FEDL22660-01
ML22660
●
Speech synthesis algorithm
This LSI contains five algorithm types to match the characteristic of playback sound: 4-bit ADPCM2 algorithm,
HQ-ADPCM algorithm, 8-bit non-linear PCM algorithm, 8-bit straight PCM algorithm, and 16-bit straight PCM algorithm.
Key feature of each algorithm is described in the table below.
Speech synthesis
algorithm
Compression
rate
Feature
4-bit ADPCM2
1/4
LAPIS original 4bit ADPCM algorithm is improved. Better
followability to the waveform improves the sound quality. Suitable for
human voices, animal crying, and natural sounds.
HQ- ADPCM
1/5
4bit ADPCM algorithm is improved. Adopting variable bit length
enables high sound quality and high data compression. Suitable for
sound effects with sharp changes in waveforms or for pulsed
waveforms.
8-bit non-linear PCM
1/2
This algorithm enables playing back a sound with 10-bit equivalent
quality in the center of the waveform. Suitable for low-amplitude
sounds that are easily distorted.
8-bit straight PCM
1/2
This algorithm has excellent followability to the waveform in all
sound areas. Suitable for sound effects with sharp changes in
waveforms or for pulsed waveforms.
16-bit straight PCM
1
This algorithm has excellent followability to the waveform in all
sound areas. Suitable for sound effects with sharp changes in
waveforms or for pulsed waveforms.
21/116
Summary of Contents for LAPIS Semiconductor ML22660
Page 107: ...FEDL22660 01 ML22660 107 116 ...