JP4988901B2 - Recording condition adjusting method, optical disc apparatus and information recording method - Google Patents

Description

  The present invention relates to a method for adjusting a recording condition for forming information on an optical disk medium for storing information by forming a recording mark having a physical property different from that of other parts on the recording medium, and an optical disk device and information using the same It relates to a recording method.

  There are many optical disk media such as CD-R / RW, DVD-RAM, DVD ± R / RW, and BD, and such media including two data layers are widely spread. As a corresponding optical disk device, a so-called DVD super multi-drive that supports recording / reproduction of CD-R / RW, DVD-RAM, and DVD ± R / RW is widely used. In the future, it is considered that high-function drives compatible with BDs will become widespread, and the appearance of optical disks with larger capacity is desired.

  With the increase in the speed and density of optical discs, a binarization technique of a reproduction signal by a PRML (Partial Response Maximum Likelihood) reproduction method has become essential. As one of the PRML systems, there is an adaptive PRML system or a compensated PRML system in which a target signal level is adaptively changed according to a reproduction signal. According to Non-Patent Document 1 “Journal of the Institute of Electronics, Information and Communication Engineers CJ90-C, pp. 519 (2007)”, by using such a PRML method, the asymmetry of the reproduction signal and the thermal interference during recording are compensated. It has been shown that a high density equivalent to 35 GB capacity can be realized with a corresponding device. It is shown that the longer the constraint length, the higher the reproduction performance under high density conditions, depending on the PRML constraint length used (bit length representing the class). In order to obtain the best binarization result, an optical equalizer that minimizes the RMS error between the reproduction signal and the PRML target signal is mounted on the optical disc apparatus having such a PRML system. The automatic equalizer is generally implemented as an FIR (Finite Impulse Response) filter with a variable tap coefficient.

  When the recording density of the optical disc is increased, the size of the recording mark is reduced compared to the size of the light spot, and the amplitude of the obtained reproduction signal is also reduced. The resolution of the light spot is determined by the wavelength λ and the numerical aperture NA of the objective lens. When the length of the recording mark having the shortest run length becomes λ / 4 NA or less, the amplitude of the repetitive signal becomes zero. This is a phenomenon generally known as an optical cutoff, and in the BD, λ / 4NA≈119 nm. When the track pitch is constant in BD, the amplitude of the 2T repetitive signal, which is the shortest run length, becomes zero when the capacity of about 31 GB or more is realized. In order to obtain good reproduction performance under such high density conditions, it is essential to use the PRML method.

  In a recordable optical disk, desired information is recorded by changing the crystal state of the recording film or the like using laser light (hereinafter, recording pulse) whose intensity is modulated in a pulse shape. As the recording film, a phase change material, an organic dye, a certain kind of alloy, oxide, or the like is used, and is widely known. In the mark edge code method used in CD, DVD, and BD, code information is determined by the front and rear edge positions.

  FIG. 43 is a diagram showing a recording pulse waveform in “N-1 recording strategy” which is one of recording methods used in BD. In this recording strategy, an NT length mark is recorded using (N-1) pulses. FIG. 43 shows a pulse train for recording 2T to 5T marks among marks having a length of 2T to 9T (T is a channel bit length) used for a BD code. The first maximum power pulse in the pulse train is called the first pulse, and the last maximum power pulse is called the last pulse. A plurality of maximum power pulses between the first pulse and the last pulse is called a multi-pulse, and increases by one for a 4T mark and by one for each subsequent increase in mark length by 1T. The maximum power pulse in the pulse train of the 2T mark is only the first pulse, and the maximum power pulse in the pulse train of the 3T mark is the first pulse and the last pulse, and does not include multipulses. The pulse immediately after the last maximum power pulse in the pulse train is called a cooling pulse.

