Mechanical Musical Instruments

Pianos that could play back music from a punched paper roll, ie a pianola were once quite common. Playing a stringed instrument mechanically is rather more complicated. The general term for a machine that plays more than one kind of instrument is an Orchestrion. They were controlled by either perforated paper rolls or pin barrels. More info came be found here. The players were often coin-operated, and were placed in ice cream parlours, billiard halls, restaurants, dance halls, theatres, and so on.

The timing of the introduction of these machines, around 1907-1912 seems a little odd; according to Wikipedia, disc recordings were popular and had almost completely superseded wax cylinders by 1912. The reason is probably that disc reproduction then was so poor that real instruments would sound very much superior. No problems with intermodulation distortion!

THE SANDELL SELF-PLAYING VIOLIN PATENT

Left: The Sandell Patent: 1907 This machine was wholly electric, with the stopping of the strings controlled by a bank of solenoids. (70) These were controlled by contacts pressing against the paper roll; when there was hole in the paper the contact touched an electrified bar and the appropriate solenoid was energised. The strings were bowed by four rosin-coated friction wheels 24; these could be pressed harder against the string by more solenoids, (88) to give dynamics, ie a louder or softer sound. These wheels were turned by the electric motor C which also moved the paper roll. There is no integral piano. This is a complicated patent that runs to 20 pages. Everything looks well worked out, strongly suggesting a working model had already been built. It claims to "improve, in matters of detail, the self-playing musical string-instrument which forms the subject of Letters Patent of the United States No. 807,871, dated December 19, 1905" The patent was assigned to the Mills Novelty Company of Chicago; see below. US Patent 856,604. June 1907

Left: The Sandell Patent: 1907 This shows the plungers for stopping the strings. The rods 76 running upwards were depressed by a bank of solenoids 70, guided by holes in the cover-plate 75. Four picker-fingers 86 were provided to allow the strings to be plucked rather than bowed, ie Pizzicato, and were operated by solenoids such as 87b. The violin 1 is partially supported by the jackscrew 4. US Patent 856,604. June 1907

Left: The Sandell Patent: 1907 This is an end-on view along the axis of the violin neck 1a. The support 71, 72 holds the solenoids 70 above the violin. US Patent 856,604. June 1907

THE WAUTERS AUTOMATIC VIOLIN

Left: The Wauters Automatic Violin: 1907 More competition for the Hupfeld. This is an article in Scientific American about the Wauters Automatic violin: An Automatic Violin Player: By George Gilbert on 28 December 1907 "In view of the present popularity of the piano player, and the marvelous perfection this instrument has attained in reproducing the work of the best musicians, it is very evident that it will be only a question of time before other musical instruments must similarly surrender to mechanical control. The latest development along this line is a machine which will play violins and kindred instruments. As may well be imagined, the violin offers difficulties which are peculiar to itself, and we are not surprised to learn that the Violin player illustrated herewith is the culmination of seven years of continuous labor and experiment." "The instrument requires no alteration in the violin itself, and any violin may be placed in the player and removed without injury. The parts are pneumatically controlled in a manner similar to that of the ordinary piano player. A perforated music sheet selects the notes which are to be sounded. This sheet travels over a "tracker board," provided with the usual ducts in which an exhaust is maintained. There are two ducts for each note, and as these are uncovered by perforations in the music sheet, the air rushing into one of the ducts acts through the medium of the usual valves and pneumatics to press a finger down on one of the violin strings at the proper point on the finger board, while the air in the other duct puts into operation the bowing mechanism of this string. The bowing is done by means of four crystal disks, one for each string."

Left: The Wauters Automatic violin: 1907 "In the accompanying drawing the details of the bowing mechanism are shown. Fig. 1 illustrates a section taken through the body of the violin A., The strings are indicated at B. The disks G, with which the bowing is done, are an inch in diameter and % of an inch in thickness. They are mounted in the ends of levers D, which are connected to the pneumatics E. When one of the bow ducts is uncovered, it operates a valve, which connects its respective pneumatic E with the exhaust chamber of the machine. The pneumatic is thus deflated, swinging the lever D to which it is connected, and bringing the disk G on this lever into contact with the selected string B." "The disk G is rotated at high speed by means of a belt, which is guided along the lever D, as best shown in Fig. 2, and runs over a pulley F at the opposite end of the lever. When the lever D is swung into operative position by the pneumatic E, the pulley F is brought into contact with a driving pulley G, and is set in motion by a frictional contact therewith. This motion is communicated to the disk G, which operates on the violin string."

