Scott Joplin’s piano rolls

Scott Joplin left no recordings except several piano rolls. They were incapable of capturing changes in volume, and some appear to have been posthumously edited. One can hear them replayed on YouTube: they sound like bar room music. The Entertainer in particular has been overexploited as a background to saloon fights in comedy films. That’s definitely not what Joplin wanted, according to John McWhorter.

Even in Joplin’s time – his active years were approximately 1895 to 1915 – there existed a more sophisticated mechanical technology. Developed by the famous Welte-Mignon company, it produced first-rate results. Gustav Mahler left us a precious twenty-five minues of Welte-Mignon rolls. Peter Gutmann of Classical Notes wrote:

True, these are piano rolls, a medium with a deservedly bad reputation. The integrity of many rolls was compromised by extensive doctoring, both to correct wrong or mistimed notes and to “enhance” the original with new harmonies, runs and doublings. Even when uncorrupted, standard rolls had no quality, as all notes sounded at the same volume and with the same flat, staccato tone. Fine for a barroom, but hardly genuine art.

Mahler’s rolls, though, were made in the new Welte-Mignon system, perfected in Germany in 1903. How did it work? We really don’t know, since the proprietary process was a closely-guarded secret and the equipment was secured after each session…

Reproduction is achieved not through a player piano, but with a so-called “vorsetzer” unit, which actually plays a concert grand using felt-tipped “fingers” activated by varying degrees of pneumatic pressure triggered by the sets of holes. The result is uncannily realistic and far superior to the limited range of the acoustic disc in conveying the “touch” of an artist…

It must have cost a lot both to record and to reproduce, and Joplin probably had difficulty convincing New Yorkers that the value of his work went beyond dance-hall and music-hall entertainment.

No acoustic recordings of Joplin’s piano playing are known – perhaps none was made – but that’s a subject for another post.


    • Sheet music can’t have all the information about the way it should be played. Even when the author leaves a very detailed notation, there’s room for adjustments of touch, tempo, volume and pedaling so the same piece wouldn’t sound the same when played by different pianists. Also, a lot of piano music is not so meticulously notated. Sometimes the composer’s directions are confusing or ambiguous. There’s room for interpretation even in a piece that’s been played to death like the Moonlight Sonata. There’s a YouTube clip of pianist Andras Schiff explaining how most players got it wrong (in his view, and he’s one of the best in the market today). It does not mean that all music should be played exactly the way its creator played it, but knowing how he did it always adds (sometimes a lot) to our understanding of the work.

      • Ah, I didn’t know all that, thanks. I kind of assumed the music would have all the necessary notation to let a player know how it should be played.

        I’m interested in this because I play a music for which almost no original recordings exist (American Bluegrass). Reasonable quality recordings started to emerge in the 20s and 30s and there is plenty of original material, but much of it is based on songs from the 19th century for which barely any music is available, let alone recordings. So the whole genre is basically made up as you go along, which makes it an awful lot of fun to play.

      • Actually, that sounds a bit contradictory so I’ll clarify. There are plenty of recordings of “original” songs by groups and performers in the 1920s and 1930s, but many of these “original” songs are either arrangements of traditional songs from the pre-recording era or are heavily based on them.

        It’s not Bluegrass but similar nonetheless, but AP Carter is credited with all the works of the Carter Family even though he went door-to-door through Appalachia collecting songs which had been passed down through the generations. He then arranged them for his own group to perform, and in many instances they were first to ever be recorded doing so. Hence the Carter Family are deemed to be the “original” artist of songs which probably pre-dated them by 100 years.

        • About a month ago, I stumbled upon a band called The Caroline Chocolate Drops, who mostly play Piedmont music, black and white, plus a smattering of country and folk from all over the place. (Their female singer, fretless banjo player and fiddler, Rhiannon Giddens, recorded a fine first solo album last year.) They say they learned some of the songs from local musicians, some from songbooks and records. But I think even when there is only one record of an obviously old, pre-existing song, musicians like the CCD, who have good knowledge of the music and performance styles, can perform it in a way that’s both different and historically plausible. The Carter filter can be reverse-engineered, as it were, although not with 100% certainty.

Comments are closed.

Discover more from Winterings in Trans-Scythia

Subscribe now to keep reading and get access to the full archive.

Continue reading