Posted by 8H Haggis (via moderator) on Mon, 01/02/2006 - 23:50.
(Originally posted 01:36:25 12/21/05)
I listened to the examples given here of the 1889 Brahms cylinder about two weeks after I expended an entire afternoon's work on the same artifact myself. My denoising was much clearer, revealing more of the details of the piano passages than the example given on this website; furthermore, it seems to me that the MIDI overlay is largely an exercise in imagination, offering notes that Brahms *did not play at all* according to my results: could the MIDI "reconstruction" be based on misperceptions of strange intermodulation distortions caused in the original recorder, which altered the content of chords, making the performance sound incoherent? And the recorder picked up octaves and even individual whole-tone progressions so unevenly that it is rare for even two nearby notes in a row to be registered with the same strength or voicing. The MIDI overlay is even MORE incoherent, and musically odd and pointless. It boggles my mind! Can this be the way the great composer Brahms would play *anything*, even as alark? (Remember that only a few years earlier he gave the world premiere of his demanding B-Flat piano concerto.) One is tempted to suspect that the phonograph experiment was conducted late in an evening devoted to far more brandy and cigars than were prudent.
Denoising the Brahms Cylinder; adding MIDI overlay
(Originally posted 01:36:25 12/21/05)
I listened to the examples given here of the 1889 Brahms cylinder about two weeks after I expended an entire afternoon's work on the same artifact myself. My denoising was much clearer, revealing more of the details of the piano passages than the example given on this website; furthermore, it seems to me that the MIDI overlay is largely an exercise in imagination, offering notes that Brahms *did not play at all* according to my results: could the MIDI "reconstruction" be based on misperceptions of strange intermodulation distortions caused in the original recorder, which altered the content of chords, making the performance sound incoherent? And the recorder picked up octaves and even individual whole-tone progressions so unevenly that it is rare for even two nearby notes in a row to be registered with the same strength or voicing. The MIDI overlay is even MORE incoherent, and musically odd and pointless. It boggles my mind! Can this be the way the great composer Brahms would play *anything*, even as alark? (Remember that only a few years earlier he gave the world premiere of his demanding B-Flat piano concerto.) One is tempted to suspect that the phonograph experiment was conducted late in an evening devoted to far more brandy and cigars than were prudent.
8H Haggis - retired audio engineer, San Jose, CA.