Audio Coding Functions

Intel^® IPP for audio coding includes general purpose functions applicable in several codecs and a number of specific functions for MPEG-4 audio encoder and decoder (see [ISO14496]), MP3 encoder and decoder. These functions implement pipeline blocks with large computational complexity.

The current set of functions is sufficient to implement a portable optimized MPEG-4 AAC Main profile decoder and a portable optimized MPEG-1, 2 Layer III encoder (see [ISO/IEC 11172-3] and [ISO13818]).

Also this chapter includes Spectral Band Replication (SBR) functions as an audio coding bandwidth extension technology, which is standardized in ISO/IEC 14496-3:2001/Amd.1:2003 [ISO14496A].

The full list of functions is given in the table below.

Intel IPP Audio Coding Functions
Function Base Name	Operation
Interleaved to Multi-Row Format Conversion Functions
Interleave	Converts signal from non-interleaved to interleaved format.
Deinterleave	Converts signal from interleaved to non-interleaved format.
Spectral Data Prequantization Functions
Pow34	Raises a vector to the power of 3/4.
Pow43	Raises a vector to the power of 4/3.
Pow43Scale	Raises a vector to the power of 4/3 with scaling.
Scale Factors Calculation Functions
CalcSF	Restores actual scale factors from the bit stream values.
Mantissa Conversion and Scaling Functions
Scale	Applies scale factors to spectral bands in accordance with spectral bands boundaries.
MakeFloat	Converts mantissa and exponent arrays to float arrays.
Modified Discrete Cosine Transform Functions
MDCTFwdInitAlloc, MDCTInvInitAlloc	Create and initialize modified discrete cosine transform specification structure.
MDCTFwdInit, MDCTInvInit	Initialize modified discrete cosine transform specification structure.
MDCTFwdFree, MDCTInvFree	Close modified discrete cosine transform specification structure.
MDCTFwdGetSize, MDCTInvGetSize	Get the size of MDCT specification structure, initialization, and work buffer.
MDCTFwdGetBufSize, MDCTInvGetBufSize	Get the size of MDCT work buffer.
MDCTFwd, MDCTInv	Compute forward or inverse MDCT of a signal.
Block Filtering Functions
FIRBlockInitAlloc	Initializes FIR block filter state.
FIRBlockFree	Closes FIR block filter state.
FIRBlockOne	Filters vector of sample through FIR block filter.
Frequency Domain Prediction Functions
FDPInitAlloc	Creates and initializes predictor state.
FDPInit	Initializes FDP state.
FDPFree	Closes FDP state.
FDPGetSize	Gets size of FDP state structure.
FDPReset	Resets predictors for all spectral lines.
FDPResetSfb	Resets predictor-specific information in some scale factor bands.
FDPResetGroup	Resets predictors for group of spectral lines.
FDPFwd	Performs frequency domain prediction procedure and calculates prediction error.
FDPInv	Retrieves input signal from prediction error, using frequency domain prediction procedure.
VLC Functions
VLCDecodeEscBlock_MP3	Parses the bitstream and decodes variable length code for MP3 using signed VLC tables.
VLCDecodeEscBlock_AAC	Parses the bitstream and decodes variable length code for AAC using signed VLC tables.
VLCDecodeUTupleEscBlock_MP3	Parses the bitstream and decodes variable length code for MP3 using unsigned multi-tupled VLC tables.
VLCDecodeUTupleEscBlock_AAC	Parses the bitstream and decodes variable length code for AAC using unsigned multi-tupled VLC tables.
VLCCountEscBits_MP3	Calculates the number of bits necessary for encoding in MP3 format.
VLCCountEscBits_AAC	Calculates the number of bits necessary for encoding in AAC format.
VLCEncodeEscBlock_MP3	Encodes an array of values into destination bitstream in MP3 format and advances bitstream pointer.
VLCEncodeEscBlock_AAC	Encodes an array of values into destination bitstream in AAC format and advances bitstream pointer.
Psychoacoustic Functions
Spread	Computes spreading function.
Vector Quantization Functions
VQCodeBookInitAlloc	Allocates memory and initializes the codebook state structure.
VQCodeBookInit	Initializes the codebook state structure.
VQCodeBookFree	Closes the codebook state structure.
VQCodeBookGetSize	Gets size of the codebook state structure.
