EP1697929B1 - Procede et dispositif pour quantifier un signal de donnees - Google Patents

Procede et dispositif pour quantifier un signal de donnees Download PDF

Info

Publication number
EP1697929B1
EP1697929B1 EP05715289A EP05715289A EP1697929B1 EP 1697929 B1 EP1697929 B1 EP 1697929B1 EP 05715289 A EP05715289 A EP 05715289A EP 05715289 A EP05715289 A EP 05715289A EP 1697929 B1 EP1697929 B1 EP 1697929B1
Authority
EP
European Patent Office
Prior art keywords
audio
values
block
threshold
value
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
EP05715289A
Other languages
German (de)
English (en)
Other versions
EP1697929A1 (fr
Inventor
Gerald Schuller
Stefan Wabnik
Jens Hirschfeld
Wolfgang Fiesel
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Original Assignee
Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV filed Critical Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Publication of EP1697929A1 publication Critical patent/EP1697929A1/fr
Application granted granted Critical
Publication of EP1697929B1 publication Critical patent/EP1697929B1/fr
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
    • G10L19/02—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders
    • G10L19/032—Quantisation or dequantisation of spectral components
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
    • G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
    • G10L19/26—Pre-filtering or post-filtering
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
    • G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
    • G10L19/26—Pre-filtering or post-filtering
    • G10L19/265—Pre-filtering, e.g. high frequency emphasis prior to encoding

