hash stringlengths 32 32 | doc_id stringlengths 7 13 | section stringlengths 3 121 | content stringlengths 0 2.2M |
|---|---|---|---|
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.1 b_smpoffst - payload sample offset flag - 1 bit | This bit, if set, indicates that the smpoffst field shall follow in the bitstream. If the current payload applies to the first sample of the AC-4 frame, this field shall be set to '0'. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.2 smpoffst - payload sample offset - variable_bits(11) | The smpoffst field indicates the offset, in units of PCM audio samples, from the beginning of the AC-4 frame to the first PCM audio sample that the data in the payload shall apply to. Note that this field may indicate a sample index that extends beyond the current AC-4 frame. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.3 b_duration - payload duration flag - 1 bit | This bit, if set, indicates that the duration field shall follow in the bitstream. If the current payload applies to all samples up to and including the final sample of the AC-4 frame, this field shall be set to '0'. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.4 duration - payload duration - variable_bits(11) | The duration field indicates the time period in units of PCM audio samples that the data in the payload applies to. Note that this field may indicate a sample index that extends beyond the current AC-4 frame. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.5 b_groupid - payload group ID flag - 1 bit | This bit, if set, indicates that the groupid field shall follow in the bitstream. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.6 groupid - payload group ID - variable_bits(2) | The groupid field provides a mechanism to indicate to a decoder that specific payloads within the UMD substream are associated with one another and that the payload data are related. For payloads that are associated with each other, the groupid field for each payload shall be set to the same value. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.7 b_codecdata - codec specific data flag - 1 bit | The 1-bit b_codecdata field enables codec-specific data to be included in the payload configuration. For payload configuration data that conforms to the present document, this field shall be set to '0'. ETSI ETSI TS 103 190 V1.1.1 (2014-04) 122 |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.8 codecdata - codec specific data - 8 bits | This is not specified here since bitstreams conforming to the present document shall have b_codecdata=0. 4.3.15.2.9 b_discard_unknown_payload - discard unknown payload during transcode flag - 1 bit This bit indicates whether the current payload is retained or discarded during a transcoding process if the transcoder doe... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.10 b_payload_frame_aligned - payload to audio data frame alignment flag - 1 bit | The 1-bit b_payload_frame_aligned field is used to indicate how closely the payload data and the audio data within the AC-4 frame need to be synchronized after decoding and/or during transcoding. If the b_payload_frame_aligned field is set to '0', the application of the payload data to the decoded audio from the AC-4 f... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.11 b_create_duplicate - create duplicate payload during transcode flag - 1 bit | In some transcode operations, the frame length of the output format may be shorter than the length of an AC-4 frame. The 1-bit b_create_duplicate field shall be set to '1' to indicate that the payload shall be duplicated in all frames of the output format that correspond to the same period of time as the current AC-4 f... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.12 b_remove_duplicate - remove duplicate payload during transcode flag - 1 bit | The 1-bit b_remove_duplicate field shall be set to '1' to indicate that payloads with the same umd_payload_id value as the current payload need to be included only once in every output frame of the transcoding process when the output frame duration of a transcoding process is longer than the length of an AC-4 frame. Su... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.13 priority - payload priority - 5 bits | The 5-bit priority field is used to indicate the priority of this payload in relation to other payloads in the UMD substream. This field may be used to indicate whether a payload can be discarded during a transcoding process to a lower data rate output format that is not able to support all payloads. A priority value o... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.2.14 proc_allowed - processing allowed - 2 bits | The 2-bit proc_allowed field is used to indicate whether a payload should be retained or discarded during transcoding if processing is applied to the metadata and/or audio data decoded from the AC-4 frame during transcoding. The value of the proc_allowed field shall be set as shown in table 169. ETSI ETSI TS 103 190 V1... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.3 umd_protection - Universal metadata protection data | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.3.1 protection_length_primary - length of protection_bits_primary field - 2 bits | The 2-bit protection_length_primary field is used to specify the length of the protection_bits_primary field. The value of the protection_length_primary field shall be set as shown in table 170. Table 170: Meaning of protection_length_primary values protection_length_primary Length of protection_bits_primary field 00 R... