Fortemedia’s SAM VE™
(Small Array Microphone, Voice Extraction) Technology

Fortemedia is at the forefront of enabling hands free communication with SAM technology and delivers industry-leading echo cancellation and noise suppression solutions to enable natural flowing full-duplex communication in mobile communication devices such as portable telematics, handhelds such as PDAs and smartphones, and Bluetooth accessories.

Combined with DSP-based voice processor ICs, SAM provides soaring advancements in voice quality and usability for the voice interface market.

New SAM VE™ technology integrates physical sound propagation model, voice production model and hearing science model into our advanced adaptive filtering techniques.

● Sound propagation model:
utilize physics of sound waves (causality, diffraction, etc) to keep only voice coming from talker’s mouth but suppress any interference outside the mouth area.

● Speech production model:
utilize the unique features of voice (pitch, harmonics, onset and offset features) to extract desired voice.

● Hearing science model:
utilize psychoacoustics knowledge (masking model, cochlear based filter bank, dynamic range compressor, etc) to enhance the characteristics more sensitive to human hearing.

● Up to 37dB Non-Stationary Noise Suppression
　-100ms Convergence Time

● Up to 0.80 MOS Score Improvement (at 6dB SNR)
　-UP to 25dB Stationary Noise Suppression

● Up to 60dB Echo Cancellation

● Implementation using two omni microphones
　-Superior Ovoid-Shaped Beamforming

● Non-Stationary Noise Suppression
　-Close Talk Voice Extraction aided by inclusion zone


FM2018-380 Overview
● 16-bit voice process with hardware accelerator
● Three 16-bit ADCs 8KHz sampling rate and one 16-bit DAC, SNR 75dB
● Microphone input has built-in PGA as on-chip pre-amp
● Built-in RAM and ROM code, No external RAM necessary
● 1.8V power source with 3.3V tolerance I/O
● Low power –27mW in two microphone mode
● Power down mode: less than 20 μA
● Support O+O BF for noise suppression, one-mic or two mics microphone
　configuration for AEC Flexible clock source
　--internal clock generation with crystal:13MHz and 12.288 MHz
　--external clock source: 13MHz, 19.2 MHz, 24MHz, 48MHz and 12.288 MHz
● Small foot print
　--22-pin CSP package (3.93 x 2.04 mm2)
　--48-pin QFN package (7 x 7 mm2)

▲TOP