Definitions

  • the present invention generally relates to quantizers of information signals and, in embodiments, to the quantization of audio signals used, for example, for data compression of audio signals or for audio coding.
  • the present invention relates to audio coding with a short delay time.
  • the currently most popular audio compression method is the MPEG-1 Layer III.
  • the sample and audio values of an audio signal are coded lossy into an encoded signal.
  • irrelevance and redundancy of the original audio signal are reduced or, ideally, removed in the compression.
  • simultaneous and temporal masking is recognized by a psychoacoustic model, i.
  • a time-varying, depending on the audio signal masking threshold is calculated or determined, indicating at which volume from a certain frequency sounds of the human hearing are perceived. This information is in turn used to encode the signal by making the spectral values of the audio signal more accurate, less accurate, or not at all quantized, depending on the masking threshold, and included in the encoded signal.
  • Audio compression methods such as the MP3 format, are then limited in their applicability when it comes to compressing audio data over a bit-rate-limited transmission channel, on the one hand, but with the least possible delay time, on the other hand.
  • the delay time is none Role, such as in the archiving of audio information.
  • low-delay audio coders sometimes called “ultra-low-delay coders,” are necessary where time-critical audio signal transmission is required, such as teleconferencing, wireless speakers, or microphones.
  • the article of Schuller G., etc. “Perceptual Audio Coding using Adaptive Pre- and Post-Filters and Lossless Compression", IEEE Transactions on Speech and Audio Processing, Vol. 10, No. 6, September 2002, pp. 379-390 , proposed an audio coding in which the irrelevance reduction and the redundancy reduction are performed not on the basis of a single transformation, but on two separate transformations.
  • the coding is based on an audio signal 902 which has already been sampled and therefore already exists as a sequence 904 of audio or sample values 906, wherein the time sequence of the audio values 906 is indicated by an arrow 908.
  • a listening threshold is calculated by means of a psychoacoustic model.
  • Fig. 13 shows a graph in which the spectrum of a signal block of 128 audio values 906 is plotted over the frequency f with the curve a, and the logarithmic threshold, as calculated by a psychoacoustic model, is plotted in logarithmic units at b.
  • the masking threshold indicates, as already mentioned, to what intensity frequencies are inaudible to the human ear, namely all tones below the masking threshold b.
  • an irrelevance reduction is now achieved by controlling a parameterizable filter followed by a quantizer.
  • a parameterization is calculated such that the frequency response thereof is the inverse of the Amount of the masking threshold corresponds. This parameterization is indicated in FIG. 12 by x # (i).
  • a constant step quantization is performed, such as a rounding operation to the nearest integer.
  • the resulting quantization noise is white noise.
  • the filtered signal is again "back-transformed" with a parameterizable filter whose transfer function is set to the amount of the masking threshold itself.
  • a gain value a # which is applied to the filtered signal before quantization, is also calculated on the coder side for each parameter set or for each parameterization.
  • the amplification value a and the parameterization x are transmitted as side information 910 in addition to the actual main data, namely the quantized, filtered audio values 912, to the coder.
  • this data ie page information 910 and main data 912, is still subjected to lossless compression, namely entropy coding, whereby the coded signal is obtained.
  • the above article proposes a size of 128 samples 906 as a block size. This allows a relatively short delay of 8 ms at 32 kHz sample rate.
  • the page information namely the coefficients x # and a #
  • the page information are transmitted only if there is enough change compared to a previously transmitted parameter set, ie if the change is a exceeds certain threshold.
  • the implementation is preferably made such that a current parameter set is not directly applied to all samples associated with the respective block, but linear interpolation of the filter coefficients x # is used to avoid audible artifacts.
  • a lattice structure is proposed for the filter to prevent the occurrence of instabilities.
  • the article still proposes to selectively multiply or attenuate the filtered signal scaled by the time-varying gain factor a by a factor other than 1, so that audible perturbations occur arise, but the bit rate can be reduced at consuming to be coded points of the audio signal.
  • a problem with the above scheme is that of having to transmit the masking threshold or transfer function of the coder side filter, hereinafter referred to as prefilter , the transmission channel is relatively heavily loaded, although yes the filter coefficients are transmitted only when a predetermined threshold is exceeded.
  • a further disadvantage of the above coding scheme is that, due to the fact that the masking threshold or the inverse thereof must be provided by the parameter set x # decoder-side to be transmitted, a compromise between on the one hand the lowest possible bit rate or a high compression ratio and on the other hand a precise approximation or parameterization of the masking threshold or the inverse thereof is to be made. It is therefore inevitable that the above by the audio coding scheme to the masking threshold matched quantization noise at some frequency ranges exceeds the masking threshold and therefore causes the listener audible audio interference. For example, with the curve c, FIG. 13 shows the parameterized frequency response of the decoder-side parameterizable filter.
  • the filtered signal may take an unpredictable form due to frequency selective filtering, in particular due to random superposition of many individual harmonics, a single or individual audio values of the encoded signal to sum up to very high values, which in turn lead to a lower compression ratio due to their rare occurrence in the subsequent redundancy reduction.
  • the object of the present invention is to provide a method and apparatus for quantizing an information signal so that with only a slight deterioration in the quality of the original information signal a higher data compression of the information signal can be realized.
  • Quantizing an information signal of a sequence of information values comprises frequency-selectively filtering the sequence of information values to obtain a sequence of filtered information values, and quantizing the filtered information values to obtain a sequence of quantized information values using a quantization step function comprising the filtered ones Maps information values to the quantized information values whose history is steeper below a threshold information value than above the threshold information value.
  • truncation of the filtered information signal above an appropriate threshold is twice the largest possible value of the original information signal to be filtered so that the artifacts artificially generated by the frequency selective filtering result from the filtered Information signal are removed or smoothed, after back-filtering hardly leads to a deterioration in the quality of the information signal filtered back after quantization, but the truncation or increasing the quantization step size above a suitable Threshold provides enormous savings in a bit representation of the filtered information signal.
  • the information signal is an audio signal in which the selective quantization above or below a certain threshold leads to a hardly audible audio quality reduction while at the same time enormously reducing the bit representation.
  • the quantization step function can alternatively be provided in order to quantize all audio values above the threshold value to a highest quantization level, or a quantization step function is used which is flatter above the threshold value or has a larger quantization step width above the threshold value, so that the artificially generated artifacts be quantified coarser.
  • Fig. 1 shows an audio encoder according to an embodiment of the present invention.
  • the audio coder indicated generally at 10, initially includes a data input 12 at which it receives the audio signal to be coded which, as will be explained later with reference to Figure 5a, is comprised of a sequence of audio values and a data output at which the coded signal is output, the information content of which will be discussed in more detail with reference to FIG. 5b.
  • the audio coder 10 of FIG. 1 is divided into an irrelevance reduction part 16 and a redundancy reduction part 18.
  • the irrelevance reduction part 16 comprises a hearing threshold detection means 20, a gain value calculation means 24, a parameterization calculation means 24, a node comparison means 26, a quantizer 28 and a parameterizable prefilter 30 and an input FIFO (first-in-first-out) buffer 32, a buffer or memory 38 and a multiplier or multiplier 40.
  • the redundancy reduction part 18 comprises a compressor 34 and a bit rate controller 36.
  • Irrelevance reduction part 16 and redundancy reduction part 18 are connected in series in this order between data input 12 and data output 14.
  • the data input 12 is connected to a data input of the means 20 for determining a listening threshold and a data input of the input buffer 32.
  • a data output of the means 20 for detecting a masking threshold is connected to an input of the means 24 for calculating a parameterization as well as to a data input of the device 22 for calculating a gain value in order to forward to the same a detected listening threshold.
  • the devices 22 and 24 calculate a parameterization or gain value based on the masking threshold and are connected to the node comparator 26 for forwarding to them these results.
  • the interpolation point comparison device 26 forwards, depending on a comparison result, as will be discussed below, the results calculated by the devices 22 and 24 as input parameter or parameterization to the parameterizable prefilter 30.
  • the parameterizable prefilter 30 is connected between a data output of the input buffer 32 and a data input of the buffer 38. Between a data output of the buffer 38 and the quantizer 28, the multiplier 40 is connected.
  • the quantizer 28 forwards to the redundancy reduction part 18 optionally multiplied or, in any case, quantized, filtered audio values, specifically to a data input of the compressor 34.
  • the node comparator 26 forwards to the redundancy reduction part 18 information from which the the parameterizable prefilter 30 forwarded input parameters are derivable, specifically to a further data input of the compressor 34.
  • the bit rate control is connected via a control connection to a control input of the multiplier 40, in order to ensure that the quantized filtered audio values as obtained from the prefilter 30 are multiplied by the multiplier 40 with a suitable multiplier, as will be discussed in greater detail below.
  • the bit rate controller 36 is connected between a data output of the compressor 34 and the data output 14 of the audio encoder 10 to appropriately determine the multiplier for the multiplier 40. On the first pass of each audio value through the quantizer 40, the multiplier is initially set to an appropriate scaling factor, such as 1.
  • the buffer 38 continues to store each filtered audio value to give the bitrate controller 36 the opportunity to multiply the bit rate controller 36 as described below to change another pass of a block of audio values. If such a change is not indicated by the bitrate controller 36, the buffer 38 may release the memory occupied by that block.
  • the audio signal when it reaches the data input 12 has already been obtained by audio signal sampling 50 from an analog audio signal.
  • the audio signal sampling is performed at a predetermined sampling frequency, which is usually between 32-48 kHz.
  • a predetermined sampling frequency which is usually between 32-48 kHz.
  • the audio signal which consists of a sequence of sample or audio values.
