of the museum is the synchronised clock system throughout the building. There are two original Bletchley Park mechanical clocks in the entrance hallway which sound incredible. Their recording makes it into the archive as it is a critical feature of the building. Every 30 seconds, throughout the entire block, a clicking thud occurs, synchronising the clocks around the museum together. Every 30 seconds… a recording is interrupted by synchronising clocks! It became a highly significant aspect to the experience of recording, inescapable and sometimes, when you forget that it happens, late at night in a dark corridor, it was shocking!

I decided to work with recording key features of as many objects as possible. This turned out largely to be demonstrating the boot sequences or the turning on of a device, carrying out a basic task with the device and switching the device off. As was the case with the clocks in the museum, other devices that were not digital computers were also making interesting sounds and so, as it was there, I recorded these as well. This saw a lot of time working with mechanical comptometers and other analogue programmed mechanical calculators.

The Imitation Archive was supported by Arts Council England’s Grants for the Arts scheme and is available at the British Library (collection C1679)

TNMOC Blog V – The Turing Welchman Bombe


The Bombe just wasn’t Bombastic enough for The Imitation Game

Between the years of it’s now legendary creation in 1940 to the unveiling of the biggest secret in the British military history courtesy of F. W. Winterbotham’s The Ultra Secret (1974), Bletchley Park’s Enigma decryption machine, the Turing-Welchman Bombe was nothing more than a series of unofficial documents and hearsay.

Upon the official revelations that Alan Turing had indeed masterminded an electro-mechanical decryption device to break the German Enigma machine’s cipher, little information was available on how the machine worked, what it looked like, how it was constructed and how it sounded.

Research and Reconstruction

Following thirteen years of investigation into how the Bombe worked and what it looked by an expert group of fascinated engineers led by John Harper, a Bombe was built; using techniques and parts that mirrored that of their original construction. Unveiled as a fully working demonstration piece for use at its birthplace Bletchley Park, the Bombe reconstruction is modelled on photographs, literature, notes and genuine artefacts that pieced together a fascinating ‘mish-mash’ of a machine, developed quickly and out of necessity, using parts from existing pieces of technology such as electronic punch card sorting machines, to produce a quick and dirty solution; solving the most sophisticated cipher encryption device of its time.

Fast forward to 2015 and inside the Manor House of Bletchley Park, amongst the phenomenal carpentry of wood panelling and antique furniture, lays the latest homage to Turing and Welchmans’s Bombe, The Imitation Game ‘museum’. Following a significant period of filming at the Manor House, the film company gave Bletchley Park permission to keep many of the props used in the making of the 2014 blockbuster including the movies reimagined prop Bombe.

Mad Props

The Bombe Reconstruction, an originally crafted design of utter complexity and logical beauty sits a mere 150 metres away across the park, grease pouring from its axles onto a brushed steel metallic under-tray as it runs the same calculation task for fascinated onlookers, time and time again, always accurate; a testament to the solid design in principle. Beside it sits an earlier reconstruction… I should say ‘stage prop’. It was built, at face-value, faithfully in accordance with the original design, for use in Michael Apted’s 2001 movie Enigma. Despite featuring Hollywood starlet Kate Winslet alongside an acclaimed Dougray Scott, the film arguably lacked ‘global appeal’ and the prop today doesn’t even warrant a plaque or label to denote its origins. Nobody would know it was featured in a major movie unless you were ‘in the know’.

Meanwhile, inside the grand manor house, lies the Imitation Game’s ‘stage prop’ Bombe. It is Bombe 2.0 XL Edition. It is Bombe on steroids; bolder and more brash, but lacking the substance beneath its overtly adorned physique (It doesn’t work). As if the machine wasn’t already absurdly complicated in its original guise, Bombe 2.0 XL Edition comes with additional VU meter readers, valves, dials and other volumetric gage displays, additional cooling blocks and oh so much more of the vivid red wiring flowing over the dark metallic frame’s top and sides, wisping gracefully to the base like a Viking goddess. The mechanical display dials that cover the front panel seem almost ‘enhanced’ somehow. The colours more vivid, more Hollywood, pre-saturated to enhance the image in HD, far more striking than the original design bothers with.