Left: The Wauters Automatic violin patent: 1912 This top-view drawing from Wauters' patent shows the complexity of the machine; the violin is shaded. The drums holding the perforated music sheet are at the bottom of the image. "The speed of revolution may run up as high as 2,000 revolutions per minute. The rate at which the disks revolve determines the loudness of the tones. A device is provided for applying rosin to the disks. This consists of a small cup attached to a spring arm and containing rosin, which bears against the revolving disks." "The fingers of the violin player are sixty-five in number, although more can be added if desired, to reach the extreme high range of the A and E strings. There is a finger for each note. The model shown employs fingers reaching the seventh position. In front of each string is stretched a rubber band, upon which the ends of the fingers strike, thus producing a touch like that of the human finger, and making it possible to imitate the "slide." The tremolo is produced by a set of four hammers, which are actuated by electric vibrators of the type used in call bells. When a hammer vibrates against, a string, next to the bridge, the tremolo effect. is produced on that string. All the strings may have this effect, or one, as the character of the music demands." (I think he meant vibrato, a cyclic variation in pitch. Tremolo is a cyclic variation in volume: ED) "Directly over the violin are four small pitch pipes, which are blown, on pressing a button, by causing air to pass through the pipes, each of which gives the tone of one of the strings, G, D, A, or E. The operator then tunes the violin in unison with the pitch pipes. Violinists know that it is hard to keep a violin in tune. But few appreciate that this is due to the sweat of the player's fingers, which makes the strings stretch. Strings on instruments placed in the violin player do not need much tuning. Silk E strings have been found to last two months, and have stayed in tune two weeks without attention."

Left: The Wauters Automatic violin: 1907 "The tempo is varied by means of a friction pinion which is moved radially on the face of a large driving wheel. This device for varying the tempo enables the simulation of rubato passages when it is operated by a skilled musician. Instruments of the violin family have four strings, each with a range of two octaves. The violin player enables each string to be treated, at will, as a separate violin, as each bow is controlled by a separate mechanism. In the model shown, the higher portions of the G and D strings are not utilized, but they can be by supplying extra fingers. Notes on a violin are found sometimes on each of the four strings. For instance, the G above the treble staff may be struck on all the strings; so that if a trill were being performed on that note on one string, an arpeggio passage containing the same note could be produced on the other strings. Of course, no human player could do that." "It is possible for the player to render a solo part, with a cello accompaniment on the bass strings, or a solo with two accompanying violin parts, all on one violin. The possibilities for combinations of orchestral effect, therefore, are seen to be many. Harmonics are produced by the application of just enough pressure to a finger to make it rest lightly in the string sounded, thus imitating the action of the human finger. Trills are produced with striking clearness by providing a series of small perforations in the music roll. The same principle applied to the bow pneumatics produces springing bow and flying staccato. In making the first music rolls for the player, the inventor, Prof. Wauters, of Binghamton, N. Y., had many technical details to solve. Instruments having fixed strings or tones are played on the tempered scale. But violins play on the untempered chromatic scale, and therefore it was necessary for Prof. Wauters to lay the groundwork for producing music rolls for instruments of that character." This shows the bundles of piping leading to the fingering pneumatics. (bellows)

THE HUPFELD PHONOLISZT VIOLINA

Left: The Hupfeld Phonoliszt Violina; mechanical playing of three violins and pianola: 1912 According to Antique HQ, the Hupfeld company “sold this type of instrument for 20 years without any competitors”; as you can see from the rest of this page, the statement about having no competitors appears to not be true. It would be good to know how many were actually sold. The instrument was manufactured by Ludwig Hupfeld at Leipzig, Germany. Source: Popular Mechanics for Mar 1912, p424

Left: The Hupfeld Phonoliszt Violina: 1912 The machine is electrically powered, but the stoppng of the violin strings is pneumatic (and so is the pianola section) so there was an electric compressor to provide air pressure. Some pianola were worked by vacuum rather than pressure (see the pianola page) but in that case the working pressure differential cannot exceed 1 Bar. However you can use as much pressure as you like, which is probably helpful when designing a large and complicated machine. The three violins were bowed by by a circular structure carrying 1300 horsehair threads which can be seen running across the narrowest part of the violin body. This, the vertical columns, and the upper circle all rotate as one unit. You might think that it would be a continual labour to keep the three violins in tune- with a total of twelve strings to adjust- but it appears to have been a manageable business. I consulted a violinist who confirmed that sweat from the player's fingers on the strings is a major source of violins going out of tune; that of course was not a factor here. Nonetheless you would have thought that thermal expansion alone would require some attention to the tuning. Another point that has just come to light is that only one string on each violin was actually played, so you only had three strings to keep tuned. Source: Popular Mechanics for Mar 1912, p424

Left: Hupfeld Phonoliszt Violina: 2008 This shows how the initially puzzling circular bow worked; it was in fact made up of a series of straight lines that approximated a circle. It was made up of 1350 horsehair strands.