  There are four types of laser light output power levels: write power PW, space power (erase power) PS, bias power PBW, and cooling power PC. The write power PW is the maximum power level in the pulse train, and is the power level of the first pulse, multipulse, and last pulse. This power level is used to input energy to the recording film to cause a state change. The space power PS is the power level applied to the portion between the marks (space). For write-once discs, the space power PS is mainly used for preheating to form the next mark, and a phase change recording film is used. The rewritable disc is mainly used for direct rewriting by erasing marks and changing them to spaces. The cooling power PC is the power level of the cooling pulse, and is used for the write-once type disc mainly for the purpose of blocking the heat diffusion to the subsequent mark recording part and reducing the thermal interference. Is mainly used for the purpose of forming an amorphous mark by rapid cooling after heating the recording film. Each power level is a uniform value regardless of the mark length.

  Parameters relating to pulse timing include a first pulse start end position dTtop, a first pulse time width Ttop, a multipulse time width TMP, a last pulse time width TLP, and a cooling pulse end position dTS (dTE). Here, dTtop and dTS (dTE) are defined as shown in FIG. 43 based on the NRZI channel bit signal of the recording data. The adjustment unit of these parameters is 1/16 of the channel bit period.

  Among the above pulse parameters, dTtop and Ttop that mainly define the formation conditions of the leading edge of the recording mark, and TLP and dTS (dTE) that mainly define the formation conditions of the trailing edge of the recording mark, improve the quality of the recorded information. Is important to keep. Therefore, in the BD, an adaptive recording pulse that adaptively changes the above parameters according to the length of the recording mark and the length of the preceding or following space is used. The values of dTtop and Ttop are classified and specified by a combination pattern of the length of the recording mark and the length of the space (preceding space) immediately before the recording mark, and the values of TLP and dTS (dTE) are the length of the recording mark. It is specified by classification according to the combination pattern with the length of the space immediately following the mark (following space). Although not shown, TMP is not classified by mark length or space length, and a uniform value is designated for all marks of 4T or more.

  Under the high density conditions as described above, since the recording mark to be formed becomes finer, it is necessary to determine the irradiation condition of the recording pulse (hereinafter referred to as recording condition) with higher accuracy than before. On the other hand, the shape of the light spot of the optical disc apparatus varies depending on the wavelength of the light source, wavefront aberration, focus conditions, disc tilt, and the like. In addition, since the impedance and quantum efficiency of the semiconductor laser change depending on the environmental temperature and changes with time, the shape of the recording pulse also changes. As described above, the adjustment technique for always obtaining the best recording conditions corresponding to the shape of the light spot and the shape of the recording pulse that varies from individual to individual and from environment to environment is generally called test writing. As the recording density increases, the technique for adjusting recording conditions by trial writing increases in importance.

  Recording condition adjustment techniques are roughly classified into two methods. One is a method using a bit error or byte error rate as an index, and the other is a method using a statistical index such as jitter. The former focuses on events that occur with a small probability with respect to recorded data, and the latter focuses on the average quality of the recorded data. For example, considering a write-once optical disc, if data is recorded / reproduced at multiple locations while changing the recording conditions, the former will cause bit errors and byte errors if there is a fingerprint at the recorded location even under the best recording conditions. Since it grows larger, it cannot be selected. Since the best recording conditions should optimize the average quality of the recorded data, in storage systems that cannot avoid the effects of media defects, fingerprints, dust, etc. like optical discs It can be said that the method using a statistical index is excellent.

  As a method for statistically evaluating the quality of recorded data corresponding to the PRML system, Jpn. J. Appl. Phys. Vol. 43, pp. 4850 (2004) (Non-patent Document 2), JP 2003- JP 141823 (Patent Document 1), JP 2005-346897 (Patent Document 2), JP 2005-196964 (Patent Document 3), JP 2004-253114 (Patent Document 4) and JP There is a technique described in, for example, 2003-151219 (Patent Document 5).

  In Patent Document 1, the probability Pa corresponding to the most probable state transition sequence and the probability Pb corresponding to the second most probable state transition sequence are used, and the quality of the reproduced signal is determined by the distribution of | Pa-Pb |. Techniques for evaluation are disclosed. Non-Patent Document 2 describes a Euclidean distance (corresponding to Pa) between a target signal of a binarized bit string (corresponding to the most probable state transition string) obtained from a reproduced signal and a noticed edge of 1 bit. From the absolute value of the difference (corresponding to Pb) of the Euclidean distance between the target signal of the shifted binarized bit string (corresponding to the second most likely state transition string) and the reproduction signal, the Euclidean distance between the two target signals A technique is disclosed in which the subtracted value is defined as MLSE (Maximum Likelihood Sequence Error) and the recording conditions are adjusted so that the average value of the MLSE distribution is zero for each recording pattern.

