Lernout & Hauspie Speech and Music coding capabilities cover a wide range of compression rates starting from 2.4 Kbps up to 16 Kbps and require the use of less processing power than many commercially available speech compression products. The company has been able to achieve this wide range of compression rates by implementing three distinct coding techniques:
Sub-Band Coding for high quality and a low rate of processing power.
Code Book Excited Linear Predictive Coding for mid-range applications where quality and high compression are required.
Variable Bit Rate encoding enables the best match between the bits required to represent a given piece of speech and the available bandwidth.
Key Pentium® II processor features/performance benefit
Features
Benefits
Faster processing and parsing of voice input
MMXTM technology improves the performance of
computationally intensive and repetitive speech functions such as:
sub-band filters
auto-correlation computations for pitch determinations & LPC filters
The multiply accumulate instruction and the additional 8 registers provide the
most significant performance benefits.