  • the encoding of the audio signal is not block-based, the audio values at the data input 12 are first combined into audio blocks in a step 52.
  • the summary of the audio blocks is, as will become apparent from the following description, only for purposes of determining the Mit Craftschwelle and takes place in an input stage of the means 20 for determining a listening threshold.
  • FIG. 5a shows at 54 the sequence of samples, each sample being illustrated by a rectangle 56.
  • the samples are numbered for purposes of illustration, again with only some of the samples of sequence 54 shown for clarity.
  • each 128 consecutive samples are combined into a block, with the immediately succeeding 128 samples forming the next block.
  • the summary into blocks could be made otherwise, for example, by overlapping blocks or spaced blocks and blocks of a different block size, although the block size of 128 is again preferred because it provides a good compromise between high audio quality and on the other hand provides the lowest possible delay time.
  • step 52 in device 20 While the audio blocks summarized in step 52 in device 20 are processed block by block in means 20 for determining a listening threshold, the incoming audio values are buffered or buffered 54 in input buffer 32 until parameterizable pre-filter 30 receives input parameters from interpolation point comparator 26 has to pre-filter, as will be described below.
  • the means 20 for determining a listening threshold starts its processing immediately after sufficient audio values have been received at the data input 12 to form an audio block or form the next audio block, which the device 20 checks by means of a check monitored in step 60. If an editable complete audio block does not yet exist, the device waits for 20. If there is a complete audio block to be processed, the means 20 for determining a listening threshold calculates a listening threshold in a step 62 on the basis of a suitable psychoacoustic model in a step 62. In order to illustrate the listening threshold, reference is again made to FIG.
  • the masking threshold which is determined in step 62, is a frequency-dependent function that may vary for successive audio blocks, and also from audio signal to audio signal, e.g. from rock to classical music pieces, can vary significantly.
  • the listening threshold indicates for each frequency a threshold below which the human ear can not perceive disturbances.
  • the device 24 now calculates the parameterization a k t such that the transfer function H (f) of the parametrizable prefilter 30 is approximately equal to the inverse of the amount of the masking threshold M (f), ie such that H f ⁇ t ⁇ 1 M f ⁇ t , where the dependence on t is again to illustrate that for different audio blocks the masking threshold M (f) changes.
  • the filter coefficients become a k t the following is obtained: the inverse discrete Fourier transform of
  • a lattice structure is preferably used for the filter 30, wherein the filter coefficients for the lattice structure are converted into reflection coefficients be reparameterized.
  • means 22 calculates a noise power limit based on the monitor threshold, namely, a limit that indicates which noise power quantizer 28 has in that may be introduced by the pre-filter 30 filtered audio signal, so that the quantization noise after the back and post filtering on the decoder side below the Mit Selfschwelle M (f) or is exactly on the same.
  • the device 22 calculates this noise power limit as the area below the magnitude square of the listening threshold M, ie as ⁇
  • the gain value a is calculated by the noise power limit device 22 by calculating the root from the fraction of quantization noise power through the noise power limit.
  • the quantization noise is the noise caused by the quantizer 28.
  • the noise caused by the quantizer 28 is, as will be described, white noise and thus frequency independent.
  • the quantization noise power is the power of the quantization noise.
  • the device 22 calculates the noise power limit in addition to the gain value a. Although it is possible for the node comparator 26 to recalculate the noise power limit from the gain value a obtained by the device 22, it is further possible that the device 22 also transmits the detected noise power limit in addition to the gain value a of the node comparison device 26.
  • the interpolation point comparison device 26 checks in a step 66 whether the parameterization just calculated differs by more than a predetermined threshold from the current parameterization last routed to the parameterizable prefilter. If the check in step 66 reveals that the parameterization just calculated differs from the current one by more than the predetermined threshold, the just calculated filter coefficients and the just calculated gain value or noise power limit are buffered in the interpolation means 26 for an interpolation to be discussed and the node comparator 26 transfers the just-calculated filter coefficients in a step 68 and the just-calculated gain value to the prefilter 30 in a step 70.
  • the support point comparison device (26) transfers to the pre-filter 30 in step 72 instead of the parameterization just calculated only the current support point parameterization, ie the parameterization that resulted in the last time in step 66 to a positive result, ie, differed by more than a predetermined threshold from a previous node parameterization.
  • the process of Fig. 3 returns to processing the next audio block, i. to query 60, back.
  • preprocessor 30 applies this interpolation parameterization to all samples of this audio block located in FIFO 32, as will be described in more detail below, whereby this current block is extracted from FIFO 32 and quantizer 28 obtains a resulting audio block receives pre-filtered audio values.
  • Fig. 4 illustrates the operation of the parameterizable pre-filter 30 in more detail for the case that it receives the just calculated parameterization and the just calculated gain value because they are sufficiently different from the current support point parameterization.
  • processing of FIG. 4 does not take place for each of the successive audio blocks, but only for audio blocks in which the associated parameterization sufficiently differs from the current support point parameterization.
  • the other audio blocks are, as described above, pre-filtered by applying the respective current node parameterization and the respective respective current gain value to all samples of these audio blocks.
  • the parameterizable prefilter 30 monitors whether a transfer of just calculated filter coefficients from the Support point comparison device 26 has taken place or from older support point parameterizations. The prefilter 30 performs the monitoring 80 until such a handover has occurred.
  • the parameterizable prefilter 30 begins processing the current audio block of audio values that is currently in the cache 32, that is, the one to which the parameterization has just been calculated.
  • Fig. 5a for example, it has been illustrated that all the audio values 56 have already been processed before the audio value with the number 0 and therefore the memory 32 already have happened.
  • the processing of the block of audio values before the number 0 audio value was then triggered because the parameterization calculated for the audio block before block 0, namely x 0 (i), is more than the predetermined threshold of the previously compared to the prefilter 30 forwarded support point parameterization.
  • the parameterization x 0 (i) is thus a support point parameterization, as it is called in the present invention.
  • the processing of the audio values in the audio block before the audio value 0 was performed based on the parameter set a 0 , x 0 (i).
  • the parameterization calculated for block 1 still in FIFO 32 differed by more than the predetermined threshold from parameterization x 0 (i) according to the exemplary example of FIG. 5 a and was therefore passed to the prefilter in step 68 30 as parameterization x 1 (i) forwarded together with the gain value a 1 (step 70) and optionally the associated noise power limit, wherein the indices of a and x in Fig. 5a should be an index for the nodes, as in the later zu discussing interpolation performed on samples 128-255 in block 1, symbolized by an arrow 82, and through steps following step 80 in FIG. 4 is realized.
  • processing would begin at step 80.
  • the pre-filter 30 determines the noise power limit q 1 corresponding to the gain value a 1 in step 84. This may be done by having the node comparison means 26 pass this value to the prefilter 30, or by re-calculating this value by the prefilter 30, as described above with reference to step 64.
  • an index j is initialized to a sample to point to the oldest sample remaining in the FIFO memory 32, or the first sample of the current audio block "Block 1", ie, in the present example of FIG
  • the parametrizable prefilter performs an interpolation between the filter coefficients x 0 and x 1 , whereby the parameterization x 0 as a base value at the support point with audio value number 127 of the previous block 0 and the parameterization x 1 as a base value at the Support point with audio value 255 of the current block 1 applies.
  • These audio value positions 127 and 255 are also referred to below as support point 0 and support point 1, wherein in FIG. 5a the reference points relating to the support point parameterizations are indicated by the arrows 90 and 92.
  • the parameterizable pre-filter 30 performs interpolation between the noise power limit q 1 and q 0 to obtain an interpolated noise power limit at the sample position j, ie, q (t j ).
  • the parameterizable prefilter 30 calculates the sample position gain value j on the basis of the interpolated noise power limit and the quantization noise power, and preferably also the interpolated filter coefficients, for example, depending on the root quantization noise q t j . Reference is made to the comments on step 64 of FIG. 3.
  • the parameterizable prefilter 30 then applies the calculated gain value and the interpolated filter coefficients to the sample at the sample position j to obtain a filtered sample for that sample position, s' (t j ).
  • the parameterizable pre-filter 30 checks whether the sampling position j has reached the current interpolation point, ie interpolation point 1, in the case of FIG. 5a the sampling position 255, ie the sample value for which the parameterization communicated to the parametrizable pre-filter 30 plus gain value directly, ie without interpolation. If this is not the case, the parameterizable prefilter 30 increments or increments the index j by 1, wherein the steps 88-96 are repeated again.
  • step 100 the parametrizable prefilter applies the amplification value last transmitted by the interpolation point comparator 26 and the from the interpolation point 26 last transmitted filter coefficients immediately without interpolation to the sample at the new interpolation point, whereupon the current block, ie in the present case, the block 1, has been executed, and the process again at step 80 with respect to the subsequent block to be processed which, depending on whether the parameterization of the next audio block block 2 differs sufficiently from the parameterization x 1 (i), may possibly be just this next audio block block 2 or else a later audio block.
  • the purpose of the filtering is to filter the audio signal at the input 12 with an adaptive filter whose transfer function is constantly adapted as optimally as possible to the inverse of the listening threshold, which also changes in time.
  • the reason for this is that, on the decoder side, the back-filtering by an adaptive filter, whose transfer function is correspondingly constantly matched to the monitoring threshold, reduces the white quantization noise introduced by quantizing the filtered audio signal; the frequency constant quantization noise, forms, namely adapts to the shape of the Mit Schulschwelle.
  • the application of the gain value in steps 94 and 100 in the prefilter 30 consists in a multiplication of the audio signal or the filtered audio signal, ie the samples s or the filtered samples s', by the amplification factor.
