Calibrate Your Ears With SoundBytes by Gabriella Cerrato

Today we inaugurate SoundBytes, the Library of Sounds of our blog. There are two main groups of sounds: real sounds, recorded from different products and/or environments, and synthesized sounds illustrating specific sound quality features. How do we think this library can be useful to readers? First of all, you can learn a lot by simply downloading these .wav files and analyzing them with your own sound analysis tools. If sound and vibration are part of your job, you likely have access to sophisticated, NVH-specific software tools (such as those from MTS and other vendors). If this is not the case, there are several audio software packages such as CoolEdit, Sound Forge and others, that still allow you to record and edit time histories of sound pressure and display spectrograms (for a list of audio editing and analysis software sonicspot.com.

Believe me, you can learn so much about what makes a sound unique by just listening to it and looking at its spectrogram. Hopefully, the sounds in our library will also help you, like they have helped me, understand that the majority of product sounds exhibit common features. As an example, the sound quality components of the noise of a refrigerator are not fundamentally different from those of an ultracentrifuge used for medical research or of an automotive climate control system, therefore, in many cases they can be measured, evaluated and diagnosed using the same tools.

All sounds in SoundBytes are in .wav format, sampled at 44.1 kHz and can be downloaded for free. If you are interested in creating your own collection of digital sounds and want more sounds, a good website is www.findsounds.com, which is a search engine for .wav and mp3 files. But back to SoundBytes - the Real Sounds comprise sounds recorded from different products and different environments. Some of these have been recorded with a single microphone (file name ending with ?) and also with a binaural head (file name ending with ?�), to show the impact of the recording technique on the perceived fidelity of the playback. The second group of sounds includes Synthesized Sounds, that is sounds that have been generated digitally. Each sound has been created to illustrate a specific sound feature, and we start with Modulation, Tonality, Loudness and Spectral Balance. These are also the same types of sounds that software manufacturers like MTS use to calibrate their algorithms. I suggest that you listen to some of the synthesized sounds to calibrate your ear to these different features. Then you can go back to the real sounds in the library and try to figure out which features you can detect and are most sensitive to.

A final technical note: to facilitate the listening experience and the comparison among the different files, unless otherwise specified, all sounds have been equalized to play back via Microsoft Media Player at approximately the same output level. This means that a very quiet sound (such a golf putter hitting the ball) will play back at the same level as a much louder sound (i.e. a 12-cylinder Ferrari driving away). In other words, these sounds are not played back at their original level. Please keep in mind that SoundBytes is a work in progress and we plan to add more sounds, from MTS and from our readers. So, if you have interesting sounds to share with the readers of the blog, please send me an E-mail with the non-proprietary .wav file(s) and a brief explanation (i.e. its origin, boundary, conditions, etc.) and we will consider adding it / them to the library.