I taped a session on a cafe, the thing is that we had a speaker on the roof playing some music. I’ve been searching on this forums but nothing works. I tried with Noise Removal and nothing.
What should I do?
Give up. Record it again. You picked one of the non-recoverable damages.
The four horsemen are echoes, overload (clipping), underload (noise), and acoustical interference (your problem). Almost always fatal.
Koz
Maybe a way to “minimise” the music?
<<<Maybe a way to “minimise” the music?>>>
Nope. This is the same problem of people who want to pull only certain segments of a mixed, produced, finished musical performance. Unless it was done with a dramatic stereo microphone system or produced in a specific type of stereo, you’re stuck.
Can we assume you were using one single microphone for the voice capture?
Koz
There is always the level keying software.
http://audacityteam.org/forum/viewtopic.php?f=28&t=5052&start=0&hilit=keypex
Koz
Yes, only one mic. It was actually a digital camera recording video and then I extracted only the audio.
The “Truncate Silence” did not work. No difference at all.
No, there’s a software package based on the old “Kepex” machines. I can’t remember what it’s called, but it waits until the average sound level has dropped below a set point and then either takes the sound to zero or reduces it according to a set amount.
Hello ssssssss How sssssss Are ssssss You? ssssss
Turns into the same sentence without the hiss or background noise. But. If the speaker looks down or does anything to change their voice, the effect fails. It also fails if you don’t get the settings right and start chopping off the very beginning and end of each word.
It’s not pretty.
I only bring that up because technically, there is a software tool that might help, but everybody who tries it gives up because it makes a slightly noisy track into science fiction voices. Trading one problem for another.
Koz
GGate does have a “fade” setting so can be set not to chop off the ends of words.
[BTW don’t set the attack on GGate to zero, use a minimum of 0.01for attack otherwise it clicks]
You can also mix the gated version which has total silence between words, with a little of the original ungated noisy version
so the end result does not sound so "science fiction ".
<<<so the end result does not sound so "science fiction ".>>>
You would think it would be easy to have a fader so you can control how much of the effect is introduced. It would be the rough equivalent of the “DB Reduction” setting in Audacity 1.3 noise reduction.
Maybe the next version.
Koz
Thanks for the help.
I guess I’ll just leave it like it is.