BNEffector sets sound level by following steps.
1. Transforms original file and stores it in floating-point format.
2. Multiplies data by a factor. The factor determines sound level.
3. Converts the floating-point data to 16bit or 24bit integers (see note).
If the factor is too large, sound clips. If too small, the resulted wav file does not use sound system capability effectively, increasing noise and distortion.
There are two ways to set the factor.
· Individual sound level optimization.
BNeffector uses different factor for each track. Sound level of each track is maximized without clipping.
· Total sound level optimization.
BNeffector uses the same factor for all tracks. Sound level differences between tracks are retained. If tracks are related each other, for example all tracks are come from the same album, use this mode.
If some tracks are jointed, BNeffector uses the same factor for jointed tracks even if individual sound level optimization is selected. See “Joint Tracks” for detail.
(note) This process is called re-quantization. BNEffector uses dither for re- quantization. Noise shaping is not available.