It’s fascinating to see how one invention can garner so much attention. How Turing can be dubbed genius for his developments and contribution to logic theory and information systems, whilst just around the corner from Bletchley Park, the somewhat annexed National Museum of Computing, possesses another, far more impressive reconstruction of the first ever programmable electronic computer Colossus (also utilised at Bletchley Park) as designed by another maverick genius of his time, the much overlooked Tommy Flowers, sees a far fewer visitors, and a much lower level of appreciation than the Bombe.

The Imitation Game is a second stab at the major motion picture cherry for global recognition of the Enigma story but the true mystery cipher that needs decrypting is why the movie industry insists on making such a meal out of it. Is it really necessary to turbo charge the device? It’s already incredibly technical and complex to look at as it is, but along with much of the story about Alan Turing and the Bombe, it is dramatic, but not quite dramatic enough to deserve ‘movie status’, so the truth is embellished, and nobody is any the wiser.

Movie Poster for Imitation Game
Movie Poster for Imitation Game
"Bletchley Park Bombe4" by Antoine Taveneaux - Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons - http://commons.wikimedia.org/wiki/File:Bletchley_Park_Bombe4.jpg#mediaviewer/File:Bletchley_Park_Bombe4.jpg
“Bletchley Park Bombe4” by Antoine Taveneaux – Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons – http://commons.wikimedia.org/wiki/File:Bletchley_Park_Bombe4.jpg#mediaviewer/File:Bletchley_Park_Bombe4.jpg

TNMOC Blog IV – Colossus

Colossus, the first large-scale electronic computer, was used against the German system of teleprinter encryption known at Bletchley Park as ‘Tunny’.

The significant achievement in computing terms of Colossus was the decision to move from a mechanical relay switch counting system to a fully electronic valve system, whose only moving part was a beam of electrons. The concepts in operation lay the foundations for the digital electronic computing revolution. However, there are still many mechanical parts in action on Colossus including the 5000rpm tape wheels and the initial calculation relay switches on the front panel of the gigantic machine.

Stepping into Block H, the original site of Colossus MKII 9, where the reconstructed Colossus exists today is probably a bit of a different story to its wartime operational environment. The machine is encircled by a steel and glass barrier, to protect not just the machine, but the visitors. Chief Engineer Phil Hayes informed me of how the valves can run up to 400 Volts. The machine itself operates faultlessly on a day to day basis at the museum, constructed out of original (or at least authentic) components; it is a marvel to the engineers who designed and built the original machines between 1943 and 1945.

The cold atmosphere in the room quickly warms up as Phil starts to engage the machine. Turned off over night in this museum piece format, the original functional machines were left to run 24/7 in concurrence with chief designer Tommy Flowers’ discovery that so long as valves were switched on and left on, they could operate reliably for very long periods, especially if their ‘heaters’ were run on a reduced current. The space of Block H would have been furnace like at times. One can see from the few archived photographs of the WRNS (Wrens) who operated the machine, how they were permitted to work with sleeves rolled up, battling the temperature inside the enclosed and most top of all top-secret environments. A switchable secondary tape roll was attached so new encrypted messages could be loaded up to reduce downtime between decrypting messages and ensure efficient operation. There was no need to stop unless technical developments were to be made.

Upon my visit to Block H, I was fortunate enough to hear Colossus go from idle to fully functioning, actively decrypting a tape message containing an authentic teletype message sent between posts of the German High Command. The sounds of the machine are more reminiscent of an industrial age than of the silicon based devices that occupy our digital computing world today. Certainly the clunking sounds of magnetic relay switches, the thudding shudder of a teleprinter output, the whine of 5000rpm motors and the sizzle of friction between wheel and paper tape are unlikely to be heard in the evolution of computing again but it’s here that the on start of a new generation of data controlling devices was born, and without even a modest fanfare for the triumph of engineering that occurred within the dense, blast proofed concrete confines of Block H, at Bletchley Park, now The National Museum of Computing.