  Patent Document 2 focuses on edge shift, uses a pattern including a virtual 1T run length as an error pattern in which the edge portion of a reproduction signal shifts to the left and right, and uses a signed sequence error based on the direction of edge shift. A technique is disclosed in which the edge shift amount is obtained by obtaining the difference between the two, and the recording condition is adjusted so that the edge shift amount approaches zero. This evaluation index is called V-SEAT (Virtual state based Sequence Error for Adaptive Target).

  In Patent Document 3 and Patent Document 4, the difference between the Euclidean distance between the reproduction signal and the correct pattern and the incorrect pattern is calculated by using a table in which a combination of the correct pattern and the corresponding incorrect pattern is stored in advance. And a technique for obtaining an estimated bit error rate SbER (Simulated bit Error Rate) obtained from the standard deviation.

  In Patent Document 5, based on the difference between the Euclidean distance between the reproduced signal and the correct pattern and the incorrect pattern, the error probability when the target edge is shifted to the left side and the error probability when the edge is shifted to the right side are respectively obtained. Discloses a technique for adjusting the recording conditions so that the two are equal. Therefore, a predetermined reproduction signal, a first pattern corresponding to the signal waveform pattern of the reproduction signal, and an arbitrary pattern (second or second) other than the first pattern and corresponding to the signal waveform pattern of the reproduction signal 3 pattern) is used. First, a distance difference D = Ee−Eo between the distance Eo between the reproduction signal and the first pattern and the distance Ee between the reproduction signal and an arbitrary pattern is obtained. Next, the distribution of the distance difference D is obtained for a plurality of reproduction signal samples. Next, the quality evaluation parameter (M / σ) of the reproduction signal is determined based on the ratio between the average M of the obtained distance difference D and the standard deviation σ of the distribution of the obtained distance difference D. Then, the quality of the reproduction signal is determined from the evaluation index value (Mgn) represented by the quality evaluation parameter.

JP 2003-141823 A JP 2005-346897 A JP 2005-196964 A JP 2004-253114 A JP 2003-151219 A

IEICE Transactions C Vol. J90-C, pp.519 (2007) Jpn. J. Appl. Phys. Vol. 43, pp. 4850 (2004)

  The most probable state transition sequence and the second most probable state transition sequence described in Patent Document 1, and the correct pattern and the incorrect pattern described in Patent Document 3 are targets for measuring the distance from the reproduction signal. It is the same in the sense of a bit string. In Patent Document 2 and Patent Document 5, there are three target bit strings, which have the same meaning. Hereinafter, these will be collectively referred to as a review bit string. Further, since the present invention aims to increase the capacity of 30 GB or more on the basis of the BD system, the description will be made on the premise of the shortest run length 2T of the modulation code.

  As described in Non-Patent Document 1, a PRML system with a constraint length of 5 or more is suitable for realizing high-density recording. As described above, under the BD optical system conditions (wavelength 405 nm, objective lens numerical aperture 0.85), when the recording density is increased in the linear direction, the capacity is about 31 GB or more and the amplitude of the 2T repetitive signal becomes zero. At this time, it is well known that a PR (1, 2, 2, 2, 1) method in which the target amplitude of the 2T repetition signal is zero is suitable as the PRML method. As a method for evaluating the quality of a reproduction signal corresponding to the PR (1, 2, 2, 2, 1) system, there are SbER disclosed in Patent Document 3 and Patent Document 4. SbER is the second most probable evaluation bit string (erroneous pattern) in addition to the binarized bit string (positive pattern). The Hamming distance from the positive pattern is 1 (edge shift), the Hamming distance is 2 (2T data shift), Using a Hamming distance of 3 (shift of 2T-2T data), each distribution is regarded as a Gaussian distribution, and the bit error rate is estimated using an error function from the average value and standard deviation.

