[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: sound events database

Ciao Ernesto,

level surely plays some role in the identification of sound sources (e.g., weight of objects).

However, to my knowledge the database you need does not exist (although check the latest work of Laurie Heller for a notable exception with nonliving environmental sounds; also check Shafiro and Gygi, 2004 for info on sound databases).

As such, I would either force your machine classifier to ignore level information, or analyze multiple "level-jittered" instances of the same signal.



title={{How to select stimuli for environmental sound research and where to find them}},
  author={Shafiro, V. and Gygi, B.},
  journal={Behavior Research Methods, Instruments, \& Computers},
  publisher={Psychonomic Society Publications}

Bruno L. Giordano, Ph.D.
Postdoctoral Research Fellow
CIRMMT – Schulich School of Music
555 Sherbrooke Street West
Montréal, Québec, Canada  H3A 1E3
Phone: +1-514-398-4535, ext. 00900
Fax : +1-514-398-2962

Neil Waterman wrote:
May be I'm confused - but a large dog barking at 5m could very easily present the same audio level generated by a slightly smaller dog at 4m?

So unless you need to reproduce Fido (being a SPECIFIC dog) at 5m (with Fido having been recorded along with an appropriate calibration tone), then I'm not sure this makes a ton of sense.

Surely what you really need is a database of typical SPL levels for certain sound sources at known distances?

I work for a company producing Level D certifiable sound systems for use in the highest fidelity sound systems, and reproduction must EXACTLY match the source aircraft, but once you accept general sounds (as opposed to SPECIFIC ones) there is no reference SPL only representative levels - I face this problem when representing the ground traffic around the aircraft while on the ground (push-back trucks, luggage carts, etc)

- Neil

On Apr 22, 2010, at 9:46 AM, Ernesto Accolti wrote:

Dear Boris,

2010/4/22 Boris Defreville (ORELIA) <boris.defreville@xxxxxxxxx <mailto:boris.defreville@xxxxxxxxx>>:
So you'd better look at widely available commercial sound effect CD's

I don't do that because the sounds in these CD's are not calibrated.

Anyway, many thanks.


__________ NOD32 4977 (20100326) Information __________

This message was checked by NOD32 antivirus system.