Left: Article on the Hupfeld Phonoliszt Violina: 1912 Source: Literary Digest; original source article by A. Troller in La Nature 8 June 1912, Paris

Left: Article on the Hupfeld Phonoliszt Violina: 1912 Source: Literary Digest; original source article by A. Troller in La Nature 8 June 1912, Paris

Left: Article on the Hupfeld Phonoliszt Violina: 1912 This confirms that the Violina worked on compressed air, and not vacuum. Source: Literary Digest; original source article by A. Troller in La Nature 8 June 1912, Paris

Left: Article on the Hupfeld Phonoliszt Violina: 1912 A close-up of the three violins. Source: Literary Digest; original source article by A. Troller in La Nature 8 June 1912, Paris

Left: Article on the Hupfeld Phonoliszt Violina: 1912 Source: Literary Digest; original source article by A. Troller in La Nature 8 June 1912, Paris

Left: Article on the Hupfeld Phonoliszt Violina: 1912 Source: The Literary Digest; original source article by A. Troller in La Nature 8 June 1912, Paris

Left: Hupfeld Phonoliszt Violina restored: 2008 The machine has been restored to its original condition by Dutch restorer Fred Bernouw. Its current location has not been discovered. Source: Classic FM, which page includes two videos of the machine in action. It is claimed that the second recording (a Chopin piece) clearly demonstrates the machine can even do vibrato, but looking at the string-stoppers I reckon its more of a trill.

A Hupfeld sold for US$ 658,000 at Bonhams in October 2010. You can read about its history here.

The Hupfeld machine was described in the magazine Music Box, summer 1970 edition.

At least one Hupfeld machine was built that played six violins. It is in Siegfried's Mechanical Music Museum at Rüdesheim, Germany.

An anonymous someone wrote a book on the Mills Violano Virtuoso; see here, claiming it was the "World's Only Self-Playing Violin And Piano" which at the time it was written (1900 is stated, but that sounds a bit early) may have been true.

There's a Violano Virtuoso for sale here.

Left: Mills Violano Virtuoso: 1911 The Mills Violina Virtuoso was operated electrically by solenoids rather than by compressed air or vacuum. See the Sandell patent above. There is more info on the Mills Company here.

Left: Mills Violano Virtuoso: photo 1968 A photograph of a Violano Virtuoso once in the Jim Hamilton collection; there is more information here. This looks absolutely identical to the machine in the two colour photographs above, and it may be the same one.

The Encore company later became that splendidly-named organisation, The American Automatic Banjo Company, Ltd. There is much information here.

Even more impressive was a complete banjo orchestra combining seven instruments: banjo, piano, snare-drum, bass-drum, triangle, tambourine, and castanets. More info here.

There is also a complete banjo orchestra machine to be heard on YouTube, playing 'Rock Around the Clock'. This version of the Banjo-Orchestra is composed of nine instruments: piano, banjo, snare drum, tambourine, triangle, wood block, castanets, bass drum, and cymbal. The instrument operates from a unique ten-tune orchestrated music roll. You can arrange to have one made for you by the D C Ramey Piano Company.The price is unknown.

You may notice that the instruments in this section are not stringed instruments. They have a temporary home here while I plan separate galleries for Auromatic Brass and Automatic Woodwind.

Left: The Trio Fantastique Robot Band: 1959 That is presumably Zenon Specht at left. What seems to be missing from all the pictures of the Trio Fantastique is the multicore electrical cable which surely must have been connnected to each robot. The cables were either very well hidden or removed before photography,

Left: Les Trio Fantastique Band: 1950s The impressive cover of a literary review of 'strange fiction'. The price shows it is a French publication. I for one welcome our new robot-musician overlords.

Left: Les Robots Music Band: 1950s Here you can see that the accordionist has three forearms. There is also a rather longer clip of the band (38 seconds) on YouTube. During this performance the saxophonist stands up, moves his head, and then sits down again. A thick cable can be seen running from the back of the stage to the right wrist of the sax player. This is the only robot band picture found so far where a control cable is visible. There is more information here, and here, and also here.

You can hear this band on YouTube, hereThe video includes the punched card reader (at 0.49) that controls the robots; judging by the great number of pipes in view the working system seem to be pneumatic, much like a pianola.

Left: Another Decap Band: 1950s This is a smaller Decap robot band; the line-up this time is saxophone, drums, and accordian. The design of the robots looks exactly the same, though they are painted differently. I have seen it stated in several places that the saxophone sound is actually produced by a Hammond electric organ behind the rear curtain of the stage, with loudspeakers either side of the performers. This seems highly plausible; I was wondering how the embouchure of a human sax player was replicated. You can see and hear a similar Decap robot band on YouTube, but it is not the same one.. There is another similar but different Decap robot band on YouTube here. And yet another similar but different Decap robot band on YouTube here, playing 'YMCA' by the Village People; obviously the art of cutting new punched cards has not been lost. There is a very informative PDF here