  The performance required for a high-precision recording condition adjustment technique necessary for realizing an optical disc system having a recording capacity of 30 GB or more based on the BD standard will be described. To this end, at least the quality of the data recorded based on the adjustment result, (1) the SbER or the like, the bit error rate, etc. are sufficiently small, and (2) the quality of the data recorded by one drive device However, it is also required that the SbER and the bit error rate are sufficiently small. The required performance (1) is a matter of course, but the required performance (2) is characteristically required in an optical disk system in which the disk medium is replaceable. It can be said that a recording condition adjusting method that does not satisfy at least two required performances is not suitable for a high-density optical disk system.

  From the viewpoint of the above two required performances, the problems of the technology inferred from the conventional technology and the combination thereof will be described.

  First, in BD, various events that occur when recording / reproduction is performed at a recording density equivalent to 30 GB / surface or higher will be described using experiments and simulation results in which the linear recording density is increased.

FIG. 2 shows the experimental results summarizing the relationship between the recording power and the number of bit errors measured using a prototype three-layer write once optical disc sample. The recording material used for the prototype disk is a Ge-based compound thin film, the layer spacing of each layer is 14 μm and 18 μm, the three-layer structure, and the thickness of the transparent cover layer from the optical head to the innermost layer is 100 μm. did. The track pitch is 320 nm. The recording / reproducing conditions were such that the data transfer rate was twice that of BD, the detection window width 1T was about 56 nm, and the recording density was equivalent to 33 GB. As the recording pulse, a general multi-pulse type recording pulse modulated between three power levels (peak power, assist power, and bottom power) was used. As the configuration of the reproduction signal processing system, an 8-bit A / D converter, a 21-tap automatic equalizer, and a PR (1, 2, 2, 2, 1) Viterbi decoder were used. The minimum value of the bit error rate was 10 ⁻⁵ or less for each layer. The peak power values that minimize the bit error rate were 13.5 mW, 15.5 mW, and 11.5 mW in the L0, L1, and L2 layers, respectively. The figure summarizes bit errors when the recording power is changed with the ratio of the three powers being constant in the L0 layer. In addition to the edge shift, one to four consecutive 2Ts are shifted together. It is the result of investigating the case of (slip). As seen in the figure, it can be seen that not only the edge shift but also the error frequency when the continuous 2T shifts collectively according to the recording power shift is equal to or greater than that. This is because the amplitude of the 2T-2T signal is zero, and in the case of the PR (1, 2, 2, 2, 1) method, the Euclidean distance for the edge shift is 14, whereas the continuous 2T is This is a result due to the fact that the Euclidean distance is as small as 12 when shifting in a collective manner.

  FIG. 3 is a simulation result summarizing the relationship between SNR and SbER. Here, an impulse response obtained when a recording mark is reproduced by a linear diffraction simulator is obtained, and a reproduction signal when recording is ideally performed is calculated by a convolution operation with a recording bit string. Noise was added as white noise, and SNR was determined as the ratio of the half-value amplitude of the 8T repetitive signal to the standard deviation of the noise. The bit error rate, SbER, and the like were calculated by processing this in a reproduction signal processing system using the PR (1, 2, 2, 2, 1) method. Patent Document 3 discloses an evaluation pattern in the case where the number of 2T continuations is up to 2, which is used by extending the number of 2T continuations to 6 (Hamming distance 1 to 7). Since the number of evaluation patterns is 18 per Hamming distance, the total number is 252. As can be seen from the figure, the value of SbER becomes almost constant when the number of continuous 2T is 2 (Hamming distance 3) or more. Although this result seems to contradict the experimental result of FIG. 2, it is not so. In the calculation of SbER, by definition, the bit error rate is estimated in consideration of the existence probability of the evaluation pattern, and therefore the overall bit error rate can be estimated even when the number of consecutive 2Ts is 2 or less.

FIG. 4 shows experimental results showing the relationship between the bit error rate and SbER. Here, in the L0 layer, five continuous tracks were recorded so as to include the influence of crosstalk, and various recording / reproducing stresses were applied to the center track for experiments. Specific stresses include radial tilt (R-tilt), tangential tilt (T-tilt), focus shift (AF), spherical aberration (SA) caused by the operation of the beam expander of the optical head, and changes in recording power. (Pw). For radial tilt, the results for the L2 layer are also shown. As can be seen from the figure, the correlation between the bit error rate and SbER is very good. When the bit error rate is around ^10-5 , the large variation is mainly due to the defect of the prototype medium.