  • the purpose is, as far as possible, the quantization noise, which is inserted into the filtered audio signal by the quantization described in more detail below, and which is adapted by the back-filtering on the decoder side to the form of the monitoring threshold set so high that it does not exceed the Mit tolerateschwelle yet.
  • the quantization noise which is inserted into the filtered audio signal by the quantization described in more detail below, and which is adapted by the back-filtering on the decoder side to the form of the monitoring threshold set so high that it does not exceed the Mit Bioschwelle yet.
  • the quantization noise power is also reduced, namely by the factor a -2 , where a is the amplification value.
  • the quantization noise power can be set optimally high, which is equivalent to increasing the quantization step size and thus reducing the number of quantization steps to be coded, which in turn increases the compression in the subsequent redundancy reduction part.
  • the effect of the pre-filter can be regarded as a normalization of the signal to its masking threshold, so that the level of the quantization noise and the quantization noise both in time and frequency can be kept constant. Since the audio signal is in the time domain, therefore, the quantization can be performed stepwise with uniform constant quantization, as will be described later. In this way, ideally, any irrelevance is removed from the audio signal, and a lossless compression scheme may be used to also remove the remaining redundancy in the pre-filtered and quantized audio signal, as will be described below.
  • the filter coefficients and amplification values a 0 , a 1 , x 0 , x 1 used must be available as page information on the decoder side, but this reduces the transmission effort for this purpose will not simply reuse new filter coefficients and new gain values for each block. Instead, a threshold value check 66 takes place in order to transmit the parameterizations as page information only if the parameterization change is sufficient, and otherwise the page information or parameterizations are not transmitted. On the audio blocks for which the parameterizations were transferred, an interpolation from the old to the new parameterization takes place over the area of these blocks.
  • the interpolation of the filter coefficients takes place in the manner described above with reference to step 88.
  • the interpolation with regard to the amplification takes place via a detour, namely via a linear interpolation 90 of the noise power limit q 0 , q 1 .
  • the linear interpolation with respect to the noise power limit leads to a better hearing result or less audible artifacts.
  • the filtered samples output from the parameterizable prefilter 30 are stored in the buffer 38 and simultaneously passed from the buffer 38 to the multiplier 40 where, in turn, as it is its first pass, it passes through unaltered, namely with a scaling factor of unity the multiplier 40 are passed to the quantizer 28.
  • the filtered audio values above an upper bound are cut off in a step 110 and then quantized in a step 112.
  • the two steps 110 and 112 are performed by the quantizer 28.
  • the two steps 110 and 112 are preferably executed by the quantizer 28 in one step by quantizing the filtered audio values s' with a quantization step function comprising the For example, in a floating-point representation, filtered samples s' are mapped to a plurality of integer quantization level indices and flat at a certain threshold for the filtered samples, such that filtered samples greater than the threshold are quantized to one and the same quantization level become.
  • a quantization step function is shown in Fig. 7a.
  • the quantized filtered samples are designated ⁇ 'in Fig. 7a.
  • the quantization step function is preferably a quantization step function below the constant step threshold, i. the jump to the next quantization step always takes place after a constant interval along the input values S '.
  • the step size to the threshold is set such that the number of quantization steps preferably corresponds to a power of 2. In comparison to the floating-point representation of the incoming filtered samples s', the threshold value is smaller, so that a maximum value of the representable range of the floating-point representation exceeds the threshold value.
  • the reason for the threshold is that it has been observed that the filtered audio signal output from the prefilter 30 occasionally has audio values that accumulate to very large values due to unfavorable accumulation of harmonics. Furthermore, it has been observed that truncating these values, as achieved by the quantization step function shown in Figure 7a, results in a high data reduction but only a slight degradation in audio quality. Rather, these isolated spots in the filtered audio signal artificially caused by the frequency-selective filtering in the parameterizable filter 30, so that a truncation of the same affects the audio quality only slightly.
  • FIG. 7a A more specific example of the quantization step function shown in Figure 7a would be one which up to the threshold rounds all filtered samples s' to the nearest integer, and from there quantizes all overlying filtered samples to the highest quantization level, e.g. 256. This case is shown in Fig. 7a.
  • FIG. 7b Another example of a possible quantization step function would be that shown in Fig. 7b.
  • the quantization step function of FIG. 7b corresponds to that of FIG. 7a.
  • the quantization step function continues with a slope that is less than the slope in the region below the threshold.
  • the quantization step size is larger. This achieves a similar effect as with the quantization function of FIG. 7a, but with on the one hand more effort due to the different step sizes of the quantization step function above and below the threshold value and on the other hand improved audio quality, since very high filtered audio values s' are not completely cut off, but rather only be quantized with a larger quantization step size.
  • the compressor 34 makes a first compression attempt and compresses side information including the gain values a 0 and a 1 at the nodes, such as 127 and 255, and the filter coefficients x 0 and x 1 at the nodes and the quantized filtered samples ⁇ 'into a preliminary filtered signal.
  • the compressor 34 is a lossless encoder, such as a Huffman or arithmetic encoder with or without prediction and / or adaptation.
  • the memory 38 which the sampled audio values ⁇ 'pass through serves as a buffer for a suitable block size, with which the compressor 34 processes the quantized, filtered and optionally scaled, audio values ⁇ ' output by the quantizer 28.
  • the block size may differ from the block size of the audio blocks as used by the device 20.
  • the bit rate controller 36 has driven the multiplier 40 at a multiplier of 1 so that the filtered audio values from the prefilter 30 will pass unchanged to the quantizer 28 and from there to the compressor 34 as quantized filtered audio values.
  • Compressor 34 in step 116, monitors whether a certain amount of compression block, ie a certain number of quantized sampled audio values, has been coded into the preliminary coded signal or whether to encode further quantized filtered audio values ⁇ 'into the current provisional coded signal are. If the compression block size is not reached, the compressor 34 continues the current compression 114.
  • bit rate controller 36 checks in step 118 whether the amount of bit needed for compression is greater than a bit amount prescribed by a desired bit rate. If this is not the case, the bit rate controller 36 checks in a step 120 whether the required bit quantity is smaller than the bit quantity prescribed by the desired bit rate. If so, the bitrate controller 36 adds the encoded signal in step 122 with stuffing bits until the achieved by the desired bit rate prescribed amount of bits is reached. Subsequently, in step 124, the output of the coded signal.
  • bitrate controller 36 could store the last compression-based compression block of filtered audio values ⁇ 'in the memory 38 multiplied by a multiplier greater than 1 multiplied by the multiplier 40 to the quantizer 28 to re-run steps 110-118 until the bit rate prescribed by the desired bit rate is reached, as indicated by a dashed step 125.
  • bit rate controller 36 changes the multiplier for the multiplier 40 to a factor between 0 and 1 only. This is done in step 126.
  • bitrate control 36 causes memory 38 to reissue the last block of filtered audio values ⁇ 'underlying the compression, then multiply them by the factor set in step 126 and re-apply to quantizer 28, whereupon Steps 110-118 are performed again and the previously provisionally coded signal is discarded.
  • step 114 the factor used in step 126 (or step 125) is also incorporated into the coded signal.
  • step 126 The point of the procedure after step 126 is that the factor increases the effective step size of the quantizer 28. This means that the resulting quantization noise is evenly above the masking threshold, resulting in audible or audible noise Noise but results in a reduced bit rate. If, after re-running steps 110-116, it is again determined in step 118 that the required bit quantity is greater than the one prescribed by the desired bit rate, the factor is further reduced in step 126, etc.
  • FIG. 5b again illustrates the resulting coded signal, indicated generally at 130.
  • the encoded signal includes page information and intermediate main data.
  • the side information comprises, as already mentioned, information from which for special audio blocks, namely audio blocks in which a significant change in the filter coefficients has resulted in the sequence of audio blocks, the value of the gain value and the value of the filter coefficients can be derived.
  • the page information further includes further information relating to the gain value used for the bit control. Owing to the interdependence between the gain value and the noise power limit q, the side information may optionally include, in addition to the gain value a # to a node #, the noise power limit q # , or even the latter.
  • the page information is preferably arranged such that the page information is related to filter coefficients and associated gain value are arranged before the main data for the audio block of quantized filtered audio values ⁇ ', from which these filter coefficients with associated gain value and associated noise power limit have been derived, ie the page information a 0 , x 0 (i) after the block -1 and the page information a 1 , x 1 (i) after the block 1.
  • the main data, ie the quantized filtered audio values ⁇ ' are exclusive of an audio block of the type which results in a significant change in the filter coefficients due to audio blocks For example, in Figs.
  • the audio values ⁇ '(t 0 ) - ⁇ ' (t 127 ) are as in the above bez Referring to FIG.
  • page information 132 alone is decodable, while the audio values ⁇ '(t 128 ) - ⁇ ' (t 255 ) are obtained by interpolation by means of the page information 132 as support values at the sample point 127 and by The page information 134 has been obtained as a support value at the interpolation point with the sample number 255 and thus can only be decoded by means of both side information.
  • page information block 132 includes gain value a 0 and filter coefficients x 0 with respect to the interpolation point at time t -1 .
  • these values are derivable from the block itself.
  • Side information concerning the interpolation point at time t 255 can not be deduced from this block alone.
  • the page information block 134 only includes information about differences in the gain value a 1 of the interpolation point at time t 255 to the gain value of the interpolation point at time to and the differences of the filter coefficients x 1 to the filter coefficients x 0 .
  • the page information block 134 thus contains only the information about a 1 - a 0 and x 1 (i) - x 0 (i).
  • the filter coefficients and the gain value should be transmitted fully and not only as a difference to the previous sample, eg every second to allow a receiver or decoder to latch into a stream of encoding data as it does will be discussed below.
  • an embodiment of an audio decoder capable of decoding the encoded signal produced by the audio encoder 10 of Fig. 1 into a decoded playable audio signal will now be described.
