Im working on a program that needs audio mixing. The problem i have is scaling the sound. For example i have 24 active samples at once. If add them all together and divide them by 24 that would not produce a good result if 1 or more samples are very low. Anyone know the best way to scale the sound so that it's always at a good limit.
Would this work ?:
x = sum of samples
y = x / 65536 (i work with unsigned data)
y is now rounded upwards.
final sound is:
x = x / y
What i need is some good formula to scale sound correctly.