  From the results of these experiments and simulations, under high-density recording / reproducing conditions that achieve a recording capacity of 33 GB / surface, not only edge shift (Hamming distance 1) but also a continuous number 2 (at least 2T) as a bit error ( It can be seen that it is necessary to evaluate errors up to the Hamming distance 3). In particular, in the method of evaluating the quality of the reproduced signal by paying attention only to the edge shift, it cannot be said that the correlation with the bit error rate or SbER is sufficient.

  Next, the distribution of the Euclidean distance difference accompanying the increase in density will be described. The Euclidean distance difference handled here is a value obtained by subtracting the Euclidean distance between the reproduction signal and the correct target signal from the Euclidean distance between the reproduction signal and the erroneous target signal. In Patent Document 1, | Pa−Pb | 3 and 4 are defined as D values. Here, the above-described simulation is used in order to consider an ideal recording state. The distribution of the Euclidean distance difference up to 2 of 2T continuous numbers was obtained by changing the recording density from 25 to 36 GB / plane (T = 74.5 nm to 51.7 nm) with an SNR of 24 dB. The configuration of the reproduction signal processing system is as described above. The results are shown in FIG. This distribution is sometimes called a SAM distribution. As described above, the PR (1, 2, 2, 2, 1) method is different from the ideal Euclidean distance of edge shift = 14, 2T shift, and ideal Euclidean distance = 12 when two consecutive 2T shifts. Therefore, in order to display them together, each Euclidean distance difference is normalized and divided by the ideal Euclidean distance. In the figure, the statistical probability when the distance difference is zero (the left end) or negative corresponds to the bit error rate. As can be seen from the figure, the spread of the distribution increases with the same SNR as the recording density increases. This indicates that the error rate increases corresponding to the improvement of the recording density, which is a reasonable result. On the other hand, focusing on the average value of each distribution (approximately equal to the peak value), in the case of edge shift, it is constant in the vicinity of 1 (= ideal Euclidean distance). However, when continuous 2T shifts, the peak value moves in a direction approaching zero as the number of continuous 2T increases to 1 or 2 and as the recording density improves. I can see it going. The reason for this phenomenon can be considered to depend on the processing capacity of the automatic equalizer. As described above, the automatic equalizer operates so as to minimize the RMS error between the reproduction signal and the positive target signal. On the other hand, since the sampling interval is 1T and is a finite value, discrete frequency characteristic calculation within a range up to ½ of the sampling frequency can be performed by the sampling theorem. As described above, since the filter characteristics obtained by the automatic equalizer are limited, the frequency component of the high frequency of the reproduction signal becomes large in the pattern section in which the continuous number of 2T included in the reproduction signal is large. Since it approaches the limit of the processing capability of the automatic equalizer, it can be considered that the deviation from the ideal Euclidean distance becomes large. As will be described later, the phenomenon that the distribution peak value (or average value) of the Euclidean distance difference shifts to a smaller side than the ideal Euclidean distance due to the improvement of the recording density is a very important matter regarding the recording condition adjustment technique. It should be noted that there is no description regarding this phenomenon in the above-mentioned publicly known documents.

Based on the results of the above experiments and simulations, the problems of the technology inferred from the prior art and the combination thereof are summarized from the viewpoint of the above-described two required performances.
(1) Method described in Non-Patent Document 2 In Non-Patent Document 2, based on the technique described in Patent Document 1, focusing on edge shift, the average value of the Euclidean distance difference distribution is the ideal Euclidean distance. The technique of adjusting is described. In “Formula (1) of Non-Patent Document 2,” the shift amount MD of a specific edge is defined as follows.