  • the decoder indicated generally at 210, includes a decompressor 212, a FIFO memory 214, a multiplier 216 and a parameterizable postfilter 218.
  • Decompressor 212, FIFO memory 214, multiplier 216 and parameterizable postfilter 218 are connected in this order between a data input 220 and a data output 222 of the decoder 210, wherein at the data input 220 the coded signal is obtained and at the data output 222 the decoded audio signal is output, which differs from the original audio signal at the data input 12 of the audio coder 10 only by the quantization noise generated by the quantizer 28 in the audio coder 10.
  • the decompressor 212 is connected at a further data output to a control input of the multiplier 216 for forwarding thereto a multiplier, and via another data output to a parameterization input of the parameterizable postfilter 218.
  • the decompressor 212 first decompresses the compressed signal applied to the data input 220 to provide the quantized filtered audio data, namely the samples ⁇ ', and the associated page information in the page information blocks 132, 134 to enter yes, the filter coefficients and gain values or, instead of the gain values, the noise power limits, indicate at the nodes.
  • the decompressor 212 checks the decompressed signal in the order of its arrival whether it contains page information with filter coefficients in self-contained form without reference to a previous page information block. In other words, the decompressor 212 searches for the first page information block 132. Once the decompressor 212 has been found, the quantized filtered audio values ⁇ 'are buffered in the FIFO memory 214 in a step 228.
  • step 228 it will first be provided with parameterization and gain value information contained in the page information received in step 226 Postfilter post-filtered and amplified in the multiplier 216, whereby it is decoded and thus the associated decoded audio block is obtained.
  • the decompressor 212 monitors the decompressed signal for the appearance of some kind of page information block, namely with absolute filter coefficients or filter coefficient differences to a previous page information block. For example, in the example of FIG. 5b, decompressor 212 would recognize the appearance of page information block 132 at step 226 in step 230 at step 230.
  • step 2208 the block of quantized filtered audio values ⁇ '(t 0 ) - ⁇ ' (t 127 ) would have been decoded using page information 132.
  • page information block 134 does not yet appear in the decompressed signal the caching and eventual decoding of blocks is continued through step 226, as described above, at step 228.
  • decompressor 212 calculates the parameter values at node 1, ie, a 1 , x 1 (i), by summing the difference values in page information block 134 to the parameter values in page information block 132 at step 232.
  • step 232 will be omitted, which will be the case, for example, every second, as described above
  • page information blocks 132 where the parameter values are absolute, that is, derivable without relation to another page information block, are arranged at sufficiently small intervals so that the on time and the dead time, respectively Turning on the audio encoder 210 in example, a radio transmission or broadcasting is not too large.
  • the number of page information blocks 134 therebetween with the difference values are also arranged in a fixed predetermined number between the page information blocks 132, so that the decoder knows when to expect a page information block of the type 132 in the coded signal again.
  • the various page information block types are indicated by corresponding flags.
  • a sample index j is first initialized to 0 in step 234. This value corresponds to the sample position of the first sample in the audio block currently remaining in the FIFO 214 to which the current page information relates.
  • Step 234 is performed by the parameterizable postfilter 218.
  • the postfilter 218 then performs a calculation of the noise power limit at the new interpolation point in a step 236, which step corresponds to step 84 of FIG. 4 and may be omitted if, for example, the smoke power limit at the interpolation points is transmitted in addition to the amplification values.
  • the postfilter 218 then performs interpolations on the filter coefficients and noise power limits corresponding to the interpolations 88 and 90 of FIG.
  • the subsequent calculation of the gain value for the sample position j based on the interpolated noise power limit and the interpolated filter coefficients from steps 238 and 240 in step 242 corresponds to step 92 of FIG. 4.
  • the postfilter 218 then applies the gain value calculated in step 242 and the interpolated filter coefficients to the sample at the sample position j. This step differs from the step 94 of Fig.
  • the interpolated filter coefficients are applied to the quantized filtered samples ⁇ 'such that the transfer function of the parameterizable postfilter does not correspond to the inverse of the listening threshold but to the listening threshold itself.
  • the post-filter does not perform a multiplication on the gain value but a division by the gain value on the quantized filtered sample ⁇ 'or already filtered-back, quantized filtered sample at the position j.
  • step 246 If the postfilter 218 has not yet reached the current sample point at the sample position j, which checks it in step 246, it increments the sample position index j at step 248 and restarts steps 238-246. Only when the interpolation point is reached does it apply the gain value and the filter coefficients of the new interpolation point to the sample at the interpolation point, namely in step 250. Again, as in step 218, the application involves division by means of the amplification value and filtering, instead of multiplication a transfer function equal to the listening threshold and not the inverse of the latter. After step 250, the current audio block is decoded by interpolation between two node parameterizations.
  • the filtering and the application of the gain value in steps 218 and 224 match the noise introduced by the quantization in the encoding in steps 110 and 112 both in shape and in height to the masking threshold.
  • a parameterization and gain value or noise power limit, as determined for a particular audio block should be considered immediately valid for a particular audio value, as in the previous embodiment, the last audio value of each audio block, ie of the 128th value in this audio block, so that for this audio value the interpolation can be omitted.
  • the parameterization determined for one audio block or the gain value determined for that audio block may also be directly applied to another value, such as the audio value in the middle of the audio block, such as the 64th audio value in the case of the above block size of 128 audio values.
  • the above embodiment referred to an audio coding scheme designed to generate a coded signal at a controlled bit rate.
  • bitrate control is not required in every application. Therefore, the corresponding steps 116-122 and 126 and 125, respectively, may also be omitted.
  • the present invention has been described above with respect to a particular audio coding scheme that allows for short delay times, the present invention is of course applicable to other audio codings as well.
  • an audio coding scheme would also be conceivable in which the coded signal consists of the quantized, filtered audio values per se, without a redundancy reduction is carried out. Accordingly, it would also be conceivable to carry out the frequency-selective filtering differently than the previously described manner, namely on the coder side with a transfer function equal to the inverse of the masking threshold and on the decoder side with a transfer function equal to the masking threshold.
  • individual aspects of the above embodiments may also be omitted.
  • the present invention is not limited to audio signals. It is also applicable to other information signals, such as video signals consisting of a sequence of frames, i. a sequence of pixel arrays.
  • the above audio coding scheme provides a way to limit the bit rate in an audio coder with very little delay time.
  • the bitrate peaks arising in the coding as a function of the audio signal are avoided by limiting the output value range of the pre-filter.
  • an upper limit on the transmission bit rate can often be met, often for example exists on wireless transmission media.
  • the change of the quantization step function above the threshold is a suitable means for limiting the bit rate to the maximum allowed.
  • the encoder consisted of a prefilter that appropriately shapes the audio signal, a quantizer with a quantizer step height, followed by an entropy coder.
  • the quantizer generated values also called indices.
  • higher indices also mean a higher bit rate associated with them, but this has been avoided by having the range of indices limited ( Figure 7a) or thinned out ( Figure 7b), but with the potential for degrading audio quality.
  • the threshold value in the quantization always remains constant or the quantization step function always remains constant, ie always the artifacts generated in the filtered audio signal with coarser quantization or cut off, whereby the audio quality may be audibly deteriorated
  • this measure only when it requires the complexity of the audio signal, namely when the bit rate required for coding exceeds a desired bit rate. In this case, in addition to those shown in Figs. 7a and 7b.
  • quantization step functions are used with a quantization step size constant over the entire possible value range at the output of the prefilter and the quantizer would respond to a signal, for example, to use either the quantization step function with always constant quantization step size or one of the quantization step functions of Figure 7a or 7b so that the signal could tell the quantizer to perform the quantization step reduction above the threshold or the clip above the threshold with slight audio quality degradation.
  • the threshold could also gradually increase be reduced. In this case, the threshold reduction could be performed instead of the factor reduction of step 126.
  • the provisionally compressed signal could only be selectively thresholded in a modified step 126 if the bit rate is still too high (118).
  • the filtered audio values would then be quantized with the quantization step function, which has a flatter course above the audio threshold. Further bit rate reductions could still be performed in the modified step 126 by reducing the threshold value and thus further modifying the quantization step function.
  • the inventive quantization scheme can also be implemented in software.
  • the implementation may be on a digital storage medium, in particular a floppy disk or a CD with electronically readable control signals which may interact with a programmable computer system such that the corresponding method is executed.
  • the invention thus also consists in a computer program product with program code stored on a machine-readable carrier for carrying out the method according to the invention when the computer program product runs on a computer.
  • the invention can thus be realized as a computer program with a program code for carrying out the method when the computer program runs on a computer.
  • the inventive scheme can also be implemented in software.
  • the implementation may be on a digital storage medium, in particular a floppy disk or a CD with electronically readable control signals, which may cooperate with a programmable computer system such that the corresponding method is executed.
  • the invention thus also consists in a computer program product with program code stored on a machine-readable carrier for carrying out the method according to the invention when the computer program product runs on a computer.
  • the invention can thus be realized as a computer program with a program code for carrying out the method when the computer program runs on a computer.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Signal Processing (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Computational Linguistics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Compression, Expansion, Code Conversion, And Decoders (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Image Processing (AREA)
  • Facsimile Image Signal Circuits (AREA)

