Meta-Critique of a Published Critique

Link to the original analysis of the 1889 cylinder recording of Johannes Brahms' piano performance of a segment of his First Hungarian Dance.

Though it is not structured in the same way, this analysis follows the most important rule of the eclectic analysis which is to form the analysis around the work.  The goal is to explore this short, noise-filled recording for the nature of Brahms' performance and a window into the past.  The analysis starts off with a bit of historical background about Brahms, and a more lengthy historical background simply about the recording itself. Here, the analysis provides enough data to teach the reader about the difficulties surrounding this out of use recording style (by cylinders).  The amount of information given about HOW to denoise, however, slightly distracts the reader from the text.  Perhaps this portion might be limited if it is to be presented to a reader who is mostly interested in the particularities of the music itself. 

The author takes a section for general observation or as we might call it, an open viewing of the recording.  Then, the reader is provided with at-length syntactical analysis in which every piece of data is scrutinized.  Once again, the diagrams provide good insight as he compares a transcription of the recording with a transcription of the score. This comparison moves the author smoothly into a discussion of improvisation in brahms' performance, which was what the analyst hoped to uncover in the recording.  This leads him into a nice conclusion, which includes his own kind of meta-critique, instructing the reader about just how difficult it was to decipher the recording and how his attempts are merely a first step in the right direction. 

 Overall, this analysis is a huge success.  It provides a multitude of syntactical data from a recording that is barely audible.  It is true that much of the anaylsis was devoted to describing how things were done, which may or may not benefit the reader.  However, this analysis was limited in nature; due to the terrible quality of the recording, it would have been impossible construct any sort of sound-in-time analysis.

The recording in question can be heard here in its original form, throughout the denoising process and reconstructed as a midi file:  (© 1999 Jonathan Berger, CCRMA, Stanford University. All Rights Reserved.)

The original wax cylinder recording 1889

First step in denoising

Second step in denoising

Third step in denoising

Fourth step in denoising

The same phrase with a synthetic sound overlaid upon the voice that was 'excavated' by the denoising process

Midi reconstruction that takes all the data from the analysis of the recording and maps it onto the entire piece