Hello. Do you apply dither when you render your final audio tracks?
I don’t because I’m always mastering at high res, it hasn’t occured to me that I ever went like “Whoa, sounds bad, I could use some dithering”.
The rule of thumb is to apply it only when you’re exporting music at a low resolution, just once before you export the final file (if you bounce already rendered tracks once or more into a final master, otherwise you will have layers upon layers of added noise).
It’s like visual dithering in files such as GIFs: dithering is when you introduce noise to make something less distorted. Sounds like an oxymoron, but it works! It’s better because we humans can judge it as better (thanks to psychoacoustics); in fact, “less distorted” to us humans can just be read as “better”.
Pic 2 and 3 have the same colors and information encoding limits, but 3, where some pixels are scrambled (thanks, noise!) looks better. Why? It’s more spread and the change between colours is less harsh and noticeable. Spreading the dots helps make the values more pleasing to the human eye, just like in the audio world with our ears: the “blurring” caused by the limits of our perceptive apparatus.
If you’re making music at low bitrates like MP3s for streaming on Soundcloud, unlike what could be your Bandcamp with 24 bit WAVs, or for a toy with a tiny janky chip, you may want to apply dithering. Otherwise, you may not want to. Next time try to export a short sample at lower and lower bitrates as you go, once with dithering, once without, and see when dithering starts helping your ears make it sound less bad and compressed. You may never need it at all.
This article is great. The ending also says what you may have realised while reading: “[…] as high fidelity recordings become compatible with more streaming services, then the need for dithering will become non-existent”.
Thanks for your interesting answer and the link.