Integer-Reversible MLT Audio Compression for Fast Lossless Transcoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional audio compression techniques face challenges such as inefficiency in predicting periodic tones, high computational overhead due to side information requirements, and reduced compression performance for larger transform lengths, especially when transcoding between lossless and lossy formats.
Innovation Solution
The STAC Codec employs an integer modulated lapped transform (MLT) for transforming audio signals into the frequency domain, combined with a backward-adaptive run-length Golomb-Rice (RLGR) encoder for compressing transform coefficients, allowing for efficient lossless and near-lossless compression with reduced computational overhead and enabling fast transcoding between formats.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If forward-adaptive prediction is used for audio compression, then compression can be achieved, but periodic tones cannot be efficiently predicted and high-order predictors require excessive computation
Solution Approach 1:
The patent replaces time-domain predictive coding with transform-domain coding. Instead of using complex predictors to model temporal correlations, the invention applies a transform (such as MDCT) to convert the audio signal into the frequency domain, where energy compaction occurs and simpler coding techniques can achieve comparable or better compression without the computational burden of high-order predictors
Solution Approach 2:
The patent changes the domain of representation from time-domain samples to transform coefficients. By transforming the signal and operating in the frequency domain, the invention achieves efficient compression of periodic tones through energy compaction in a few coefficients, avoiding the need for complex time-domain predictors
2Loss of substance
If lossless transform coding is used with large block sizes, then frequency-domain coding efficiency improves, but integer-reversible transforms accumulate rounding errors
Solution Approach 1:
The patent uses floating-point arithmetic for the transform coefficients instead of integer arithmetic. This allows for more precise representation of transform coefficients without the rounding errors that accumulate in integer-based reversible transforms, thereby maintaining high reconstruction accuracy while benefiting from frequency-domain coding efficiency
Solution Approach 2:
The patent introduces a quantization step in the transform domain as an intermediary process. By quantizing the transform coefficients controllably and using appropriate bit-depth representations, the invention manages precision requirements and avoids the rounding error accumulation problem while maintaining compression efficiency
3Reliability
If conventional lossless coding is used, then exact reconstruction is achieved, but transcoding to lossy format requires full decoding and re-encoding
Solution Approach 1:
The patent performs the transform operation during the initial encoding process and retains the transform coefficients in the compressed representation. This preliminary transformation allows for fast transcoding to lossy formats by operating directly on the existing transform coefficients without requiring full decoding to the time domain, thereby saving significant processing time while maintaining exact reconstruction capability when needed
4Productivity
If frequency-domain coding is used, then transcoding efficiency improves, but transform computations are required for lossless coding
Solution Approach 1:
The patent performs the computationally intensive transform operation during the initial encoding phase and stores the transform coefficients. This preliminary action eliminates the need to perform transform computations during transcoding operations, as the coefficients are already available in the frequency domain, thereby reducing computational overhead for subsequent operations while maintaining high transcoding speed
Data Source
AI summary
A “STAC Codec” provides lossless audio compression and decompression by processing an audio signal using integer-reversible modulated lapped transforms (MLT) to produce transform coefficients. Transform coefficients are then encoded using a backward-adaptive run-length Golomb-Rice (RLGR) encoder to produce losslessly compressed audio signals. In additional embodiments, further compression gains are achieved via an inter-block spectral estimation and data sorting strategy. Further, compression in the transform domain allows the bitstream to be partially decoded, using the corresponding RLGR decoder, to reconstruct the frequency-domain coefficients. These frequency-domain coefficients are then directly used to speed up various transform-domain based applications such as transcoding media to lossy or other formats, search, identification, visualization, watermarking, etc. In other embodiments, near-lossless compression is achieved by right-shifting transform coefficients by some number of bits such that quantization errors are not perceived as distortion in the decoded audio signal.