I recently started getting into home-recording, primarily to do voice-overs on videos or to record screen tutorials for my parents. Of course I don’t have a quiet sound studio at my disposal, so my study has to do. When I turn off the radiator and close the door, it is actually pretty quiet except for the fan on my iMac. That plus background recording noise won’t go away and is quite distracting on plain voice recordings. Some searching online pointed me in the direction of using a noise gate filter:
In its most simple form, a noise gate allows a signal to pass through only when it is above a set threshold: the gate is open. If the signal falls below the threshold no signal is allowed to pass: the gate is closed. A noise gate is used when the level of the ‘signal’ is above the level of the ‘noise’. The threshold is set above the level of the ‘noise’ and so when there is no ‘signal’ the gate is closed. A noise gate does not remove noise from the signal. When the gate is open both the signal and the noise will pass through.
In Audio Hijack Pro, it’s pretty easy to set up:
- Start a new recording session and choose your audio interface as the input and output
- Click on Advanced… and choose your microphone input channel for both left and right input. Choose channels 1 & 2 for the output. That way, the mono microphone will be heard on both the left and the right
- Hijack the signal
- In the effects patch, place an mda SpecMeter to show the level of background noise
- Add an mda Dynamics filter, turn off the compressor and turn on the gate
- Now place another mda SpecMeter after the Dynamics filter
- Be quiet, and observe the noise level in the SpecMeter. Select a level slightly larger than your background noise as the threshold. Any sound quieter than this will not get recorded.
- Choose a very short time for the attack. I usually use 1ms. This is how long it takes the gate to open up once the threshold is surpassed.
- Choose a longer value for decay. This determines how quickly the gate closes again after the signal falls below the threshold. A value of 0.5 seconds works for me. If you choose this too short, the audio will get cut off rapidly after you stop speaking, it may even cut out between words and cause the audio to sound choppy.
Now record your audio, listen to it, and adjust the parameters if you’re not happy. When no word is spoken, the gate will close and absolutely nothing is being recorded such that your recording is clean without a hiss. When you speak, the gate opens and your sound is being recorded. There will of course still be background noise then, but it won’t be as noticeable because there are other sounds.
For comparison, do a recoding without the gate filter and then you will clearly hear the difference in the recording. I find that even the noise during the recording is more pronounced without a gate filter because it seems to continue from the time when I am not speaking to when I am speaking. With the gate, the noise comes in at the time as my voice and therefore is not as obvious.