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.3.2 protection_length_secondary - length of protection_bits_secondary field - 2 bits | The 2-bit protection_length_secondary field is used to specify the length of the protection_bits_secondary field. The value of the protection_length_secondary field shall be set as shown in table 171. Table 171: Meaning of protection_length_secondary values protection_length_secondary Length of protection_bits_secondar... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 4.3.15.3.3 protection_bits_primary - primary UMD substream protection data - 8 to 128 bits | The protection_bits_primary field contains protection data that may be used to verify the integrity of the UMD substream and the payload data within the UMD substream. Calculation of the value of the protection_bits_primary field is implementation dependent and is not defined in the present document. 4.3.15.3.4 protect... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5 Algorithmic details | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1 Audio spectral frontend (ASF) | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.1 Introduction | The audio spectral frontend tool is one of two spectral frontend tools which are used to create spectral lines of audio tracks from sf_data() elements in an AC-4 frame. These spectral lines are typically routed through the stereo and multichannel processing tool before being routed into the IMDCT. The ASF tool consists... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.2 Entropy coding of spectral values | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.2.1 Introduction | Data and Control Interfaces The input to the Huffman decoding tool is the asf_spectral_data() part of a sf_data() element: spectral_data Huffman coded spectral data (collection of Huffman codewords and sign bits). Note that the length of the spectral data is only known after the Huffman decoding process. The output fro... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.2.2 Decoding process | The quantized spectral lines are stored as Huffman encoded values in the bitstream. To retrieve the quantized spectral lines, Huffman decoding shall be applied. For each Huffman codeword in the bitstream, the used Huffman codebook number sect_cb[g][s] is known from the section data. All the valid Huffman codebooks for ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.3 Quantization reconstruction and scaling | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.3.1 Introduction | Data and Control Interfaces The input to the quantization reconstruction and scaling tool is a vector of quantized spectral lines and the asf_scalefac_data() part of a sf_data() element: quant_spec vector of quantized spectral lines NOTE 1: The maximum allowed value of the elements of quant_spec is 8191. scale_factor_d... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.3.2 Decoding process | To convert the quantized spectral lines into scaled spectral lines a quantization reconstruction and scaling step shall be applied. The following (non-uniform) reconstruction process shall be used to apply the inverse operation to the non-uniform quantization operation used by the encoder for the spectral lines. The re... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.4 Spectral noise fill | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.4.1 Introduction | Data and Control Interfaces The input to and the output from the spectral noise fill tool is a vector of spectral lines in bitstream order: scaled_spec vector of (scaled) spectral lines in bitstream order The bitstream parameter and helper elements used by the spectral noise fill tool are: num_window_groups number of w... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.4.2 Decoding process | Spectral lines in scaled_spec which are 0 because either the corresponding sfb_cb[g][sfb] is 0 or the corresponding max_quant_idx[g][sfb] is 0 shall be replaced by random noise of a transmitted level if noise fill data is present as indicated by b_snf_data_exists=1. If b_snf_data_exists=0, the spectral noise fill tool ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.5 Spectral ungrouping tool | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.5.1 Introduction | Data and Control Interfaces The input to the ungrouping tool is a vector of spectral lines in bitstream order: scaled_spec vector of (scaled) spectral lines in bitstream order The output from the ungrouping tool is a vector of reordered spectral lines: spec_reord vector of reordered (scaled) spectral lines, i.e. ordere... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.1.5.2 Decoding process | When more than one transform block is used in an AC-4 frame, spectral lines are stored in groups. To undo this grouping, the AC-4 decoder shall apply reordering as specified in the following pseudocode: Pseudocode k = 0; win = 0; spec_reord[] = 0; /* init all spectral lines with 0 */ for (g = 0; g < num_window_groups; ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2 Speech spectral frontend (SSF) | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.1 Introduction | The Speech Spectral Frontend (SSF) is an alternative spectral frontend to the Audio Spectral Frontend (ASF). It is designed to be used for speech content. Like the ASF, the SSF will provide blocks of spectral lines which need to be inverse transformed using the same inverse MDCT as for the ASF output. The SSF data whic... