With the knowledge of algorithms and many years of experience in firmware development, SEDA Solutions® has unique insight into many multimedia digital algorithm applications and their transformations onto floating and fixed-point processor platforms. Among many audio coding implementations, such as WMA and AAC codecs, SEDA Solutions® has achieved considerable success with its MPEG-1 and MPEG-2 decoder (commonly known as MP3) fixed-point implementation. Our uniquely designed MP3 Encoder algorithm is twice as efficient in complexity as to that of the standard yet it only requires half the complexity in delivering the same level of encoded quality.

Years of firmware services and multiple porting opportunities make SEDA Solutions® one of the leaders in providing high quality audio functionality on your system. Our rigorously tested implementations have been shipped out on several audio consumer products. No matter what the final target platform, SEDA Solutions'® audio firmware services deliver not only high quality coding but also efficient implementation resulting in optimum usage of memory and clock cycles. Savings in resource utilization ultimately reduce cost and power consumption.

Our fixed-point implementations have met all the standard test vectors. Depending on the choice of precision and resource availability, the decoder can be implemented to be either in partial or full compliance with respect to the standards requirements. As well as precision quality and optimum resource utilizations, SEDA Solutions® offers post-filtering options to enhance the quality of perceived signal. Other service options provided are technical system assistance and assurance in safeguarding against lost bit stream data and protection against MIPS overflow.

MP3 Encoder
SEDA Solutions® has re-designed and implemented a version of the Mpeg-1 and Mpeg-2 (Lower Sampling Frequency Extension) Layer III encoding algorithm desirable for floating or fixed-point processors with a data width of multiples of 16 bits. The encoding algorithm is guaranteed to provide the LOWEST MIPS for a given processor architecture core. The proprietary modified encoder algorithm is so simple that although it closely maintains the quality of a full encoder at bit rates of 64 kb/s and above, it requires only 50% to 60% the complexity of a normal encoder algorithm. The implementations have focused on maintaining a high quality and lowering the computational complexity while keeping the data precision to a minimum of required fixed-point bits. Table below shows a typical implementation by SEDA Solutions® of the MP3 algorithm described above on a conventional proprietary 16 bit fixed-point processor.

Module Name
MP3 Decoder
MP3 Encoder
Data Size
16/32
16/32
Instruction Width
32
32
Data RAM
12218
12888
Data ROM
12004
12004
Program Size
< 8 K
< 8K
Cycle MIPS
43
49


SEDA Solutions® MP3 Encoding implementation services will support all or a selection of the following options:

 

  • MPEG-1 and MPEG2 (LSFE) Layer III only
  • Sampling Rates: 16kHz, 22.05 kHz, 24 kHz, 32kHz, 44.1kHz, 48kHz
  • Bit Rates: 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144, 160, 176, 192, 224, 256, 288, 320, 352, 384, 416, 448 kbit/s
  • Modes: Mono, Dual Channel, Stereo, Joint Stereo
  • Constant bit rates

TOP OF PAGE

MP3 Decoder
SEDA Solutions® has implemented the Mpeg-1 Layers I, II and III plus Mpeg-2 Lower Sampling Frequency (LSF) extension as well as the Multi Channel extension audio decoding standards onto several 16 bit, 24 bit and 32 bit fixed data path processors. The implementations have focused on maintaining high quality and lowering computational complexity while keeping the data precision to a minimum of required fixed-point bits.

SEDA Solutions® MP3 Decoding implementation services will support all or a selection of the following options:

  • MPEG-1
  • MPEG2 (LSFE) or (MCE)
  • Layers: I, II and III
  • MPEG-1 Sampling Rates: 32kHz, 44.1kHz, 48kHz
  • MPEG-1 Bit Rates: 32, 40, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256, 288, 320, 352, 384, 416, 448 kbit/s
  • MPEG-2 Sampling Rates: 16kHz, 22.05 kHz, 24 kHz
  • MPEG-2 Bit Rates: 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144, 160, 176, 192, 224, 256 kbit/s
  • Modes: Mono, Dual Channel, Stereo, Joint Stereo
  • Variable and constant bit rates
  • CRC check
  • De-emphasis filtering
  • Fast forward and reverse speed play

Our fixed-point implementations have met all the ISO standard test vectors. Depending on the choice of precision and resource availability, the decoder can be implemented to be either in partial or full compliance with respect to the standards requirements.

TOP OF PAGE

Windows Media Audio
As an active member in the development and design of Audio application based devices, SEDA Solutions® is a licensee of Microsoft Corporation's Windows Media Audio (WMA) technology.

SEDA Solutions® has implemented the WMA Version 8.0 decoding technology on several programmable platforms. The implementations have focused on maintaining high quality and lowering computational complexity while keeping the data precision to a minimum of required fixed-point bits.

SEDA Solutions® WMA coding implementation services will support all or a selection of the following options:

  • Constant Bit Rate
  • Variable Bit Rate
  • Mono or Stereo
  • Bit Rates from 128 bps to 192 kbps
  • Sampling rates from 8 kHz to 48 kHz
  • Versions 7.0, 8.0 and 9.0

Our fixed-point implementations have met all the WMA standard test vectors.

TOP OF PAGE

Advanced Audio Coding
The AAC algorithm is another high quality compression algorithm that is part of the MPEG-4 Part 4 Audio Standard supported by ISO/IEC. Unlike the MP3 and WMA codecs, the AAC codec's superior quality has made it a standard in many mobile applications such as the 3GPP. The MPEG-4 AAC standard is based on its MPEG-2 AAC predecessor and is segmented into several tools that allow lower bit rate and added scalability with error resilience. The AAC Codec is divided into four different profiles:

  • AAC Low Complexity (LC)
  • AAC Long-Term Prediction (LTP)
  • AAC Scalable Sampling Rate (SSR)
  • Error Resilient AAC Low Delay (ER AAC LD)

SEDA Solutions® AAC coding implementation services will support all or a selection of the following options:

  • All profiles
  • Sampling Rates: 8 kHz to 96 kHz
  • Bit Rates: 16 to 576 kbps
  • Channels: 1 to 48 (including Stereo, Joint Stereo, Mono or Multi)
  • Constant Bit Rates


For more information on Audio related products and services please contact us by email at info@sedasolutions.com.

TOP OF PAGE

 

Copyright 2005 SEDA SOLUTIONS® All rights reserved.
info@sedasolutions.com