What is big-data in the world of environmental acoustic recordings?
07 Dec 2015
The ecosounds lab is hosting more than 21 years of audio recordings of the environment. This is equivalent to about 186 000 hours of continuous audio. Total file count (at the time of writing, 07/Dec/2015) is 334 302. This seems like a lot of data. Who could ever possibly listen to that much sound? However, these figures should be put in perspective.
The British Library (see picture) curates some 6.5 million recordings of many kinds of sound. Some of its collection goes back to the very origins of sound recording in the 19th century.
The British Library (foreground). St.Pancreas Railway Station and Hotel (background) They don’t build public buildings like they used to.
The Cornell Lab of Ornithology (see picture) holds three million hours of marine recordings (there is much money to fund whale research) and one million hours of terrestrial recordings. Its collection goes back to about 1926 when the first 78rpm record albums were sold featuring birds songs of the USA. The Cornell lab stores not just the recordings in digital but also much of it is in the original format such as the albums of 78’s.
The Cornell Lab of Ornithology viewed from the far side of Sapsucker Woods Pond.