OS: Windows 7 Ultimate SP1 x86-64
Audacity: 2.0.6 Installer
So I have a file which contains the speaker output, and I have another file which contains the speaker output and voices. Is there any easy way to cancel out the speaker output in the second file? I tried using noise removal, but it doesn’t work very well. I’ve also tried matching it up manually, by inserting extra time before the track, but that takes me hours to align the files. Is there an easier way?
Vocal Isolation.
I bet there’s a failure reason you haven’t thought of. Are they MP3 files? MP3 compression and sound damage changes with content, so the music/background part of the two files isn’t exactly the same. That may be why you can’t get them to cancel manually. They will probably never cancel.
Pick a portion of both files with music only. Repeatedly magnify them (Drag-Select > Control-E) until you get down to samples (attach 1). Then use the Time Shift Tool (two sideways black arrows) to gently shift one track (attach 2). If there are no places that match to that kind of accuracy, then there is no show.
Koz
that takes me hours to align the files.
…and since there are 26,000 files, it may take me a very long time. What’s the actual job and how involved and complicated is it?
Koz
Noise Removal only works on long sound events that don’t change. So microphone hum and air conditioner noise is fair game, but dog barking, car starting up, talking and music is not. So we keep drifing back to finding out what the job is.
Koz
So I have a file which contains the speaker output, and I have another file which contains the speaker output and voices. Is there any easy way to cancel out the speaker output in the second file?
Probably not… This can ONLY work in very-special cases where you have the identical digital-data in both files.
For example, if you have a singer on one recording/track and a guitar on another recording/track, you can digitally mix them and then later subtract-out the vocal or guitar. But,if you have the original guitar, you probably also have the original vocal, and there’s no reason do this. If you no longer have the separate-original guitar track, you cannot record the guitar again and subtract the new guitar recording because it’s digitally different.
If your files are not already time-aligned, obviously they are different. If I record myself saying “Hello” twice and I subtract one from the other, it will NOT sound “subtracted”… In fact, subtracting two different recordings sounds exactly the same as adding (mixing) the recordings and it will sound something like me and my imaginary twin saying “Hello”, together.