If you’re a beginner producer, you might not have heard of dithering just yet. If you’re an experienced engineer, you probably have questions about how dithering can take your sound to another level.
Dithering has taken on several different uses over the last few decades as music production and presentation have taken on different forms.
At one point, music was all recorded through analog devices like tape machines and then converted to digital for mass production. Music now is more than likely recorded and mixed in a virtual space, always remaining digital through each stage past the microphone. But we still convert files from high-quality WAVs to lower quality MP3s to make up for streaming sizes.
This is why we still need dithering.
So what is dithering? What does it do? The goal of dithering is to make your tracks sound just as good as they do on your sound system before and after converting to a lower resolution.
Let’s talk about what dithering is.
Dithering is the process of adding random noise to your track to reduce the amount of digital distortion caused by the reduction of bit rate and file format in what’s known as quantization distortion.
Sound complicated yet? That’s because it sort of is.
To fully understand it, we need to explore a few other slightly complicated audio terms, like bitrate. Let’s break down how it affects the sound of your music, what digital distortion is, and why we need to convert our audio files to smaller sizes in the first place.
Bit rate is the value of steps available in each sample of your music. Sample rate and bit rate work hand in hand to create either high-quality or extremely low-quality audio.
The bit rate usually found on CDs and streaming services is 16. That means that for every sample, which is a microscopic vertical slice of your audio, there are 16 steps in the waveform.
Picture a line moving diagonally from bottom to top as one sample. But our line is a straight A-to-B linear line–it’s a series of steps. Less steps mean getting further from a straight line. The more steps, the closer to a straight line.
So if we have 16 steps going from top to bottom, that’s much closer than if we had only eight steps. But 16 is noticeably less of a straight line than 24.
In professional recording studios, most tracking is done at a 24-bit rate. This means the incoming audio is as high quality as it can be entering the system.
But this amount of data can be huge and won’t fit on streaming services without weighing down the whole thing.
So instead, we convert these tracks to lower bit rates to conserve space, from 24 to 16. We lose a bit of detail regarding the audio quality, and audiophiles would say it makes a major difference.
But most people honestly can’t tell unless they have high-quality equipment, and the amount of space saved is typically worth the minor reduction in quality.
However, how we convert to lower bit rates and formats affects the quality, and by using dither, we can control some of these variables to get the best quality.
The last bit of information to understand before we answer this (we promise) is something we mentioned before called digital quantization distortion, and no, we aren’t talking about Axen’s wicked “Droid Bass” sample pack, although that is some sick, intentional digital distortion.
Essentially what this is is the rounding error that digital conversion tends to make when reducing bit rate, or transferring from analog to digital.
Because we’re reducing the steps on the line based on the line itself, sometimes the digital conversion will have to round certain values to the closest step.
Imagine that our line is no longer straight and has a bit of a curve to it. At 24 bits per sample, we have a good idea of the curve, but by reducing the step value to 16 some of the steps might end up between the original 24.
While some distortion can be excellent, especially when emulated off of analog distortion, this rounding leads to unpleasant digital distortion, and when mastering our audio, we’re obviously trying to avoid as much unpleasantness as possible.
Dithering, as we said, is the introduction of random noise into the mix. Don’t worry–we aren’t talking about anything loud enough that you can hear, but your computer certainly will.
The low amount of noise reduces the dynamics of your track and therefore reduces the range of curves within your waveform. This means your lines have less potential for quantization distortion because there’s less variation to round up or down.
Basically, it’s like implementing your own controlled level of distortion instead of dealing with whatever random distortion the computer decides to kick in while converting.
This might sound a bit backward but trust us: You would much rather have an extremely low level of noise that you control in your track instead of the random, digital distortion byproduct that conversions can produce.
Dithering allows us to produce high-quality sounding tracks at small file sizes, so when you use Aaron Richards Talk About It vocal pack, you can expect to hear the crisp samples on streaming services the same way to come out of the pack.
There are several dithering tools you can use when decreasing the bit rate of your audio. Some might even come with your DAW if you use certain software that professionals call the industry standard.
Most mastering plugins will come with dithering that you can shape to fit your track best. Maximizers and limiters will be a good option. Make sure to do your research and find a tool that will be high quality and the right fit for your system.
Although you technically can, we wouldn’t recommend using iTunes to convert your music to the proper format for streaming services.
Especially if you’re looking to upload the sick new track you just made with BIJOUS new Street Knowledge sample pack to Spotify and Apple Music, you want your audio to be already converted to their preferred bitrate and file size; otherwise they may try to do it for you, and they aren’t going to be particularly gentle about it.
Dithering has a lot of complicated math and terminology behind it. But it’s essentially the introduction of a sound that you control that reduces the amount of distortion that you don’t control.
Remember, we really aren’t talking about an audible amount of noise. Not to humans anyway.
We want to reduce the amount of digital distortion in the mix at all steps. Though don’t think that dithering will save your tracks from clipping if you have them too high.
Do your research and find a reliable mastering tool that can implement dithering for you. You might not immediately notice what it’s doing, but it’s going to save you some headaches when it comes time to finish your tracks and get them ready for release.