Sound Quality: Math Degrading Bit Depth?

If I take an audio signal in Reaktor and divide it by 10 and then multiply it by 10, am I degrading the sound quality?

Compare what happens in an audio editing app. If I use a gain command to reduce an audio segment to a very low amplitude, and then normalize it, I'm adding tons of noise.

So is mathmatical manipulation in Reaktor degrading the bit depth?

(What about if you did something really radical, like dividing and multiplying the audio signal by a million a few times?)

 

AFAIK, the Reaktor engine uses 32-bit floating point math. This means that you don't need to worry about losing resolution. Multiplying and dividing a signal by a large number won't degrade your sound, since you're using floating-point math instead of fixed-point math.

So, you've got more resolution than you're gonna need.

Usually. ;^)

There are limits to 32-bit numbers. For example, if you add one (1) to ten billion (10,000,000,000), you'll lose it. But then, that kind of difference will get thrown out by your DAC anyway.

Perhaps that audio editing app is editing fixed-point signals. That would explain the added noise, cause then you're actually losing bits when reducing the signal.

Chet

 

that is the best emoticon i've ever seen! and i don't usually like noses

 

I think the worst emoticon is ;_;.

Its supposed to be a crying face... wtf is that.

 

spooky kabuki? i dunno that looks really wierd.

 

on second thought, i declare this the "confused pikachu"

 

haha