Bit-Depth Audio and Harmonic Distortions

Audio Basis - articles about audio

Bit depth changing from 32-bit float or 24-bit integer to 16-bit integer causes some harmonic distortions. Read what's happens there.

Pure source signal contains sine wave 4410 or 4430 Hz is resolution 44.1 kHz, 32-bit float or 24-bit integer.

This signal is converted to 16-bit integer format.

File with sine wave 4410 Hz hasn't noise, but it has 3-th harmonic with significant level (-114 dB).

File with sine wave 4430 Hz has noise with a level below -133 dB.

This is the result of conversion of 32-bit float to 16-bit integer. Source signal is sine waves 4000 and 4030 Hz with sample rate 192 kHz:

Here we see input (without conversion or other processing) signal with frequency 4010 Hz in 16-bit / 44100 Hz:

3-rd harmonic with level -118 dB is appeared.

Here we see the input signal with 4030 Hz frequency in 16-bit / 44100 Hz:

No separate harmonics. Noise appears.

Conclusions

For 32-bit float to 16-bit integer conversion:

If a frequency of sine wave is close value that is multiple sampling rate, harmonic distortion appears as pure sine waves with noise same noise of HD formats (very low).
Otherwise, non-linear distortions are uniformly allocated (like white noise). Products at all frequencies have low level.

As final format 24-bit allows using full abilities of the modern DACs.

It gives impoving signal-to-noise ratio for quiet music signals, in comparison 16-bit resolution.