Linked by Sean Cohen on Tue 13th Apr 2004 06:52 UTC
Today I'm going to talk about why software - any software, all software - actually matters, what the different types of software are, and why you should care about its properties (no matter who you are, or what you do).
Permalink for comment
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.
Recent Original Stories
Recent Comments
Headlines
Random Comments
Random Stories
Random OS Link
Lossy compression mainly discards sounds that are inaudible due to masking. There are two kinds, transient masking (you don't hear something quiet if something loud happens straight after it) and frequency masking (you don't hear a particular frequency if there is a louder one within a few hz of it).
The sounds discarded are not out of the range of human hearing, you hear them, but ignore them in preference to other louder sounds at the same time.
Different lossy compression methods use different methods and thresholds to work out what should and should not be audible.
Therefore, what one codec assumes is important information may be discarded by another.
So when you re-encode with another codec it discards a different set of information.
Re-encoding with the same codec also produces a different result, as the quieter sounds that led it to discard information during the initial encoding are no longer present.
The net effect of this is that transients become smeared (temporal masking) and quiter sounds are rendered more inaudible (frquency masking).
When A/B tests have been done, the artifacts of this process are audible to most people.