A better sine sweep test

As I learned from Dan Worrall, a sine sweep stays at a constant loudness throughout the whole frequency range in a way that doesn’t occur in most recorded sounds.^[ Linear, Logarithmic, Exponential & Perspective] To solve this problem, I created a sine sweep test that decreased in volume the higher it went, in accordance to how real-world sounds decrease in volume at the higher frequencies. Because I know that pink noise is found in many sounds in nature,^[ Pink noise - Occurence] I created a sine sweep that decreased in decibels per octave at the same rate pink noise decreases. To do so, I had to answer the question, At what rate does pink noise decrease in volume?.

If I’m at 48kHz sample rate, the highest frequency I can replicate is 24k. So I worked backwards from there, dividing each decade until I reached a suitable lowest frequency. The sine sweep would start at 2.4Hz and linearly decrease in volume by 40dB once it reached 24kHz. In iZotope RX, I created a sine sweep that starts at -18dB and ends at -58dB.

Satin was my first test subject with this.

I decided to start the test tone at -16dB, which gives a 2dB margin on either side of my -18dB RMS gain stage target. ! [[ Screen Shot 2022-01-07 at 11.02.23 PM.png ]]