If you have never heard of the word dithering, you are in the right place. We’ll explain what it means and how it pertains to sound.
If you are doing sound projects, there may be times when you need to know when to dither. We’ll explain what it means and how it all works.
After reading this, you’ll probably feel like an audio expert. But there are plenty of other terms that you still need to learn. So let’s get right to it and talk a bit more about dithering.
What is Dithering?
Dithering is when you are adding noise to a signal. The purpose of this is to mask harmonics that are in a higher order. It will also lessen the perception of distortion.
Distortion – What to Know Before Dithering
If you want to know what dithering is, you’ll need to understand distortion and what it does. Distortion can be a blessing or a curse when it comes to audio. Some of it may be pleasant, while others may not be.
Distortion can either belong in an audio recording or not. If the harmonics are low, then the distortion will be good. Specifically, these distortions are based on tape, tube, or transistors.
The higher-order harmonics are the ones that can make distortion an unpleasant thing. That’s because the frequency is higher. And it can cause a shrill noise making it unpleasant to your ears.
Next, we’ll discuss another term that you need to understand pertaining to dithering. This is known as quantization.
What is Quantization?
If you are converting an analog audio signal to one that is digital, that is quantization. This can be done using a piece of equipment known as a quantizer. Now, let’s dive into the other stuff like quantization noise and distortion.
Quantization Noise
Quantization noise is a type of error that occurs when an analog signal is being encoded into a digital signal. These errors usually happen when the process of rounding occurs.
Quantization Distortion
This occurs if a waveform is rounded or squared off, causing inaccurate representations of the amplitude. You’ll usually see this happen with the lower bit depths as the number of bits available is limited. Therefore, the bits won’t be rounded as they should be.
This will cause higher harmonics to occur, eventually making distortion something worth getting rid of in an audio recording.
How Does this All Tie Into Dithering?
So now, you’re probably asking, what does this have to do with dithering? When you introduce noise to the signal to cover the harmonics created from quantization distortion, that’s dithering at work.
What happens is that the harmonics are masked and become more randomized. This will make distortion hard to perceive.
Dithering is swapping the harmonics from quantization distortion and exchanging it for noise. So now that you know the dithering function, let’s discuss when you should use it.
When Should You Use Dithering?
The only time when dithering should be used is when you need to cover up any quantization distortion. You’ll commonly find quantization distortion during the post-production process of any recording.
If the float file is a 32-bit or even a 24-bit, you can reduce it to 16-bit depth if necessary. Dithering can also occur during the exporting process as well. If you have a limiter, keep in mind that there is a dithering function available when you need it.
When Should You Not Use Dithering?
There is a time when you don’t need to use dithering. That’s when there is no quantization distortion present. Even if you use dithering when it isn’t needed, it won’t affect the signal in a large way.
But still, you should avoid adding unnecessary dithering. The last thing you want to do is add noise to a signal when it’s not warranted. But, on the other hand, if you overdo it, the noise will become loud enough to the point where your listeners will have a less than ideal experience.
Many advanced sound engineers will do their best to avoid unwarranted dithering. They are trained well enough to know when to and not to use it.
The Three Types of Dithering
There are three types of dithering to be aware of. You have Types 1, 2, and 3 POW-r. POW-r stands for psychoacoustically optimized worth length reduction. Let’s break down each type:
- Type 1: This is used when you are working with range mixes that are low-dynamic. Most pop and rock recordings will use this type of dithering. You don’t need to use any noise shaping since the noise will not be affected.
- Type 2: This type will be used if speech or spoken word are involved; this can be used for masking quantization distortion while amplifying the noise from 2kHZ to approximately 14kHZ or thereabouts.
- Type 3: If you are recording orchestral music, this is the type you’re going to use. This is similar to the noise shaping used in Type 2. The EQ curve to the noise will usually be greater than other musical recordings outside of the orchestral genre.
Conclusion
We hope you found this guide about dithering to be informational and understanding. You’ll know exactly when and when not to dither if you are a sound engineer. We also hope that you understand the science behind how it works.
Distortion can be a good thing. But it can also be a bad thing depending on some factors. If the distortion is good, you won’t need to dither at all.
Practice dithering with sample recordings whenever you get the chance. Then listen to the sound files (both pre-dither and post-dither) in their entirety so you can make comparisons between the two.