Red Dog Music | Oct 9, 2018 | 0
What is dithering?
Just gone to export your latest main room anthem from Ableton Live and noticed that little ‘Dither Options’ box? Not sure what it is or when to use it?
What is dithering?- Quantisation
We had a post a while ago that [very] briefly described how digital audio works, but we didn’t discuss dither. What an oversight! We shall address that right now…
You know how when you play a drum beat in live in Maschine and it tightens things up by snapping those hits to the grid? That’s quantisation. And that’s what happens to your analogue waveform when it’s being converted into digital. It gets quantised to the nearest level, distorting the waveform.
In your audio interface’s analogue to digital converter, dither is applied to get around these errors by taking advantage of the wonderful world of statistics. Dither is really just adding a quiet bit of random noise into the system.
Imagine you have an analogue waveform with a level of an arbitrary 0.6, and the available levels to which it can be ‘snapped’ are 0 and 1. If the signal was just rounded, it would always be 1, but add some slight level fluctuations by adding the dithering noise, the signal might occasionally be quantised to 0. If it snaps to 1 three times out of five, and to 0 twice out of five, you get an average of 0.6.
And that’s (sort of) how dither works, but what about those mix bounce-down options we started with?
Why should I use dither when bouncing down my mix?
Applying dither when exporting audio is the same sort of thing as dithering at the analogue to digital converter, but a bit different…
Inside your DAW, there is a lot of maths going on. Whenever you change the level of a sound, add some reverb or some compression, your DAW has to do some sums to do on all those audio samples.
While you might be using 16-bit or 24-bit recordings or samples, your DAW will be using a lot more bits than that to keep things as accurate as possible right through from initial sound to final output.
For the sake of argument though, let’s just say that the audio in your DAW is 24-bit, but you want to file to burn a CD, which is 16-bit. How do you get from 24 to 16? The obvious answer is to subtract 8, but if you just take those extra 8 bits away – called a truncation – then you add distortion to your audio. This distortion can be particularly audible with low-level signals, such as reverb tails.
By adding dither, we introduce that random noise again that brings our statistical friends back to the party and use mathematics and probability to improve how things sound.
Bob Katz’ book Mastering Audio, the art and the science, has an excellent description of dither in chapter 4, and is definitely – along with the rest of the book – worth a read!
By adding noise, dither does increase the noise-floor of your system but, counter-intuitively, means that we can hear sounds that would otherwise be so quiet that we wouldn’t be able to (they would otherwise be quantised to a level of 0).
And back to our mixdown settings in our DAW then. If you’re exporting your audio down to a shorter word length (16-bits for CD for example), then you should dither as the last step in the chain.
But what about those dithering options? Not all noise is created equal. Some dithering algorithms are ‘noise-shaped’ to improve performance. Rather than the noise being equal across all frequencies, some lower the noise level in the frequencies to which our ears are more sensitive allowing the dither to work, but making the noise less apparent.
Different dithering types have been created for different applications so, while standard dithering types should generally cover most bases, it never hurts to experiment!
Red Dog Music is the UK’s friendliest musical instrument and pro-audio dealer. Between our 5000 square foot Edinburgh shop filled with an incredible range of products, and our London showroom in Clapham specialising in high-end instruments, dj and pro-audio, Red Dog Music has you covered from north to south and from performance to playback.