VQPreliminarySelect	Selects candidates for the nearest code vector of codebooks.
VQMainSelect	Finds optimal indexes with minimal distortion.
VQIndexSelect	Finds optimal vector set for specified number of codebooks.
VQReconstruction	Reconstructs vectors from indexes.
MP3 Encoder Functions
AnalysisPQMF_MP3	Implements stage 1 of MP3 hybrid analysis filterbank.
AnalysisFilterInit_PQMF_MP3	Initializes PQMF MP3 analysis specification structure.
AnalysisFilterInitAlloc_PQMF_MP3	Allocates memory and initializes the PQMF MP3 analysis specification structure.
AnalysisFilterGetSize_PQMF_MP3	Returns size of PQMF MP3 analysis specification structure.
AnalysisFilterFree_PQMF_MP3	Frees memory allocated for PQMF MP3 analysis specification structure.
AnalysisFilter_PQMF_MP3	Transforms PQMF MP3-processed subband signals into time domain samples.
MDCTFwd_MP3	Implements stage 2 of the MP3 hybrid analysis filterbank.
PsychoacousticModelTwo_MP3	Implements the ISO/IEC 11172-3 psych-acoustic model recommendation 2 to estimate the masked threshold and perceptual entropy associated with a block of PCM audio input.
JointStereoEncode_MP3	Transforms independent left and right channel spectral coefficient vectors into combined mid/side (MS) and/or intensity (IS) mode coefficient vectors suitable for quantization.
Quantize_MP3	Quantizes the spectral coefficients generated by the analysis filterbank.
PackScaleFactors_MP3	Applies noiseless coding to the scale factors and then packs the output into the bitstream buffer.
HuffmanEncode_MP3	Applies lossless Huffman encoding to the quantized samples and packs the output into the bitstream buffer.
PackFrameHeader_MP3	Packs the content of the frame header into the bitstream.
PackSideInfo_MP3	Packs the side information into the bitstream buffer.
BitReservoirInit_MP3	Initializes all elements of the bit reservoir state structure.
MP3 Decoder Functions
UnpackFrameHeader_MP3	Unpacks the audio frame header.
UnpackSideInfo_MP3	Unpacks the side information from the input bitstream for use during the decoding of the associated frame.
UnpackScaleFactors_MP3	Unpacks scale factors.
HuffmanDecode_MP3, HuffmanDecodeSfb_MP3, HuffmanDecodeSfbMbp_MP3	Decodes Huffman data.
ReQuantize_MP3, ReQuantizeSfb_MP3	Requantizes the decoded Huffman symbols.
MDCTInv_MP3	Performs the first stage of hybrid synthesis filter bank.
MDCTInvWindow_MP3	Performs 36 or 18 points IMDCT and windowing depending on the block type.
SynthPQMF_MP3	Performs the second stage of hybrid synthesis filter bank.
SynthesisFilterInit_PQMF_MP3	Initializes PQMF MP3 synthesis specification structure.
SynthesisFilterInitAlloc_PQMF_MP3	Allocates memory for PQMF MP3 synthesis specification structure and initializes it.
SynthesisFilterGetSize_PQMF_MP3	Returns size of PQMF MP3 synthesis specification structure.
SynthesisFilterFree_PQMF_MP3	Frees memory allocated for PQMF MP3 synthesis specification structure.
SynthesisFilter_PQMF_MP3	Transforms PQMF MP3-processed subband signals into time domain samples.
AAC Decoder Functions
UnpackADIFHeader_AAC	Gets the AAC ADIF format header.
UnpackADTSFrameHeader_AAC	Gets ADTS frame header from the input bitstream.
DecodePrgCfgElt_AAC	Gets program configuration element from the input bitstream.
DecodeChanPairElt_AAC	Gets channel_pair_element from the input bitstream.
NoiselessDecoder_LC_AAC	Decodes all the data for one channel.
DecodeDatStrElt_AAC	Gets data stream element from the input bitstream.
DecodeFillElt_AAC	Gets the fill element from the input bitstream.
QuantInv_AAC	Performs inverse quantization of Huffman symbols for current channel in-place.
DecodeMsStereo_AAC	Processes MS stereo for pair channels in-place.
DecodeIsStereo_AAC	Processes intensity stereo for pair channels.
DeinterleaveSpectrum_AAC	Deinterleaves the coefficients for short block.
DecodeTNS_AAC	Decodes for Temporal Noise Shaping in-place.
MDCTInv_AAC_32s16s	Maps the time-frequency domain signal into time domain and generates 1024 reconstructed 16-bit signed little-endian PCM samples.
MDCTInv_AAC_32s_I	Computes inverse modified discrete cosine transform (MDCT), windowing and overlapping of signals.