Claims (10)

  1. Dispositif pour quantifier un signal de données d'une suite de valeurs de données, le signal de données étant un signal audio et les valeurs de données étant des valeurs audio, aux caractéristiques suivantes:
    un moyen (20) pour déterminer un premier seuil d'écoute pour un bloc de valeurs audio parmi une succession de valeurs audio;
    un moyen (24) pour calculer une version d'une paramétrisation d'un filtre paramétrisable de sorte que sa fonction de transmission corresponde environ à l'inverse de la quantité du premier seuil d'écoute;
    un moyen (30) pour filtrer de manière sélective en fréquence la succession de valeurs audio, pour obtenir une succession de valeurs audio filtrées;
    un moyen (28) pour quantifier les valeurs audio filtrées, pour obtenir une succession de valeurs audio quantifiées, au moyen d'une fonction d'étages de quantification qui reproduit les valeurs audio filtrées sur les valeurs audio quantifiées, et dont l'évolution au-dessous d'une valeur de données de seuil est plus abrupte qu'au-dessus de la valeur de données de seuil;
    le moyen (30) pour filtrer de manière sélective en fréquence présentant le moyen suivant:
    un moyen pour filtrer un bloc prédéterminé de valeurs audio de la succession de valeurs audio par le filtre paramétrisable à l'aide d'une paramétrisation prédéterminée qui est fonction de manière prédéterminée de la version de la paramétrisation, pour obtenir un bloc de valeurs audio filtrées.
  2. Dispositif selon la revendication 1, dans lequel le moyen pour déterminer un seuil d'écoute est réalisé de manière à déterminer, par ailleurs, un autre deuxième seuil d'écoute pour un autre, deuxième bloc de valeurs audio, et le moyen pour calculer est réalisé de manière à calculer une version d'une autre, deuxième paramétrisation du filtre paramétrisable de sorte que sa fonction de transmission corresponde environ à l'inverse de la quantité du deuxième seuil d'écoute, le moyen pour filtrer de manière sélective en fréquence présentant les moyens suivants :
    un moyen pour interpoler entre la version de la première paramétrisation et la version de la deuxième paramétrisation, pour obtenir pour une valeur audio prédéterminée du bloc prédéterminé de valeurs audio une version d'une paramétrisation interpolée; et
    un moyen pour appliquer la version de la paramétrisation interpolée à la valeur audio prédéterminée du bloc prédéterminé de valeurs audio.
  3. Dispositif selon la revendication 2, présentant, par ailleurs, un moyen (22) pour déterminer une première limite de puissance de bruit en fonction du premier seuil de masquage et une deuxième limite de puissance de bruit en fonction du deuxième seuil de masquage, et dans lequel le moyen pour filtrer présente un moyen (90) pour interpoler entre la première limite de puissance de bruit et la deuxième limite de puissance de bruit, pour obtenir une limite de puissance de bruit interpolée pour une valeur audio prédéterminée du bloc prédétermine de valeurs audio, un moyen (92) pour déterminer une valeur de modulation intermédiaire en fonction d'une puissance de bruit de quantification créée par une quantification selon une règle de quantification prédéterminée et de la limite de puissance de bruit interpolée, et un moyen (94) pour appliquer la valeur de modulation intermédiaire à la valeur audio prédéterminée, pour obtenir une valeur audio filtrée modulée.
  4. Dispositif selon la revendication 3, dans lequel le moyen pour interpoler effectue, entre la première limite de puissance de bruit et la deuxième limite de puissance de bruit, une interpolation linéaire.
  5. Dispositif selon la revendication 3 ou 4, dans lequel le moyen pour déterminer la valeur de modulation intermédiaire présente un moyen pour former la racine à partir du quotient du bruit de quantification divisé par la limite de puissance de bruit interpolée.
  6. Dispositif selon l'une des revendications précédentes, dans lequel le moyen pour quantifier est réalisé de manière à effectuer la quantification à un signal de régulation.
  7. Dispositif selon l'une des revendications précédentes, présentant, par ailleurs, un moyen de compression sans perte destiné à comprimer les valeurs audio filtrées, pour obtenir un flux audio comprimé, le moyen de compression étant réalisé de manière à réguler un débit binaire du flux audio comprimé, et à envoyer, au cas où le débit binaire est supérieur à une valeur de régulation, le signal de régulation au moyen pour quantifier.
  8. Dispositif selon l'une des revendications précédentes, dans lequel la fonction d'étages de quantification évolue de manière plane au-dessus de la valeur de données de seuil, de sorte que les valeurs audio filtrées supérieures à la valeur de données de seuil soient quantifiées à une valeur d'étage de quantification maximale.
  9. Procédé pour quantifier un signal de données d'une succession de valeurs de données, le signal de données étant un signal audio et les valeurs de données étant des valeurs audio, aux étapes suivantes consistant à:
    filtrer de manière sélective en fréquence la succession de valeurs audio, pour obtenir une succession de valeurs audio filtrées;
    quantifier les valeurs audio filtrées, pour obtenir une succession de valeurs audio quantifiées, au moyen d'une fonction d'étages de quantification qui reproduit les valeurs audio filtrées sur les valeurs audio quantifiées, et dont l'évolution au-dessous d'une valeur de données de seuil est plus abrupte qu'au-dessus de la valeur de données de seuil;
    déterminer un seuil d'écoute pour un bloc de valeurs audio; et
    calculer une version d'une paramétrisation d'un filtre paramétrisable de sorte que sa fonction de transmission corresponde environ à l'inverse de la quantité du premier seuil d'écoute,
    l'étape de filtration sélective en fréquence présentant l'étape suivante consistant à:
    filtrer un bloc prédéterminé de valeurs audio de la succession de valeurs audio par le filtre paramétrisable à l'aide d'une paramétrisation prédéterminée qui est fonction de manière prédéterminée de la version de la paramétrisation, pour obtenir un bloc de valeurs audio filtrées.
  10. Programme d'ordinateur avec un code de programme adapté pour réaliser le procédé selon la revendication 9 lorsque le programme d'ordinateur est exécuté sur un ordinateur.
EP05715289A 2004-02-13 2005-02-10 Procede et dispositif pour quantifier un signal de donnees Active EP1697929B1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102004007184A DE102004007184B3 (de) 2004-02-13 2004-02-13 Verfahren und Vorrichtung zum Quantisieren eines Informationssignals
PCT/EP2005/001343 WO2005078703A1 (fr) 2004-02-13 2005-02-10 Procede et dispositif pour quantifier un signal de donnees