“It is regretted that it is not possible to give an adequate idea of the fascination of a Colossus at work; its sheer bulk and apparent complexity; the fantastic speed of thin paper tape round the glittering pulleys; the childish pleasure of not-not, span, print main header and other gadgets; the wizardry of purely mechanical decoding letter by letter (one novice thought she was being hoaxed); the uncanny action of the typewriter in printing the correct scores without and beyond human aid; the stepping of the display; periods of eager expectation culminating in the sudden appearance of the longed-for score; and the strange rhythms characterizing every type of run: the stately break-in, the erratic short run, the regularity of wheel-breaking, the stolid rectangle interrupted by the wild leaps of the carriage-return, the frantic chatter of a motor run, even the ludicrous frenzy of hosts of bogus scores.”

Good, Michie & Timms 1945, p. 327 in 51. Introductory: Impressions of Colossus

TNMOC Blog III – Powers Samas Obsolescence


Possibly one of the most powerful moments for me during my residency at The National Museum of Computing, Bletchley Park has been trying to record the Power Samas Punch Card Reader in full function.

The engineer who was with me was on his last day of volunteering at the museum before fully retiring had been obsessively restoring various mechanical and electro-mechanical counting machines such as the giant Powers Samas Card Reader that had previously been used by a large Chicken farm in the north of England during the 1940s.

As with so many of these now obsolete machines, there is no need or desire to maintain them with any kind of level of functionality, even just for posterity sake. This was particular machine was a unique example. Sadly when we tried to record the sound, just a few seconds into operation something broke.

This is the very last moment that anyone, anywhere is ever likely to hear the sound of the Powers Samas Model 041 Punch Card Reader in operation. It was the only known ‘working’ model in existence.

I was interviewed by Brian Anderson for the Vice/Motherboard blog where I discuss this and other activities if you would like to read more.

TNMOC Blog II – RCA AR-88LF

RCA AR-88LFI spent some time working through the process used to decrypt the cipher made by the Lorenz SZ40/42 in-line cipher machine used by The German Army in World War II. These were messages that came from the German High Command. High level stuff.

It’s a fascinating story of incredibly high level skill in logical analysis and something I couldn’t do justice describing myself. I recommend a visit to TNMOC to find out more. I’ve started off at the point of entry… The Radio Frequency transmission of wireless telegraphy.

I’ve recorded many sounds from the RCA AR-88LF at the museum, covering as many sonic aspects that I can. Here is a little teaser of a recording taken from the rear of the unit, picking up super low electromagnetic frequencies.

TNMOC Blog I – Mechanical Calculators

Documenting Desktop Mechanical Calculators

Vintage III - Mechanical Calculator

I have been busy making a detailed audiovisual archive of the below mechanical calculators at The National Museum of Computing, Bletchley Park as part of my project The Imitation Archive.

Trinks~Triplex - Mechanical Calculator Vintage II - Mechanical Calculator

I am going to be working on creating audiovisual archive material for the next month. This is my first post. A series of simple images of mechanical calculators, all of which have a similar aesthetic to them, by and large but from different manufacturers. A variation of a theme. Vintage - Mechanical Calculator Klowther 400 - Mechanical Calculator

These are common devices used pre-computing. They are the original data processors. Moving on, I will begin to investigate the calculation power of electronic counting machines (early computers) and will continue to post audiovisual material as I progress through the collection. Facit - Mechanical Calculator Contex - Mechanical Calculator Brunsviga II - Mechanical Calculator Brunsviga - Mechanical Calculator

The Imitation Archive is a residency project with The National Museum of Computing, with support from The Arts Council England, The British Library and Birmingham Conservatoire.

High definition audio and further commentary coming soon.

Photos in order of appearance:

Bell Punch Company Limited
Model No. 509/D/865.686
UK

Grimme, Natalis and Co.
Model Name. System Trinks-Triplex
Denmark

Monroe Limited
Model No. Unknown
USA

The London Computator Corporation Limited
Model No. LC/509/SF/392
UK

The London Computator Corporation Limited
Mode No. LC/512/SF/158
UK

Atvidaberg Facit
Model No. C1-19
Serial – 522373
Sweden

Contex
Model No. 136557
Model Name. – +-x
Denmark

Grimme, Natalis and Co.
Model Name. Brunsviga 13ZK
Serial No. 220217
Germany