DecodeMainHeader_AAC	Gets main header information and main layer information from bit stream.
DecodeExtensionHeader_AAC	Get extension header information and extension layer information from bit stream.
DecodePNS_AAC	Implements perceptual noise substitution coding within an ICS.
DecodeMsPNS_AAC	Implements perceptual noise substitution coding in the case of joint coding.
DecodeChanPairElt_MP4_AAC	Gets channel_pair_element from the input bitstream
LongTermReconstruct_AAC	Uses Long Term Reconstruct to reduce the redundancy of a signal between successive coding frames.
MDCTFwd_AAC_32s	Generates spectrum coefficient of PCM samples.
MDCTFwd_AAC_32s_I	Computes forward modified discrete cosine transform (MDCT) of windowed signals.
EncodeTNS_AAC	Performs reversion of TNS in the Long Term Reconstruct loop in-place.
LongTermPredict_AAC	Gets the predicted time domain signals in the Long Term Reconstruct (LTP) loop.
NoiselessDecode_AAC	Performs noiseless decoding.
LtpUpdate_AAC	Performs required buffer update in the Long Term Reconstruct (LTP) loop.
SBR Audio Encoder Functions
DetectTransient_SBR	Estimates stationarity of signal by calculating deviation of spectrum audio signal.
EstimateTNR_SBR	Measures Tonality-to-Noise Ratio of complex QMF subband samples.
AnalysisFilterEncGetSize_SBR	Returns sizes of analysis SBR specification structures and work buffers.
AnalysisFilterEncInit_SBR	Initializes analysis SBR specification structure.
AnalysisFilterEnc_SBR	Performs subband filtering of the input audio signal.
AnalysisFilterEncInitAlloc_SBR	Allocates memory and and initializes analysis SBR specification structure.
AnalysisFilterFree_SBR	Closes analysis SBR specification structure.
SBR Audio Decoder Functions
AnalysisFilterInitAlloc_SBR	Allocates memory and initializes analysis SBR specification structure for real and complex signals.
SynthesisFilterInitAlloc_SBR	Allocates memory and initializes synthesis SBR specification structure for real and complex signals.
SynthesisDownFilterInitAlloc_SBR	Allocates memory and initializes downsample synthesis SBR specification structure for real and complex signals.
AnalysisFilterFree_SBR	Closes analysis SBR specification structure for real and complex signals.
SynthesisFilterFree_SBR	Closes synthesis SBR specification structure for real and complex signals.
SynthesisDownFilterFree_SBR	Closes downsample synthesis SBR specification structure for real and complex signals.
AnalysisFilterGetSize_SBR	Returns size of analysis FilterSpec_SBR specification structures, init and work buffers.
SynthesisFilterGetSize_SBR	Returns size of synthesis FilterSpec_SBR specification structures, init and work buffers.
SynthesisDownFilterGetSize_SBR	Returns size of downsample synthesis FilterSpec_SBR specification structures, init and work buffers.
AnalysisFilterInit_SBR	Initializes analysis SBR specification structure.
SynthesisFilterInit_SBR	Initializes synthesis SBR specification structure.
SynthesisDownFilterInit_SBR	Initializes downsample synthesis specification structure.
AnalysisFilter_SBR	Transforms time domain signal output from the core decoder into frequency subband signals.
SynthesisFilter_SBR	Transforms SBR-processed subband signals into time domain samples.
SynthesisDownFilter_SBR	Transforms SBR-processed subband signals into time domain samples and performs downsampling at the same time.
PredictCoef_SBR	Obtains prediction filter coefficients using covariance method.
PredictOneCoef_SBR	Obtains one prediction filter coefficient using covariance method
Parametric Stereo Functions
AnalysisFilter_PS	Increases frequency resolution of the first lower subbands by hybrid filtering.
DTS Audio Coding Functions
SynthesisFilterInit_DTS	Initializes DTS synthesis filter specification structure.
SynthesisFilterInitAlloc_DTS	Allocates memory for DTS synthesis filter specification structure and initilaizes it.
SynthesisFilterGetSize_DTS	Returns size of DTS synthesis filter specification structure.
SynthesisFilterFree_DTS	Frees memory allocated for DTS synthesis filter specification structure.
SynthesisFilter_DTS	Transforms QMF DTS-processed subband signals into time domain samples.