Publications (2)

Publication Number Publication Date
EP1697929A1 EP1697929A1 (fr) 2006-09-06
EP1697929B1 true EP1697929B1 (fr) 2007-10-31

Family

ID=34853461

Family Applications (1)

Application Number Title Priority Date Filing Date
EP05715289A Active EP1697929B1 (fr) 2004-02-13 2005-02-10 Procede et dispositif pour quantifier un signal de donnees

Country Status (16)

Country Link
US (1) US7464027B2 (fr)
EP (1) EP1697929B1 (fr)
JP (1) JP4444295B2 (fr)
KR (1) KR100813193B1 (fr)
CN (1) CN1918630B (fr)
AT (1) ATE377243T1 (fr)
AU (1) AU2005213767B2 (fr)
BR (1) BRPI0506627B1 (fr)
CA (1) CA2555639C (fr)
DE (2) DE102004007184B3 (fr)
ES (1) ES2294685T3 (fr)
HK (1) HK1093814A1 (fr)
IL (1) IL177164A (fr)
NO (1) NO337836B1 (fr)
RU (1) RU2337413C2 (fr)
WO (1) WO2005078703A1 (fr)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102004009955B3 (de) * 2004-03-01 2005-08-11 Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. Vorrichtung und Verfahren zum Ermitteln einer Quantisierer-Schrittweite
US7627481B1 (en) * 2005-04-19 2009-12-01 Apple Inc. Adapting masking thresholds for encoding a low frequency transient signal in audio data
JP4640020B2 (ja) * 2005-07-29 2011-03-02 ソニー株式会社 音声符号化装置及び方法、並びに音声復号装置及び方法
EP1852848A1 (fr) 2006-05-05 2007-11-07 Deutsche Thomson-Brandt GmbH Procédé et appareil d'encodage sans perte d'un signal source utilisant un courant de données encodées avec perte et un courant de données d'extension encodées sans perte
DE102006022346B4 (de) * 2006-05-12 2008-02-28 Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. Informationssignalcodierung
US8655652B2 (en) * 2006-10-20 2014-02-18 Dolby International Ab Apparatus and method for encoding an information signal
US7823092B1 (en) * 2007-11-23 2010-10-26 Altera Corporation Method and apparatus for implementing a parameterizable filter block with an electronic design automation tool
EP2830054A1 (fr) 2013-07-22 2015-01-28 Fraunhofer Gesellschaft zur Förderung der angewandten Forschung e.V. Encodeur audio, décodeur audio et procédés correspondants mettant en oeuvre un traitement à deux canaux à l'intérieur d'une structure de remplissage d'espace intelligent
AU2015291897B2 (en) 2014-07-25 2019-02-21 Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung E.V. Acoustic signal encoding device, acoustic signal decoding device, method for encoding acoustic signal, and method for decoding acoustic signal
DE102014220687A1 (de) * 2014-10-13 2016-04-14 Continental Automotive Gmbh Kommunikationsvorrichtung für ein Fahrzeug und Verfahren zum Kommunizieren
RU2754497C1 (ru) * 2020-11-17 2021-09-02 федеральное государственное автономное образовательное учреждение высшего образования "Казанский (Приволжский) федеральный университет" (ФГАОУ ВО КФУ) Способ передачи речевых файлов по зашумленному каналу и устройство для его реализации

Family Cites Families (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3506912A1 (de) 1985-02-27 1986-08-28 Telefunken Fernseh Und Rundfunk Gmbh, 3000 Hannover Verfahren zur uebertragung eines audiosignals
DE3820038A1 (de) * 1988-06-13 1989-12-14 Ant Nachrichtentech Verfahren zur aufbereitung und uebertragung einer bildsequenz
DE3820037A1 (de) 1988-06-13 1989-12-14 Ant Nachrichtentech Bildcodierverfahren und einrichtung
JP3186290B2 (ja) * 1993-01-20 2001-07-11 ソニー株式会社 符号化方法、符号化装置、復号化装置及び記録媒体
US5581653A (en) * 1993-08-31 1996-12-03 Dolby Laboratories Licensing Corporation Low bit-rate high-resolution spectral envelope coding for audio encoder and decoder
DE19537338C2 (de) * 1995-10-06 2003-05-22 Fraunhofer Ges Forschung Verfahren und Vorrichtung zum Codieren von Audiosignalen
GB2307833B (en) * 1995-12-01 2000-06-07 Geco As A data compression method and apparatus for seismic data
ATE249672T1 (de) * 1996-07-05 2003-09-15 Univ Manchester Sprachkodier- und dekodiersystem
US6370477B1 (en) * 1996-11-22 2002-04-09 Schlumberger Technology Corporation Compression method and apparatus for seismic data
US6131084A (en) 1997-03-14 2000-10-10 Digital Voice Systems, Inc. Dual subframe quantization of spectral magnitudes
KR100335609B1 (ko) 1997-11-20 2002-10-04 삼성전자 주식회사 비트율조절이가능한오디오부호화/복호화방법및장치
US6195633B1 (en) * 1998-09-09 2001-02-27 Sony Corporation System and method for efficiently implementing a masking function in a psycho-acoustic modeler
KR100440896B1 (ko) * 1998-12-30 2004-09-18 주식회사 대우일렉트로닉스 양자화복호화장치
KR100893281B1 (ko) * 1999-04-16 2009-04-17 돌비 레버러토리즈 라이쎈싱 코오포레이션 오디오 코딩을 위해 이득-적응형 양자화 및 비균일 심볼길이를 사용하는 방법 및 장치
US6370502B1 (en) * 1999-05-27 2002-04-09 America Online, Inc. Method and system for reduction of quantization-induced block-discontinuities and general purpose audio codec
WO2001033718A1 (fr) 1999-10-30 2001-05-10 Stmicroelectronics Asia Pacific Pte Ltd. Procede de codage des coefficients de frequence dans un codeur ac-3
US7110953B1 (en) * 2000-06-02 2006-09-19 Agere Systems Inc. Perceptual coding of audio signals using separated irrelevancy reduction and redundancy reduction
JPWO2002080573A1 (ja) * 2001-03-28 2004-07-22 ソニー株式会社 量子化装置、量子化方法、量子化プログラムおよび記録媒体
DE10217297A1 (de) * 2002-04-18 2003-11-06 Fraunhofer Ges Forschung Vorrichtung und Verfahren zum Codieren eines zeitdiskreten Audiosignals und Vorrichtung und Verfahren zum Decodieren von codierten Audiodaten
CA2388352A1 (fr) * 2002-05-31 2003-11-30 Voiceage Corporation Methode et dispositif pour l'amelioration selective en frequence de la hauteur de la parole synthetisee

