If I run the Equalization effect repeatedly does this degrade the file each time or is the original file used to generate the final mix each time? In otherwords, is this like a jpg file compression that degrades photos every time it is run?
OB
If I run the Equalization effect repeatedly does this degrade the file each time or is the original file used to generate the final mix each time? In otherwords, is this like a jpg file compression that degrades photos every time it is run?
OB
When an effect is applied it changes the audio data. If you apply the effect a second time, then it is that changed data that is processed not the original. This is why it is best to work in 32-bit float format when processing audio. 32-bit float is extremely precise so it is possible to apply thousands of changes to the data with negligible degredation.
OK, so now I expose my nubeity - how or what is 32 bit floating?
OB
32-bit float is a shorthand way of saying that Audacity performs all its mathematical calculations using 32-bits (binary digits - 0s and 1s) floating point arithmetic (numbers with a decimal point and digits after the decimal point) rather than 16-bit integer arithmetic (whole numbers only -no fractions). This gives greater precision and keeps the overall technical quality of the data as high as possible. That doesn’t mean that it gives superior sound quality. If you have a bad sound recording, Audacity preserves that badness to a very high mathematical precision. Similarly, Audacity tries its hardest to keep a good recording sounding good as you apply your choice of effects to it.
The default “Quality” setting in Audacity is to use 32-bit float, and as PGA wrote Audacity works internally in this format.
If you look on the left end of an audio track (just below the track name) the sample rate and bit format are shown. By default this is “44100 Hz” “32-bit float”.
(see also: http://manual.audacityteam.org/man/Track_Drop-Down_Menu )
So, Audacity doesn’t use the original source file, it processes and reprocesses the “working file” every time an action is done.
So a good workflow be to import the raw file (mine are usually performance sets with 8 to 10 songs), pull out a segment (song) to process as a separate project, run effects, such as Eq., then export to a final file or finished track. If repeated tries are done I should use a copy of my original file so that i am working with my original data each time until I am happy with the final effect.
OB
That’s one way of working, but remember that for as long as the project is open you can always “Undo” operations. (Ctrl+Z or select “Undo” from the Edit menu).