Reproducing Pianos
Pianola Journal
Friends of the Institute
Concerts and Recordings
Recent News
Contacts and Information
Site Map
Text Only Version
Click Images to enlarge

The Pianola Journal - Volume 7, 1994.

The Pianola Journal - Volume 7, 1994

  • Editorial
  • Debussy and Welte: Roy Howat
  • On the Roll - Easthope Martin and Other Pianolists: Rex Lawson
  • The Reproducing Piano (Part 3): John Farmer
  • A Note on the Technique of Recording: Reginald Reynolds
  • Concert Review
  • Book Review


Debussy and Welte: Roy Howat

The Pianola Journal has sometimes made history by printing first musical editions of pieces composed and initially notated for the pianola, This article is about a counterpart to that, the first systematic use - as far as I know - of piano rolls to help establish an accurate edition of music already well known.

Debussy recorded fourteen piano pieces for Welte: La soirée dans Grenade from the Estampes of 1903 (roll no. 2735), D'un cahier d'esquisses of early 1904 (roll 2734), the six pieces of his Children's corner of 1908 (roll 2733), the humorous waltz La plus que lente of 1910 (roll 2736), and five Preludes from Book 1 of 1909-10 (three on roll 2738 and two on roll 2739). Already this raises a question of what, if anything, went under the serial numbers 2737 and 2740-1, for which no rolls were ever issued. Was there more Debussy? Unfortunately no answer is known, and the archives that might have told us seem to have long vanished.

Another question is recording date. It is often assumed Debussy recorded in 1913, the year some of these rolls were issued, and the date written on his eulogy to Edwin Welte (see facsimile opposite), but various factors suggest a little earlier. In particular, had Debussy recorded in 1913 he'd surely have included something from his second book of Preludes, completed early that year. Also, the techniques of preparing a roll (see Rex Lawson's comments on this in The Pianola Journal no. 5, page 33) suggest that Debussy's rehearing of a roll would more likely have been considerably after he recorded it. So we can probably conclude between 1910 and 1912 - probably nearer 1912, for it would seem unlikely for Welte to sit on such prime unpublished material for longer than necessary.

For many years, both as pianist and as a researcher of Debussy's music, I'd been interested by the rolls, both for what they revealed (or at least implied) of Debussy's performing habits, and for some musical differences from the printed editions. The most radical difference was a succession of tempo changes in the prelude La Cathédrale engloutie not marked in the printed score or Debussy's manuscript (of which more below); there were also many others, mostly added or omitted chords and notes and changed figurations and harmonics. I was already using some of them in my performances - at least as far as I could discern them from the Telefunken LP transfers, which were then my only source for what the rolls contained,

A Note on the Technique of Recording: Reginald Reynolds

Very many pianists perform at the AEolian Hall, in London, leaving no trace of those delightful sounds with which they have charmed their audience during a brief recital; yet fortunately other and greater pianists have found their way to the top storey of the AEolian building, and there discovered a means of perpetuating their interpretations for all time.

In a secluded room stands a Weber grand piano, in tone and in outward appearance not different from the usual model, nor does the touch betray the magic power beneath the keys. Upon closer inspection the secret is partially revealed by the electric cable which can be seen coming from beneath the instrument; and if it were possible to trace this back into the piano, there would be found 160 wires, half of them leading to specially devised contacts under the keys, the remainder running to positions near the point where the hammers strike the strings, while the cable itself passes through the wall of the room, coming out into a sound-proof chamber, in which is installed the amazing mechanism that constitutes the Duo-Art recording apparatus. Here the other ends of the wires are attached to electro-magnets, which operate the punches in the powerful perforating machine, each punch corresponding with each key of the piano. The pianist plays - the punches perforate - the record is produced!

This method of recording ensures absolute accuracy of reproduction, the length of the perforations being determined by the period for which the key is held down. Thus staccato notes produce little round holes about 1-32 of an inch in diameter, a tribute to the agility of the fingers and also to the rapidity of the recording punches which are working at 4,000 pulsations per minute. The rhythm is determined by the spacing of the perforations in the music roll as it passes through the recording machine at a uniform speed (usually 8 feet in one minute), and this spacing is in exact accordance with the interval between the notes played by the pianist, so that when the music roll is placed upon a Duo-Art piano, and caused to play at the same speed, there must result an exact reproduction of all the most subtle nuances of rhythm.

Similarly the touch of the pianist is recorded and reproduced; still by means of perforations in the music roll, in conjunction with the most ingenious mechanism, both in the recording machine and in the Duo-Art piano. By the use of only 8 "dynamic controls" no less than 32 variations of touch can be produced, extending over the whole range of finger power, from the lightest pianissimo to the strongest accent, and in combination with the well-known "Themodist" device ("Pianola" patent) the melody is differentiated from the accompaniment, each having its own free modulation of tonal effect.

back to top

The Pianola Journal - Volume 8, 1995.

