Home / Opinions / Fixing Noisy Audio / Subtraction and Recombination with Video

Subtraction and Recombination with Video

Next, choose the Noise Removal filter, found under the Effect menu entry. In the resulting dialog box, press the Get Noise Profile button. This tells the software what the background noise is supposed to sound like, so it can be differentiated from the sounds you actually want to hear.

You now want to use your noise sample to process the entire sound file, so highlight the whole waveform. Choose Select>All under the Edit menu, or simply press CTRL+A. Then call up the Noise Removal filter again. This time, press the Remove Noise button. This will subtract the background noise specified in the sample taken in the previous stage from the entire audio file. You can also use the Preview button to see how the effect will sound, and adjust the Less/More slider until you find a happy balance. Too much noise removal will make your audio tinny and compressed; too little and you will still hear the background noise, although it will be hard to escape some detrimental effects. We usually find the middle setting too aggressive, so pushing the slider towards Less is typically a necessity.

As you can see, we now have an audio waveform with some regions of almost complete silence. All the obtrusive background noise has been eliminated. The side-effect is a slight sense of compression, which will make speech sound a bit like it’s underwater. But this will be much better than having your interviewee’s voice drowned out by hiss. With very heavy background noise, however, this effect can become overbearing, so you will have to compromise between the two.

Now the noise has been removed from your audio file, use the File > Export as WAV command to save out the results. You will now need to recombine it with the video.

If you have been using VirtualDub to output the audio, you can return to the software and load your original AVI file. Under the Audio menu, choose WAV Audio… instead of Source audio. Then select your treated audio file. You can then create a new AVI in your chosen format with the audio integrated.

Otherwise, the method for recombining the audio will vary depending on your chosen editing app. If your editing software keeps its audio and video tracks separate, you could simply overlay the new audio over the old waveform by dragging it on top. If this isn’t possible, you will need to drag the file to another audio track instead. Then use your software’s audio mixer or rubber band controls to reduce the volume of the original audio track to zero – or mute the track entirely, if this option is available.

In all cases, make sure you line up the beginning of the treated audio file with the beginning of the video file, otherwise they will be out of synch. Similarly, whichever method is required with your software, be aware that your video and audio are now separate, non-associated files. So now if you move one of them along the timeline the other one will not follow. Each time you move the video, the audio will need to be moved manually and lined up again. It’s a hassle, but worth the trouble for the benefit of clean, noise-free sound.

comments powered by Disqus