Using Izotope To Clean Up Audio

Izotope is a great alternative tool to Audacity that you can use to clean up your audio when recording videos.

Recommended Size: Any

Ensure you're recording in a quiet location, without air conditioners, traffic noises, and so on. Try to record in a carpeted area, as you really, really don't want echo in your audio. That's almost impossible to clean up. Face the camera and mic square on.

The mileage of your mic will vary, but one thing to keep in mind is that some of the background static in your audio is simply from your camera's pre-amp, which may have to do a lot of work to pick up your audio. I believe most mics are not very sensitive and thus the pre-amp in the camera has to work hard to ensure the sound is audible. That's why a separate amp is always preferred, to take away the work load from the camera. Or if the mic is expensive enough, it will have its own. I believe most mics are at 0DB, but this gets boosted with other equipment.

Generally a cheap mic will record everything around it, and thus the pre-amp, and so you get a lot of background static. A shotgun mic can fix this issue, as the cone is tighter and can be pointed directly at you, but you just have to ensure it doesn't pick up your camera's pre-amp.

You can use a recorder and sync the audio later if you want, and there are programs that can help you do that, else doing that manually is a pain. I couldn't be bothered, so I'm using a wireless lavalier mic setup that has its own amp that attaches to the camera, and so far sounds amazing. And it better, the thing was $400!

The steps:

  • Leave 10 seconds of silence (where you simply don't speak) at the start of your video. This will be used as the noise profile later on. Ensure this silence is free from other noises, otherwise that will mess up your cleaning efforts. Try not to cough or breath loudly either.
  • Speak clearly and enunciate all your words, and try for a bell-shape pattern of speech, whereby your voice starts out low, reaches a high-point, and then tapers off toward the end of the sentence. This sounds better to us and makes editing easier. This will require practice. But as you do this, watch how other YouTubers do it, for example Phillip DeFranco. Notice even how he talks in his intro when he says "Sup you beautiful bastards, hope you're having a fantastic (day), welcome to the Philip DeFrance show and let's just jump into it". Notice how his voice starts and builds up and then tapers off, and this seems natural, but it really isn't. It will take a long time to build up this skill.
  • Put your video into your favourite program. Mine is Sony Vegas 15 Edit.

Thumbnail

  • See the spikes? Those are either me taking in a sharp breath or moving and effecting the mic. Try to avoid both for a cleaner sample. Now render out just the audio, I do it just as a .wav file.

Thumbnail

  • Obviously you need Izotope. Import your audio and you will get a similar screen (My version is a bit older so yours may look a little different).
  • Okay a few things to notice here, as the program is not exactly intuitive the first time you use it. You must zoom in as you have to be precise here.

Thumbnail

  • Select around 10 seconds of pure, clean audio. You can see here I am avoiding selecting an area with spikes, which were most likely caused by sharp intakes of breath. It's 9.1 seconds, which is fine.

Thumbnail

  • Click De-noise and then click 'learn' This now builds up a noise profile, which will be compared to the rest of the audio.

Thumbnail

  • Close the screen and then select the entire audio. But be careful that you actually select all the audio, that you zoom out again!
  • Now comes the settings...never use the automatic settings as this won't work. Always use manual, not adaptive settings. You will have to tweak the settings as required, but I will show you the settings I use; which are for a small room, carpeted and no outside noise interference.

Thumbnail

  • Now wait for it to process and then listen to your new audio! If it's not good, then undo and try other settings, but if it is, you can now render it into a .wav file and import again into your video editing program.

Related