Also Published As

Publication number Publication date
JP2007522509A (ja) 2007-08-09
BRPI0506627B1 (pt) 2018-10-09
AU2005213767A1 (en) 2005-08-25
JP4444295B2 (ja) 2010-03-31
CA2555639C (fr) 2012-07-10
DE502005001821D1 (de) 2007-12-13
NO337836B1 (no) 2016-06-27
DE102004007184B3 (de) 2005-09-22
EP1697929A1 (fr) 2006-09-06
IL177164A0 (en) 2006-12-10
WO2005078703A1 (fr) 2005-08-25
AU2005213767B2 (en) 2008-04-10
RU2006132742A (ru) 2008-03-20
ES2294685T3 (es) 2008-04-01
RU2337413C2 (ru) 2008-10-27
US7464027B2 (en) 2008-12-09
US20070043557A1 (en) 2007-02-22
CN1918630B (zh) 2010-04-14
IL177164A (en) 2010-11-30
CN1918630A (zh) 2007-02-21
KR100813193B1 (ko) 2008-03-13
BRPI0506627A (pt) 2007-05-02
HK1093814A1 (en) 2007-03-09
CA2555639A1 (fr) 2005-08-25
NO20064091L (no) 2006-11-10
ATE377243T1 (de) 2007-11-15
KR20060113999A (ko) 2006-11-03

Similar Documents

Publication Publication Date Title
EP1687808B1 (fr) Codage audio
EP1697929B1 (fr) Procede et dispositif pour quantifier un signal de donnees
EP1697928B1 (fr) Codage audio
EP1687810B1 (fr) Dispositif et procede pour determiner un pas de quantification
DE102007017254B4 (de) Vorrichtung zum Kodieren und Dekodieren
DE60214599T2 (de) Skalierbare audiokodierung
DE60307252T2 (de) Einrichtungen, verfahren und programme zur kodierung und dekodierung
EP0954909B1 (fr) Procede de codage d'un signal audio
EP1145227B1 (fr) Procede et dispositif pour masquer une erreur dans un signal audio code, et procede et dispositif de decodage d'un signal audio code
EP1025646B1 (fr) Procede et dispositif de codage de signaux audio ainsi que procede et dispositif de decodage d'un train de bits
DE60313332T2 (de) Audiocodierung mit niedriger bitrate
WO2005083678A1 (fr) Dispositif et procede pour traiter un signal multicanal
EP3544003B1 (fr) Dispositif et procédé de détermination d'une valeur d'évaluation
DE19959156C2 (de) Verfahren und Vorrichtung zum Verarbeiten eines zu codierenden Stereoaudiosignals
WO2007131564A1 (fr) Codage de signaux d'information
DE102012009435A1 (de) Vorrichtung und Verfahren zum Streaming hybridcodierter Audiodaten
DE202004002269U1 (de) Vorrichtung zum Quantisieren eines Informationssignals
DE202004002264U1 (de) Audiocodierung
EP0905918A2 (fr) Procédé et dispositif de codage de signaux audio
DE10065363B4 (de) Vorrichtung und Verfahren zum Decodieren eines codierten Datensignals
DE10252327A1 (de) Verfahren zur Erweiterung der Bandbreite eines schmalbandig gefilterten Sprachsignals, insbesondere eines von einem Telekommunikationsgerät gesendeten Sprachsignals
DE19515805A1 (de) Verfahren zur Datenreduktion eines digitalisierten Tonsignals
DE102006053976A1 (de) Verfahren zur Umschaltung von Fenster und zur Entscheidung des Bandstatus der Mittel/Seitenkodierung für fortgeschrittene Audiokodierung

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20060720

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU MC NL PL PT RO SE SI SK TR

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: FRAUNHOFER-GESELLSCHAFT ZUR FOERDERUNG DER ANGEWAN

REG Reference to a national code

Ref country code: HK

Ref legal event code: DE

Ref document number: 1093814

Country of ref document: HK

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

DAX Request for extension of the european patent (deleted)
GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU MC NL PL PT RO SE SI SK TR

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

Free format text: NOT ENGLISH

REG Reference to a national code

Ref country code: IE

Ref legal event code: FG4D

Free format text: LANGUAGE OF EP DOCUMENT: GERMAN

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

REF Corresponds to:

Ref document number: 502005001821

Country of ref document: DE

Date of ref document: 20071213

Kind code of ref document: P

GBT Gb: translation of ep patent filed (gb section 77(6)(a)/1977)

Effective date: 20071129

REG Reference to a national code

Ref country code: HK

Ref legal event code: GR

Ref document number: 1093814

Country of ref document: HK

REG Reference to a national code

Ref country code: SE

Ref legal event code: TRGR

REG Reference to a national code

Ref country code: ES

Ref legal event code: FG2A

Ref document number: 2294685

Country of ref document: ES

Kind code of ref document: T3

ET Fr: translation filed
PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

Ref country code: PT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20080331

Ref country code: PL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

Ref country code: LT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

Ref country code: IS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20080229

Ref country code: BG

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20080131

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

Ref country code: CZ

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: RO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

Ref country code: SK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20080801

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: EE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

Ref country code: GR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20080201

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: CY

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: HU

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20080501

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: TR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20071031

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 12

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 13

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20230220

Year of fee payment: 19

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: SE

Payment date: 20230220

Year of fee payment: 19

Ref country code: IT

Payment date: 20230228

Year of fee payment: 19

Ref country code: BE

Payment date: 20230220

Year of fee payment: 19

P01 Opt-out of the competence of the unified patent court (upc) registered

Effective date: 20230512

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: LU

Payment date: 20240220

Year of fee payment: 20

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: ES

Payment date: 20240319

Year of fee payment: 20

Ref country code: NL

Payment date: 20240220

Year of fee payment: 20

Ref country code: IE

Payment date: 20240216

Year of fee payment: 20

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: AT

Payment date: 20240216

Year of fee payment: 20

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: MC

Payment date: 20240219

Year of fee payment: 20

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FI

Payment date: 20240219

Year of fee payment: 20

Ref country code: DE

Payment date: 20240216

Year of fee payment: 20

Ref country code: CH

Payment date: 20240301

Year of fee payment: 20

Ref country code: GB

Payment date: 20240222

Year of fee payment: 20