Dither: A better way?

C

cwillu

New member
Hello all

After reading several articles about how and why dither works, I had got this idea: why add low level noise to a signal when you could add a high frequency tone at 22050hz (which works nicely at 44.1khz) at the same low level. It would give the same averaging effect as a random noise, and is inaudible. I suppose it might mess with some depth cues, but other then that, it should work just as well as normal dither...

Anybody have any thoughts on this?

William Underwood
 
Hmm, qbasic can be a wonderful thing. If it's added at exactly the correct level (amplitude up to the 17th bit), it gives the equivilant of 17 bits, with the 17th bit recorded at 22050hz: seems to be trading sampling rate for bit depth.

When you half the frequency of the tone (which turns out to be a triangular wave), you increase the bitrate by one, and halve the sampling rate at the same time. Unfortunately, a 11025hz tone is within the audiable range, so it starts to loose it's usefulness.

Okay, scratch that idea...

William Underwood
 
You must log in or register to reply here.
Back
Top