we are releasing the Good-sounds Dataset. It contains monophonic recordings of instrumental sounds and accompanying metadata. The dataset is composed by 8750 recordings of 12 different instruments (flute, cello, clarinet, trumpet, violin, alto sax, tenor sax, baritone sax, soprano sax, oboe, piccolo and double bass) recorded by 15 different musicians. Some of this recordings were performed using different microphones at the same time, leading to a total amount of 16308 audio files.
Two kind of recordings are available: single notes of almost the whole range of each insturment as well as some scales.
The dataset is freely available for download here.
It was described in the publication "A real-time system for measuring sound goodness in instrumental sounds" and it is used in the one we will present in the next ISMIR'16: "Good-sounds.org: a framework to explore goodness in instrumental sounds"
Oriol Romaní Picas
Music Technology Group
Universitat Pompeu Fabra
note:this is also added to Freesound under user MTG.
Georgetown Univ. Computational Linguistics Group held its first meeting in Sep.16, 2016, for the Fall term, organized by Nathan Schneider(our new professor of LING and CS). It is very exciting to see some ~20 people from LING and CS depts who are working on CompLing and IR topics, among others. After years of development I feel that the CompLing is growing at a good rate here. Look forward to many more activities and collaborations to come.