AUDIO BANDWIDTH EXTENSION
A revolutionary technology for clear wideband audio!
Audio coding at low bit rates has many established and emerging applications
such as Satellite and Terrestrial Digital Audio Broadcasting, audio delivery
over the mobile network, high quality audio communication over the IP and mobile
network, Internet music download and streaming, solid-state audio playback
devices, etc. In many of these applications the demand for higher compression
efficiency continues to grow. In fact there appears to be a proliferation of
applications demanding CD like quality stereo at bit rates of 32-48 kbps and
high quality FM grade mono audio at bit rates of 16-24 kbps. These in turn
continue to spur the demand for newer algorithms for audio bit rate reduction.
Audio Bandwidth Extension has emerged as a key technique for achieving higher
compression efficiency and hence high subjective quality at low bit rates.
ATC Labs has developed a number of elegant solutions providing critical
advantages.
Licensing/Support :
- For licensing details please Contact us at sales@atc-labs.com
- For details on implementations on other platforms please Contact us at info@atc-labs.com
Unmatched Audio Quality
ATC Labs patent pending technologies ASR, FSSM and MBTAC are embodied in its codec technology product, Audio Bandwidth Extension Toolkit (ABET). These recreate clear wide band audio based on sophisticated signal analysis and synthesis at low bit rates.
Absolute Flexibility
Rapid integration and deployment with any existing baseband coding scheme. Highly configurable algorithms which can be adapted to suit any bit rate.
Low Complexity
Reduced complexity algorithms for use in low power devices such as mobile phones.
Blind Bandwidth Extension Mode
BET now also incorporates a mode where by audio bandwidth can extended without any forward information as a post-processing option (see Audio Post Processing).
Technologies
Accurate Spectral Replacement
Accounts for the specificity of the coherent (i.e., sinusoidal) components of an audio signal, as well as the specificity of the incoherent (i.e., noise) components of an audio signal, namely with respect to their different perceptual impact and their different spectral nature and fine spectral structure.
Fractal Self Similarity Model
The self-similarity model provides a broad description of the intra-spectral correlation structures that may exist in the signal. The model, which includes replication as a special case, allows for a better approximation to the original spectrum for a wider class of signals.
Multi Band Temporal Amplitude Coding (MBTAC)
Tools for the efficient coding of envelope which look for typical and/or perceptually significant patterns in the time-frequency envelopes. These include techniques for noiseless coding and grouping based on perceptual criterion.
Highlights
- FSSM and ASR algorithms utilized in an adaptive framework and to allow for the combination of aspects of FSSM and ASR synthesis. In other words, ASR and FSSM may either be used independently or in combination to exploit their complementary nature.
- ASR and FSSM are applied in the domain of a high frequency resolution filterbank such as the Odd Discrete Frequency Transform (ODFT) or the Modified Discrete Cosine Transform (MDCT).
- The presence of efficient and high quality coding tools for the stereo envelope allows for a rapid integration of a parametric audio coding scheme offering accurate reproduction of stereo envelopes with any existing baseband coding scheme.
- Accurate reconstruction of the synthesized high frequency spectrum.
Publications :
- Anibal Ferreira and Deepen Sinha, “Accurate Spectral Replacement”, In the Preprint of 118th AES Convention, Barcelona, Spain, May 2005... PDF
- Deepen Sinha, Anibal Ferreira and Deep Sen, “A Fractal Self-Similarity Model for the Spectral Representation of Audio Signals”, In the Preprint of 118th AES Convention, Barcelona, Spain, May 2005... PDF
- Deepen Sinha, Anibal Ferreira and Harinarayanan E. V., “A Novel Integrated Audio Bandwidth Extension Toolkit (ABET)”, in the preprints of 120th Convention of the Audio Engineering Society, Paris, France, May 2006... PDF