The Pianola Journal - Volume 8, 1995

  • Editorial
  • Hugo Cole: an obituary
  • Whose Fingers on What Piano?: Denis Hall
  • Punching Pianola Rolls: an education project at the Huddersfield Contemporary Music Festival: Bill Vince
  • The Art of Player Piano Transcription: Edward Schaaf
  • Two Pleyela Recordings of The Rite of Spring: Louis Cyr
  • Other Minds 2: Rex Lawson


Whose Fingers on What Piano?: Denis Hall

A criticism sometimes levelled at the use of reproducing pianos to recreate performances of many years ago is that the piano played by the pianist is not the same as the one on which the piano roll record is replayed, and that the acoustic of the recording studio cannot be the same as the room or concert hall in which the reproducing piano is playing. Since any pianist will respond both to the instrument and the environment in which he is playing, then by their very nature, reproducing pianos and their recordings must be fundamentally flawed in their attempts to 're-enact the artist', the phrase used by Ampico in the 1920s. This article will concentrate solely on this topic and will deliberately ignore other strengths and weaknesses of piano roll recordings.

There is undoubtedly something in this criticism, but a deal less than the detractors of the reproducing piano would have us believe. That fine English pianist, Harold Bauer, wrote in his autobiography Harold Bauer - His Book in 1948:

"I spent many hours in the offices (of the Aeolian Company) editing and correcting the paper rolls on which my performances had been mechanically recorded for the pianola, later electrified and re-named the Duo-Art.
"I made from first to last some two hundred records, taking infinite pains in the editing that was essential to their completion.
"The final result was always somewhat discouraging in spite of all this trouble, for the reason that the dynamics, set to produce certain effects on the piano which was being used for such editorial purposes, varied when the record was played on another piano. This was due to minute differences in quality of tone and resistance within the action, and there was no way of overcoming the difficulty."

On a slightly different tack, but on the same related subject, he wrote in a letter to W. C. Woods, the American Duo-Art recording producer and editor, in 1922:

"I am very curious to see if I shall be able to do work which will be satisfactory for both Europe and America, for there of course is the difficulty which has been hitherto encountered. I was particularly struck by Cyril Scott's records, which he corrected in London and revised in New York. I heard the two rolls (of the same piece) and in my opinion there can be no question that, admitting the New York version to sound better in New York, the original London version undoubtedly sounds better in London.
"This is very curious and interesting and I shall try to examine very carefully wherein this difference lies. One thing I have already noticed. In the regulation of the soft pedal, the hammers are brought closer to the strings than over here, so that soft pedal effects will produce more difference in New York than in London. I have examined a number of upright pianos and feel fairly sure that I am not mistaken, but it might be well for you to compare measurements. I shall let you know of anything further which strikes me. I am rather inclined to think that as public taste in piano tone here is different from prevailing standards in America, your dynamics will frequently prove too high to give best results on instruments regulated for the greater lightness and brilliancy which has always characterised European pianos."

These are the only such comments the author has come across by a recording artist, and they cannot be ignored. Nevertheless, Bauer made his first rolls for the Duo-Art reproducing piano in 1915 (5627 - Etude Op 10/ 12 - Chopin) and was active right up to the very late (7451 - Etude Post. Nos. 1 & 3 - Chopin) issued in 1932, so one may deduce that he was not as discouraged as all that. He was prepared to record the solo part only of Saint-Saëns' Second Piano Concerto for the Duo-Art specifically for public performance with orchestra, hardly the act of an artist with no faith in the process. To put this roll recording activity into perspective, he did not start to make discs until Victor signed him up in 1924. From then until 1929, 12 double-sided 78 rpm records were issued, far fewer titles than of piano roll recordings.

The Art of Player Piano Transcription: Edward Schaaf

Introduction by Rex Lawson
Dr Edward Oswald Schaaf is an intriguing musical figure of the early part of this century, whose compositions and arrangements occasionally turn up in American roll collections, but only very rarely find their way across the Atlantic to Europe. His musical style was somewhat florid, partly on account of his conception of the player piano as an instrument that could not easily bring out melodies without the doubling of voices or other contrapuntal assistance. Hitherto the player piano world has known very little about his life and work, which was surprisingly fertile, including nine operas, two masses, three string quartets, several symphonic works, and large quantities of music for military band.

It is a difficult matter to obtain some sort of unity and order in the plan of presentation of a subject which has had no technical treatment, and in entering upon untrodden ground the subject is approached with all due diffidence. The whole treatment of the subject will no doubt have to be revised as often as the treatment of a rapidly developing medical science.

The player piano is rapidly adjusting itself to the highest artistic requirements, and is opening out a scope of hitherto undreamed magnitude to the imagination of the composer of piano music, and the task which the author sets out to accomplish is an exposition of the chief principles which govern player piano transcription generally, and to explain technical details, by which special effects may be secured.

This volume is not intended as an exposition of the art of interpreting player piano music; it deals with the intellectual, rather than the aesthetic, and concerns itself with the principles of expression only, in so far as they relate to the technical side of the subject.

Newark, N.J., 1914.

back to top