Hi Everybody,
Does anyone know of the existence of broadband pulse signal made up of 1 cycle per frequency?
e.g. 48 kHz
19980 sines (from 20 Hz to 20 kHz) all starting at the same time with a duration of 1 cycle each.
Could such a signal be programmed into a WAV file for a given sample rate and bit depth? The duration would be 50 ms (the period of 20 Hz, the lowest frequency).
Sincerely,
Merlijn van Veen
Does anyone know of the existence of broadband pulse signal made up of 1 cycle per frequency?
e.g. 48 kHz
19980 sines (from 20 Hz to 20 kHz) all starting at the same time with a duration of 1 cycle each.
Could such a signal be programmed into a WAV file for a given sample rate and bit depth? The duration would be 50 ms (the period of 20 Hz, the lowest frequency).
Sincerely,
Merlijn van Veen
Have to do some math in Excel and see what the waveform looked like.
With all equal amplitude it would probably look like a single spike close to the 20kHz end.
Probably have to use decreasing amplitude.
With all equal amplitude it would probably look like a single spike close to the 20kHz end.
Probably have to use decreasing amplitude.
Hi,
It is a single spike, by definition. 0V to top rail, to bottom rail, back to zero.
Length and upper bandwidth depending on half the sampling rate. Bit depth will
have no real effect *. Repeat rate will determine the lowest measured frequency.
Spectrum is the same as white noise (constant power per Hz).
The top octave will contain half the power, the next lower
octave a 1/4, the next lower 1/8 etc 1/16, 1/32, 1/64 ....
Conversely white noise can be used to derive the impulse
reponse of an electronic audio chain, but keep it away from
your loudspeakers unless you like your tweeters deep fried.
rgds, sreten.
* Other than you clearly can't add up a series of sine
waves when the amplitude of each is meaninglessly
inaccurate, the maths I suspect are utterly hopeless.
It is a single spike, by definition. 0V to top rail, to bottom rail, back to zero.
Length and upper bandwidth depending on half the sampling rate. Bit depth will
have no real effect *. Repeat rate will determine the lowest measured frequency.
Spectrum is the same as white noise (constant power per Hz).
The top octave will contain half the power, the next lower
octave a 1/4, the next lower 1/8 etc 1/16, 1/32, 1/64 ....
Conversely white noise can be used to derive the impulse
reponse of an electronic audio chain, but keep it away from
your loudspeakers unless you like your tweeters deep fried.
rgds, sreten.
* Other than you clearly can't add up a series of sine
waves when the amplitude of each is meaninglessly
inaccurate, the maths I suspect are utterly hopeless.
Last edited:
- Status
- Not open for further replies.