Difference between revisions of "Useful approximations"
From CCRMA Wiki
| Line 9: | Line 9: | ||
* Humans can usually hear from about 20 Hz to 20,000 Hz (20 kHz) | * Humans can usually hear from about 20 Hz to 20,000 Hz (20 kHz) | ||
* Perception of "loudness" for different frequencies is (very) roughly flat from about 100 Hz to 10 kHz, for most decibel levels. Pitches sound much less loud outside of this range. | * Perception of "loudness" for different frequencies is (very) roughly flat from about 100 Hz to 10 kHz, for most decibel levels. Pitches sound much less loud outside of this range. | ||
| + | * A quarter of a decibel change in intensity is the threshold of perception. | ||
== Digital Audio == | == Digital Audio == | ||
Latest revision as of 15:30, 30 March 2010
This is an attempt to keep track of the various numbers and approximations that are good to have in your hip-pocket when studying audio and music.
Physics
- Sound travels roughly 1 foot every millisecond.
- Typical t60 for a small, house-sized room is .5 seconds, for a larger room (eg, classroom) it's about 1 second.
- A 20 Hz signal has a period of 50 milliseconds ( .05 seconds), a 20,000 Hz signal has a period of 50 microseconds ( .00005 seconds)
Perception
- Humans can usually hear from about 20 Hz to 20,000 Hz (20 kHz)
- Perception of "loudness" for different frequencies is (very) roughly flat from about 100 Hz to 10 kHz, for most decibel levels. Pitches sound much less loud outside of this range.
- A quarter of a decibel change in intensity is the threshold of perception.
Digital Audio
- About 6 dB of Signal to Noise Ratio is gained per bit when quantizing continuous values.
- In order to avoid aliasing, the sampling frequency must be two times that of the highest frequency present in the signal.
- 44100 Hz = .00002675737 seconds between samples, 48000 Hz = .000020833333 seconds between samples