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.2 Top level structure of the SSF | Figure 4: SSF decoder block diagram Figure 4 shows the SSF decoder block diagram. The predictor decoder, the spectrum decoder and the envelope decoder have bitstream elements as input. The output of the spectrum decoder is combined (added) with the output of the subband predictor and sent to the inverse flattening whic... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.3 Envelope decoder | Data and Control Interfaces The input to the envelope decoder is a vector of envelope indices, a vector of gain indices and a vector decoded envelope values from the previous frame: env_idx[band] vector of num_bands envelope indices ETSI ETSI TS 103 190 V1.1.1 (2014-04) 132 NOTE 1: The first index describes an absolute... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.4 Predictor decoder | Data and Control Interfaces The inputs to the predictor decoder are SSF bitstream elements. The outputs from the predictor decoder are the following predictor parameters: f_pred_gain[block] vector of num_blocks predictor gain values f_pred_lag[block] vector of num_blocks predictor lag values The state internal to the p... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.5 Spectrum decoder | Data and Control Interfaces The inputs to the spectrum decoder are envelope allocation values and the predictor gain for the current block: env_alloc[band] vector of num_bands envelope allocation values (integer) for current block f_pred_gain predictor gain value for the current block block block index of the current b... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.6 Subband predictor | Data and Control Interfaces The input to the subband predictor is a vector of SSF output spectral lines from the previous block, a signal envelope vector for the current block and the predictor parameters from the predictor decoder for the current block: f_spec[bin] vector of num_bins spectral lines from the previous b... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.7 Inverse flattening | Data and Control Interfaces The input to the inverse flattening is the spectrum from the spectrum decoder, the spectrum from the subband predictor and the signal envelope values for the current block: f_spec_res[bin] vector of num_bins residual spectral lines from the spectrum decoder f_spec_pred[bin] vector of num_bin... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.8 Parameterization | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.8.1 C matrix | The extractor of the subband predictor uses a matrix C, selected by tab_idx. The setup of all 37 possible matrices, stored in a big matrix Call[tab_idx], is described in the following. Each matrix C is a three-dimensional array of prediction coefficients of the size (2Rf + 1)×65×Rt, where Rt and Rf are dependent on tab... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.8.2 Arithmetic coding | A single instance of an arithmetic coder is used during the arithmetic decoding of the following three parameter types within an SSF granule, each using a different statistical model: • Envelope indices using a static pre-trained CDF table • Predictor gain indices using a static pre-trained CDF table • Transform coeffi... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.2.8.3 Dither and random noise | Two instances of a random number generator are used during the SSF decoding - one for the dither signal setup and the other one for noise signal values. The implementation of the random number generator shall conform to the code in the following pseudocode since it is important that the SSF encoder and the SSF decoder ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3 Stereo and multichannel processing | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.1 Introduction | The multichannel tools take their input from one or more spectral frontend(s), receiving n-tuples of spectra with matching time/frequency layout. The multichannel processing tool applies a number time- and frequency varying matrix operation on these tuples. Between two and seven spectra are transformed at a time. Param... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.2 Parameter extraction | All parameters for the multichannel tool are transmitted in one or more chparam_info() elements. The following pseudocode describes parsing the chparam_info() element and extracting parameters a, b, c, d: Pseudocode // initialize a[g][sfb], b[g][sfb], c[g][sfb] and d[g][sfb] after parsing of // chparam_info() for (g = ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.3 Processing the channel data elements | In the various channel data elements, a number n > 1 of input tracks (I0, …, In) are matrix transformed into the n output tracks (O0, …, On). They share the same sf_info and so share the same time/frequency resolution. The matrix equations below are understood to apply to each time/frequency tile separately, with separ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.3.1 Processing tracks of the mono_data element | Let I0 be the track decoded from the sf_data element contained in the mono_data element. The decoder shall map the input track directly to the output O0, i.e.: <O= = <I= ETSI ETSI TS 103 190 V1.1.1 (2014-04) 157 |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.3.2 Processing tracks of the two_channel_data or stereo_data element | Let I0 and I1 be the tracks decoded from the two sf_data elements contained in the two_channel_data element, or of the stereo_data element. If b_enable_mdct_stereo_proc = 0, no further processing is necessary and O0,=I0, O1=I1. Otherwise, the decoder shall extract the parameters a, b, c, d from the contained chparam_in... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.3.3 Processing tracks of the three_channel_data element | Let I0, I1, and I2 be the tracks decoded from the three sf_data elements contained in the three_channel_data element. The decoder shall extract parameters ai, bi, ci, di (i = 0…1) from the two contained chparam_info elements, and from the chel_matsel element determine the transform matrix M as described in table 172. F... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.3.4 Processing tracks of the four_channel_data element | Let I0…I3 be the tracks decoded from the four sf_data elements contained in the four_channel_data element. The decoder shall extract parameters ai, bi, ci, di (i = 0…3) from the four contained chparam_info elements. Finally, the output tracks O0… O3 shall be derived by: E O O O O F = E aa ba a b b b ca da c b... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.3.5 Processing tracks of the five_channel_data element | Let I0…I4 be the tracks decoded from the five sf_data elements contained in the five_channel_data element. Next the decoder shall extract parameters ai, bi, ci, di (i = 0…4) from the five contained chparam_info elements, and from the chel_matsel element determine the transform matrix M as described in table 173. ETSI E... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.4 Processing the channel elements | Tracks routed into the stereo and multiprocessing tool (s,-.,[| ||… ]) may originate from four different channel elements: channel_pair_element, 3_0_channel_element, 5_X_channel_element or 7_X_channel_element. Processing and mapping from input tracks to output channels (s,-.,[/|0|1|… ]) differs depending on the chann... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.4.1 Processing tracks of a channel_pair_element | If stereo_codec_mode {SIMPLE,ASPX,ASPX_ACPL_1}, the 2 tracks of the stereo_data() element shall be processed according to clause 5.3.3.2. The mapping of input tracks and output channels shall be done as follows: >I I ? = @s,-., s,-., A @s,-.,/ s,-.,0A = >O O ? If stereo_codec_mode = ASPX_ACPL_2, the single track of ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.4.2 Processing tracks of a 3_0_channel_element | If 3_0_coding_config = 0, the 2 tracks of the stereo_data() element (sSMP,0, sSMP,1) are processed according to clause 5.3.3.2. The mapping of input tracks and output channels shall be done as follows: >I I ? = @s,-., s,-., A @s,-.,/ s,-.,0A = >O O ? The third track of the additional mono_data() element, sSMP,2, sha... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.4.3 Processing tracks of a 5_X_channel_element | The up to 5 tracks acquired from a 5_X_channel_element shall be processed according to the channel data element they originate from. Tables in this clause describe which tracks shall be taken as input, and which channels are produced as output when processing the various channel data elements according to clause 5.3.3.... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.4.3.3 5_X_codec_mode = ASPX_ACPL_3 | If 5_X_codec_mode ASPX_ACPL_3, the two audio tracks (sSMP,0, sSMP,1) are acquired from a stereo_data element and processed according to clause 5.3.3.2. The mapping of input tracks and output channels shall be done as follows: >I I ? = @s,-., s,-., A G H H H I s,-.,/ s,-.,0 s,-.,1 s,-.,/2 s,-.,02J K K K L = G H H H I1... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.3.4.4 Processing tracks of the 7_X_channel_element | The 7_X_channel_element enables the decoder for two more channels beyond L, R, C, Ls and Rs (refer to table 87 for details). Analog to the 5_X_channel_element, the up to 7 tracks of a 7_X_channel_element shall be processed according to the channel data element they originate from. Hence, table notations are borrowed fr... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.4 96 and 192 kHz decoding | If bitstream element b_hsf_ext is set, the ac4_hsf_ext_substream structure contains further spectral data, scalefactor data and spectral noise fill data extending the data beyond 24 kHz. If the decoder is not capable of decoding into 96 kHz or 192 kHz output, it shall ignore the additional coefficients. If the decoder ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.5 IMDCT transform equations and block switching | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.5.1 Introduction | The choice of analysis block length is fundamental to any transform-based audio coding system. A long transform length is most suitable for input signals whose spectrum remains stationary, or varies only slowly, with time. A long transform length also provides greater frequency resolution, and hence improved coding per... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.5.2 Transforms | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.5.2.1 Introduction | The Inverse Modified Discrete Cosine Transform (IMDCT) applies the inverse of the frequency mapping that was carried out in the encoder. See clause 5.5.3 for a full list of supported transform lengths. ETSI ETSI TS 103 190 V1.1.1 (2014-04) 165 Data and Control Interfaces The input to the IMDCT is a block of N scaled sp... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.5.2.2 Decoding process | In the following procedure, steps 1 through 4 describe the computation of the IMDCT, which will transform the N spectral lines into an intermediate single real data block of length 2N using a single N/2 point complex IFFT with simple pre- and post-phase-shift operations. Steps 5 and 6 describe the unfolding, windowing ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.5.3 Block Switching | The following full and partial block lengths can be employed in the AC-4 transform block-switching procedure if the sampling frequency is 44,1 kHz. • 2 048, 1 024, 512, 256, 128 ETSI ETSI TS 103 190 V1.1.1 (2014-04) 168 If the sampling frequency is 48 kHz, the following additional block lengths are possible: • 1 920, 9... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.6 Frame alignment | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.6.1 Introduction | 5. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 6.1 Introduction | AC-4 is designed to enable artefact free switching and splicing between different streams, as may happen in adaptive streaming [1]. When the input of the codec is switched to a different adaptation set, it is important to take into account the codec internal signal and metadata delays. Because some control data apply t... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.6.2 Decoding process | 5. |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 6.2 Decoding process | A decoder shall apply a delay between the input and output PCM samples according to: !" = #[ −$_%&'] where d_pcm is defined per table 182. Table 182: Delay in frame alignment tool frame_rate PCM signal delay d_pcm [samples] Control data delay d_ctrl [frames] 23,976 288 1 24 288 1 25 352 1 29,97 96 1 30 96 1 47,952... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7 QMF domain processing | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.1 Introduction | AC-4 decoders employ a complex QMF analysis/synthesis filter bank pair to enable alias suppressed frequency-domain processing. The output of the spectral frontend is transformed into a matrix of num_qmf_subbands complex QMF subbands as rows and num_qmf_timeslots time slots as columns, where num_qmf_timeslots is equal t... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.2 QMF control data alignment | The control data of tools operating in QMF domain is at all times aligned with the corresponding spectral frontend data, contained in the sf_data element. This means, that within a raw AC-4 frame, the control data of any QMF domain tool shall be applied to the spectral data encoded in the very same raw AC-4 frame, rega... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.3 QMF analysis filterbank | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.3.1 Introduction | A QMF bank is used to split the time domain signal from the core decoder into num_qmf_subbands subband signals. The output from the filterbank, i.e. the subband samples, is complex-valued and thus oversampled by a factor two compared to a regular cosine modulated QMF bank. Data and Control Interfaces The input of the Q... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.3.2 Decoding process | The number of QMF subbands in the analysis QMF filter bank, num_qmf_subbands, is always 64. The number of QMF slots is calculated as follows: !'_(')_"#'*+, "+ = _ !" # _$ _ #%% & Table 183 shows the possible values for num_qmf_timeslots. Table 183: Number of QMF time slots frame_length 2 048 1 920 1 536 1 024 ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.4 QMF synthesis filterbank | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.4.1 Introduction | Synthesis filtering of the QMF subband signals is achieved using a num_qmf_subbands-subband synthesis QMF bank. The output from the filterbank is real-valued time domain samples. Data and Control Interfaces The input to the QMF synthesis filterbank is a set of num_qmf_subbands complex-valued subband signals: QinQMF,ch ... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.4.2 Decoding process | The synthesis filtering comprises the following steps, where a vector qsyn_filt consisting of 1 280 samples is assumed: • Shift the samples in the vector qsyn_filt by 128 positions. The oldest 128 samples are discarded. • The num_qmf_timeslots new complex-valued subband samples are multiplied by the matrix N, where: 5(... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.5 Companding tool | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.5.1 Introduction | The companding tool is used to mitigate pre- and post-echo artefacts. It is applied on the QMF domain data. The encoder applies attenuation to signal parts with higher energy, and gain to signal parts with lower energy. The decoder applies the inverse process, which effectively shapes the coding noise by the signal ene... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.5.2 Decoding process | The QMF subband range [sb0, sb1] that the companding tool works on shall be determined according to: +- = 8acpl_qmf_band." in case the codec mode is ASPX_ACPL_1 0 otherwise , = __ Let K= sb1-sb0 be the number of affected subbands. Absolute sample levels ( , ) shall be calculated according to: ,... