Audacity does support some low bit-depth formats, such as Microsoft ADPCM (4-bit), u-Law (8-bit) and A-Law (8-bit), but note that these do not correspond directly with low bit-depth "linear PCM" because they are encoded differently.
Probably the best way to illustrate the impact on the sound quality of reducing the bit-depth of PCM digital audio is to "quantize" the sample values. The number of "bits per sample" determines the maximum number of possible sample values.
For example, for 8-bit PCM, the sample values must be between binary 0000 0000 and 1111 1111. In decimal that is 256 possible "levels" for each sample value. This is usually arranged as either 127 values above zero and 127 values below zero, and zero, totalling 255 values, or as 128 values below zero and 127 values above zero and zero, totalling 256 values. (these two schemes are implementation differences and have very little effect on the resulting sound).
Audacity has a scripting language built in, called "Nyquist". Nyquist is a programming language based on LISP, that is designed for generating and processing audio. One of the available commands is "QUANTIZE" http://www.cs.cmu.edu/~rbd/doc/nyquist/ ... l#index565
To use this command. select some audio, then open the Nyquist Prompt effect (http://manual.audacityteam.org/man/nyquist_prompt.html
To quantize the audio to the equivalent of 4-bit audio (binary 0000 to 1111 = 0 to 15 decimal) we would want 7 values above zero, 7 values below zero, and zero (total 15 values), and we can do that with this code:
- Code: Select all
(quantize *track* 7)