Recording Method
All recordings that are part of this manual were made on the 10th of March 2022 in the recording studios at the Koninklijk Conservatorium Den Haag with sound engineer Pablo Gastaldello.
Materials
- Two Neumann KM184 paired microphones were placed about 2 meters from the saxophone.
- One Neumann U87 microphone was placed about 1.5 meters from the saxophone.
- Focusrite RedNet2 with a Solid State Logic (SSL) XLogic VHD Pre preamp was used for audio recording.
- Reaper recording software was used to record raw material.
- Sonic Visualizer with the YIN plugin. YIN is a vamp implementation of the Yin algorithm for monophonic frequency estimation. It is used to analyze the fundamental frequencies of single monophonic pitches and was developed by Matthias Mauch and Simon Dixon.
Saxophone Models Used in this Study
- Soprano Saxophone – Henri Selmer Paris – Super Action 80 Series III – Solid Silver Neck
- Alto Saxophone – Henri Selmer Paris – Supreme – Lacquered Neck
- Tenor Saxophone – Henri Selmer Paris – Super Action 80 Series III – Solid Silver Neck
- Baritone Saxophone – Henri Selmer Paris – Super Action 80 Series II – Lacquered Neck
Methodological Steps
- Recordings were made of each individual note produced by the four saxophones without mouthpiece. For the sake of this study only the standard fingerings were used. No alternate, eighth-tone, quarter-tone, or multiphonic fingerings were used. (However, these can be integrated through a future study).
- Four to six test recordings were made to ensure accuracy of the material.
- Each note was examined in Sonic Visualizer using the YIN plugin. Below you can see a screenshot from Sonic Visualizer with its pitch analyses.
- This material was analyzed and a hertz value was extracted for each of the four test pitches.
- This value was confirmed by ear with a pure sine wave of the same value to ensure accuracy. This step was important: despite being powerful software, it is not infallible and it is only a tool to aid in analysis. Everything had to be tested through the software and then confirmed aurally.
- The recorded hertz values of all four tests were combined to create an average hertz value. The standard deviation was also calculated in order to understand the variability between the individual tests. This statistic shows how unstable the pitch result for a specific fingering might be.
- The average hertz value was tested by ear with a pure sine wave of the same hertz value.
- The average hertz value was converted to the precise musical pitch using the following formula:
mp is the musical pitch and f0 is the frequency of the individual averaged pitch.
- This formula gives the MIDI number which represents the closest tempered semitone, plus how far out of tune this pitch is from that semitone.
- I converted this pitch value to the nearest eighth-tone.
- From the nearest eighth-tone value, I converted this to musical notation.