You hopefully should know when to dither and why. Hopefully, because thereʼs still a lot of misunderstanding on this topic among some even experienced engineers!
If you want to know a lot more about dithering topic, read an entire article on our blog page.
All you need to know is that there’s real world and virtual world. The real world, which is also the distribution one, works with fixed point arithmetics (usually 24- and 16-bit). That’s a derivative of digital equipment (like f.e. digital to analog converter) most used interface (AES/EBU), that has 24-bit word-length limit.
And there’s virtual world. It’s been using floating point math for some time now. This is DAWs world today. The exceptions are some old ProTools versions being still in use occasionally, that feature 32bit floating point engine with 24bit fixed point mixer (not a dithered one).
These worlds meet each other when you for example export your DAW work for distribution.
The export procedure requires your attention. DAW can automatically change format without any warning. This is where potential truncation might happen! It will do for example on 32-bit (DAW/float) to 24-bit (fixed) change.
This is when dithering needs to be applied. Why? To avoid truncation distortion. You can hear how something like this sounds on our blog page here.
The other scenario is when you for example check option in your DAW to record files in 16-bit instead of 24-bit. Please remember, your analog to digital converter will very likely produce 24-bit words.
If you prefer to record to 16-bit files for some reason, you need to apply dithering then too!
Yet another scenario, pretty common today, as collaborations and audio files exchange iterations are growing. People often save/export audio as 24-bit and/or 16-bit a little frivolously from their DAWs. With each save like this dithering needs to be used! When audio is being truncated numerous of times, truncation distortion stacks up quickly and becomes pretty audible.
If you often collaborate, please send this information within your team.
Link to this FAQ/info itself: https://www.sonicsciencelab.com/quick_faq/to-dither-or-not-to-dither/
With that said, it all comes down just maintain bit-depth/word-length across all you do with audio with awareness. in a conscious way.
#1: It is always best to deliver mix file at your DAW native bit-depth/word-length to your mastering engineer. So, check your DAW documentation. File sizes and network transfers are no longer an issue in terms of saving some space/bandwidth.
#2 (actually a small warning): It might happen that you find some engineers over the internet, discouraging from using dithering. In vast majority of cases opinions are coming from those who have no idea what dithering is, why it needs to be engaged and when. People naturally tend to ignore things they can’t understand. Be aware.
Some argue that dithering brings noise to audio. The fact is that dithering noise is being applied at very low levels. In order to have it becoming an issue, audio must have been undergoing a lot of iterations. A lot. Truncation distortion accumulates way faster!
Please enter your name and email address below.
0
