Dither - To UV22 or Not to UV22: That Is the QuestionDigital audio’s weakness is its ability to accurately reproduce extremely low-level signals. As the amplitude decreases, there are fewer and fewer bits available to accurately represent the audio wave.
In an analog system the audio simply sinks into the noise floor. Even when covered by noise, the waveform is still audibly intact. In a digital system, low level audio simply becomes inaccurate and sounds really bad. A few bits at the bottom of the amplitude range aren’t capable of faithfully converting the analog information to digital, let alone reinterpreting the digital data into analog variations in air pressure.
Dither is a low-level noise that combines with your pristine audio at low levels to help keep enough digital bits active to replicate an accurate waveform throughout the conversion process from digital to analog to digital—go figure that it takes noise to make a system work that has no noise.
UV22 is a very good-sounding type of dither that makes a way for more accurate reconstruction of this low-level audio. The frequency of the UV22 dither noise is lower than the conventional process and therefore provides a more transparent reconstruction.
As a practical application, if your music is to be mastered by someone else in a professional mastering facility, you might not need to use the UV22 process—too much dither can accumulate too much noise. If you’re seeing the project through to the end you have three primary options:
•Don’t dither. If your music is always loud—loud, loud, loud—you might not need to worry about dither because dither really only helps audio quality at low levels. However, if there is a fade out in your song, then of course everything isn’t always loud and you’ll want to take advantage of the UV22 process to feather the fade out with excellent audio clarity.
•Dither once. The conventional approach has been to avoid dithering until the final version is mastered (prepared for duplication). In this case, wait until the final mix is printed to the replication master, then use a Normal UV22 setting on the Stereo I/O in the Setup window. Dither only once on Normal setting; dither noise is cumulative so too much dither equates to senseless noise.
If your music contains only a few low-level passages try using the “Low” setting on the Stereo I/O. “Low” equates to a lower level of noise and, though it might not provide as much smoothing of low-level audio, it might be quite sufficient—and less noise just seems like a better choice wherever appropriate.
•Dither twice. If your music contains a wide dynamic range, you might not be able to resist the urge to use dither as you record to the multitrack. In this case, plan to use the UV22 process twice: use it once on the multitrack tracks—on the “Low” setting; use it again on the final printing of the master take—on the “Low” setting. If you use the UV22 process twice, using it on Low setting both times will closely equate to the dither noise resulting from using it once in “Normal” position.