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6 Advanced spectral extension tool - A-SPX | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.1 Introduction | The A-SPX tool (Advanced Spectral Processing Extension) is used to parametrically code the higher frequencies of the audio signal. IWC (Interleaved Waveform Coding) may be employed in addition, if the parametric model cannot re-instate the correct frequency components, or to achieve a more accurate temporal reproductio... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.2 A-SPX specific variables | sbg_master array of master QMF subband groups, of length num_sbg_master+1 sbg_sig_highres array of QMF subband group borders for high frequency resolution signal envelopes, is of length num_sbg_sig_highres+1 sbg_sig_lowres array of QMF subband group borders for low frequency resolution signal envelopes, is of length nu... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.3 Decoding A-SPX control data | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.3.1 Subband groups | QMF subband samples are combined to form subband groups, short-handed "sbg". These are described by subband group tables, which are of six types: 1) Template subband group tables, sbg_template_[high|low]res: Static high and low resolution subband group tables serving as template for deriving the master table. 2) Master... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.3.2 Low band filter and QMF delay line | The QMF time-frequency samples from the analysis filterbank, stored in QinASPX, are truncated at the cross over subband, sbx, and delayed by ts_offset_hfgen before being fed into the HF Generator according to the following pseudocode. Pseudocode for (ts=0; ts<ts_offset_hfgen; ts++) { for (sb=0; sb<num_qmf_subbands; sb+... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.3.3 Time/frequency matrix | The A-SPX time/frequency matrix is derived from the QMF matrix Q_low. Similarly, it has time on the horizontal axis and frequency on the vertical axis. The A-SPX matrix is generated using two methods. • Framing, to determine the time slot borders of the signal and noise envelopes • Tiling, to determine the subband grou... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.3.4 Decoding A-SPX signal and noise envelopes | For each A-SPX signal envelope and noise envelope, the signal and noise scale factors respectively are delta-coded either along time or frequency directions. Upon crossing an A-SPX interval boundary (along the time axis), the first envelope in the current A-SPX interval may be delta-coded using Huffman decoding with re... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.3.5 Dequantization and stereo decoding | Two quantization steps are possible for the quantization of the signal scale factors: • aspx_quant_mode_env = 0 corresponds to a quantization step of 1,5 dB; and • aspx_quant_mode_env = 1 corresponds to a quantization step of 3,0 dB. For a 1-channel element and for a 2-channel element with aspx_balance = 0, i.e. when s... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.4 HF signal construction | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.4.1 HF generator tool | 5.7.6.4.1.1 Introduction The HF generator patches subband signals from consecutive subbands of the matrix Q_low to consecutive subbands of the matrix Q_high based on parameters transmitted in the bitstream. The tonal to noise ratio of the subband signals Q_high are adjusted to the levels signaled in the bitstream by th... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.4.2 HF envelope adjustment tool | The envelope adjustment process takes Q_high as input from the HF generator and produces as output a QMF matrix Y. The adjustment is performed upon the entire A-SPX range spanning the time slots of the current A-SPX interval (given by atsg_sig) and num_sb_aspx QMF subbands, starting at subband sbx. ETSI ETSI TS 103 190... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.4.3 Noise generator tool | Overview The noise generator tool is used to inject noise into the recreated high-frequency spectrum in order to decrease its tonality and to more closely match the original signal. The tool outputs a noise envelope that is subsequently added to the generated HF signal. Data and Control Interfaces Input • noise_lev_sb_... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.4.4 Tone generator tool | Overview The tone generator tool is used to match the tonality of the original signal by adding the appropriate missing sinusoids to the recreated high band. The tool outputs an envelope worth of values, the dimensions of which are specified in the bitstream by the encoder. Data and Control Interfaces Input • sine_lev_... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.4.5 HF signal assembling tool | The compensated gain values are applied to the input subband matrix Q_high, for all signal envelopes of the current A- SPX interval as follows. Note, in case of variable A-SPX interval borders, the previously assembled HF signals, Y_prev are pushed into the current output buffer Y. Pseudocode tsg = 0; /* Get delayed QM... |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.5 Interleaved waveform coding | |
d42eb522967d110c6c642c9d4bb78e76 | 103 190 | 5.7.6.5.1 Introduction | A-SPX involves patching the spectral content of suitable low frequency regions to higher frequencies and adjusting them such that they perceptually match the high frequency content that was present in the original signal. The entire signal spectrum above the A-SPX crossover frequency is derived from the spectrum below ... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.