By
bidi
Fri Oct 08, 2004 3:50 pm
its a bit hard to describe, but i'll give it a go.
basically in a sound file when there is silence the amplitude should be zero. but sometimes (for any number of reasons) you'll record something and the silence (standard reference point) will be non-zero. in extreme cases, the whole wave will be raised or lowered quite noticeably when you see it in a graphical display. this is called DC offset. DC because it acts as if there is a direct current overlaid onto your wave file (this makes perfect sense if you know your way around an oscilloscope - where sounds and voltage can be used to produce similar waveforms.
the problem with dc offset comes from running it through speakers. think of the zero line of you wave file being where your speaker cone is at rest. when the amplitude is above the zero line the speaker cone is pushed outward and when the amplitude is below the zero line the cone is pulled inward. the rapid back and forth motion of the speaker cone vibrates the air and makes sound.
so when you have dc offset going, the system is doing work to keep the speaker cone pulled in or pushed out even when it should be at rest. if you figure out how to run a dc wave through a speaker (which would look like a horizontal line but not at the zero point) and look at your speaker you will see how its being pulled or pushed off center even though there is no sound playing.
so getting rid of a bad dc offset will lower the energy pushing the speakers, and stop different sounds from interfering with each other. cause you can imagine having one sound trying to keep the speaker pushed all out and vibrate would interfere with another trying to vibrate the cone from the zero point. basically any complex mix of sounds yields a single complex sine wave and removing dc interfence considerably simplifies that wave, removing alot of inaudible garbage.
that make any sense?
