Audio Basis - articles about audio
Sometimes you need to adjust an audio file for a device or music format with CD quality (16-bit/44.1 kHz). It adds distortions due to information losses. These losses are not ordinary white noise. They as correlate with the music. It can cause additional listening discomfort. And there is technology that can partially solve the issue.
Dithering audio is a suppression method of non-linear audio distortions when bit depth is decreased. It is adding specially generated very weak noise to a digital signal. The dithering algorithm of AuI ConverteR 48x44 (alphaD) observes the signal behavior. It allows for minimizing noise influence on hearing and open minute details of a music.
We will not tell you about dithering mathematics. You can read it at another place. We'll just show what you get.
You have 24-bit or 32-bit (integer or float point) source. You need 16-bit source.
Pure modeled sine signal 24-bit or 32-bit (float point also)
Dithering is optional processing. If you use it, make sound check for better result. Dithering is most effective for music with low loudness. We should refer loudness relative maximum loudness of used resolution. In instance, the maximum loudness of 16 bit resolution is about 32000. The minimum is 0. You can follow your own perception when choose control level. It may be the quietest place of tested recording.
Now we make 24-bit to 16-bit conversion without dithering. When 24-bit signal truncated to 16-bit one, rounding to thr nearest integer is doing.
Pure sine after 24-bit to 16-bit truncation. Appear non-linear distortions
We see in this picture non-linear distortions ("wood of blue sticks"). Middle level of peaks of the "sticks" is about -135 dB.
Levels of these distortions increase and decrease together with the useful signal. In ther words, the distortions correlate with the useful signal.
No signal = No these distortions
The "digital sounding" is consequence of these "sticks" and their correlation with useful signal. And it is not "not very pleasant".
Now we will see 24-bit to 16-bit conversion with Audiophile Inventory's dithering audio alphaD.
In settings of AuI ConverteR 48x44 we set minimal value (Light) of Dithering mask depth... slider
Dithering audio settings of AuI ConverteR 48x44 - Light value (recommended)
That is result of 24 to 16 bit conversion
Conversion 24 to 16 bit with AuI's dithering audio alphaD
Now we see 2 things:
1. Mid level of peaks of noise about -145 dB.
2. The noise does not depend on the behavior of the signal. Now we loss "digital sounding".
If necessary we can add dithering intensivity for mask non-linear distortions by moving of Dithering mask depth... slider to right.
Let's look to generated without dithering source 16-bit sine.
Source 16-bit sine generated without dithering
"Wood of sticks" again :)
Below we will see how does slider Dithering mask depth... position to signal spectrum. How it mask the non-linear distortions.
We can mask source non-linear distortions with changing Dithering mask depth... slider position.
Audio file converter AuI ConverteR provides own dithering technology alphaD. The method smooth "16-bit artefects" and keep low noise.
It supported in all version of the program.
In configurable Modula-R edition, dithering is provided by "Base (core)" module.
The audio dithering is used to smooth non-linear distortions in 16-bit resolution.
See more...
The audio dithering is important because it allow significantly eliminate resolution-decreasing losses.
See more...
Also read: