Replicating Background Noise – Susumu Yokota’s, Saku

It may seem strange to want background noise in a recording, but it gives it a certain realism as opposed to the cleanliness of a digitally-mastered DAW creation. I like the sound I hear in the background of Susumu Yokota’s Saku from his Sakura album, so I was able to replicate it using common pink noise and equalization. Here’s how:

I’m not sure why, but the noise in Sasumu Yokota’s track, Saku, sounds good to me. It gives his recording a depth and realism that I want in mine. I can hear it clearly at the beginning of the track. It is there a second or two before the first instrument plays.

First, I downloaded the purchased track from Bandcamp. I used my audio editor (Edison in my case) and cut out the first second, trimming it so I had only the background sound and not the silence before or the instrument after.

I wanted to replicate it, and it is noise, so I thought that maybe different colors of noise would work, or even pre-recorded studio noise. I tried these by playing them either as clips in my DAW or by using Denise Audio Noize 2 and loading them in. None came close. This noise was somehow unique and special.

So, I played the noise and side-chained it into a channel that I had a pink noise generator (Denise Audio’s Noize plugin) on. Following the Noize plugin in the chain, I had Fab Filter’s Pro-Q3 equalizer. The Pro-Q3 was receiving the side-chained signal I wanted, while processing the pink noise.

By using the EQ Match feature, I let Pro-Q3 characterize and filter the pink noise to match the one from the recording, and it worked.

I now had the noise I created and equalized, sounding nearly identical to that of the album noise on Saku. But, it was mono because I used only one channel, and although it was a stereo channel, when Pro-Q3 EQ Matches, it works on the stereo signal but produces only one EQ curve applied to both left and right channels.

But the noise was different in each channel on the original, so to get that I just had to process the left and right channels separately and do the same procedure.

Here you can see subtle differences in the matched EQ curve for each channel. Left is on the left and right is on the right.

Now the sound was nearly exact and I can easily capture as long of a segment of it as I need to use in my own music.

This could be done with any noise that you may want to replicate and use in the background of your recordings.