Here, X is the level of the reproduced signal, likely state transition to the target signal level and 1-bit edge shift bit string (2nd against P _A and P _B are each binarized bit string (the most likely state transition sequence) Column), and d _min is an ideal Euclidean distance corresponding to the edge shift. Supplementing according to the result shown in FIG. 5, the present method is a method corresponding to adjusting the recording condition so that the distribution of the edge shift becomes the ideal Euclidean distance (= 1). On the other hand, FIG. 3 shows that the correlation with SbER (or bit error rate) is not sufficient when focusing only on edge shift under high-density recording conditions. From this result, it can be seen that the present method focusing only on edge shift is not sufficient in the high density recording condition in view of the above required performance (1). Further, in “Table 2 of Non-Patent Document 2,” a location where two 2Ts continue, that is, the front edge of the 2T mark (Tsfp (2s, 2m)) when the preceding space is 2T, and the subsequent space is 2T. It is shown that there is no adjustment index at the trailing edge (Telp (2s, 2m)) of the 2T mark in this case. Also in this respect, in view of the result shown in FIG. It can be said that this method is not sufficient for the case of density recording conditions.
(2) Method described in Patent Document 2 The method described in Patent Document 2 is also a method of obtaining an index of recording adjustment by paying attention only to edge shift, but by introducing a virtual 1T mark or space, Recording adjustment is also possible at a location where two 2Ts continue. However, similar to the above, since attention is paid only to edge shift, it cannot be said that the correlation with SbER (or bit error rate) is good, so this method may not be sufficient in view of the required performance (1). I understand.
(3) Method described in Patent Document 5 Since the method described in Patent Document 5 selects an error bit string so as to satisfy the run length restriction, not only edge shift but also 2T continuously shifts. Is a method excellent in the correlation between the index and SbER (or bit error rate). In this method, as shown in “FIG. 3 of Patent Document 5”, in order to adjust the recording condition including the 2T mark, evaluation is performed depending on whether the target mark edge is shifted to the left side or the right side. The hamming distance between the erroneous bit string and the positive bit string is different. For example, when Tsfp (3s, 2m) is viewed according to the notation of Non-Patent Document 2, the bit string described is as follows.

  Considering the PR (1, 2, 2, 2, 1) method, in the case of a left shift bit string, the Hamming distance from the positive bit string is 1, and the Euclidean distance is 14. In the case of the right shift bit string, the Hamming distance from the positive bit string is 2, and the Euclidean distance is 12. As can be seen from the results shown in FIG. 5, when the Hamming distance is different, the average and standard deviation values of the respective distributions are different. In Patent Document 5, in order to cope with this problem, the concept of SbER is introduced, each error probability is estimated using an error function, and the condition that both error probabilities are equal is an adjustment target. According to this method, it is considered possible to define a recording condition that minimizes SbER (or bit error rate). On the other hand, as described above, the simulation result shown in FIG. 5 is a result when the recording mark is formed in an ideal state (edge shift = 0). As seen in FIG. 5, the center value and the standard deviation are different depending on the difference in the Hamming distance. Therefore, according to the method described in Patent Document 5, it is necessary to shift the recording mark formation conditions so that the error probabilities of three distributions (probability that the Euclidean distance difference is 0 or less) are equal. In view of the required performance (2) regarding the guarantee of disc compatibility described above, there remains a question as to whether this method is an ideal method for adjusting the recording conditions of a high-density optical disc. In order to quantitatively consider this point, the above-described simulation was used.

  In order to define the edge shift amount detected by the method of Patent Document 5, the concept was extended. According to “Expression (13) of Patent Document 5,” the edge shift equivalent amount Ec is

Is defined. Here, M ₂ , M _3, σ ₂ , and σ ₃ are the average and standard deviation of the distribution of the Euclidean distance difference when the edge of interest is shifted left and right by 1 bit, respectively. As described above, the two distributions were normalized by the ideal Euclidean distance to obtain the result of FIG. Similarly, if it is assumed that the ideal Euclidean distance corresponds to 1T and M ₂ , M _3, σ ₂ , and σ ₃ are each normalized and used with the ideal Euclidean distance, the edge shift Ec ′ in the time axis direction from the edge shift equivalent amount Ec is used. Can be calculated.

  FIG. 6 shows a distribution obtained by simulation, and it is understood that the same result as that schematically shown in “FIG. 6 of Patent Document 5” is obtained. FIG. 7 shows the result of examining the value of Ec ′ when the SNR is changed. As can be seen from the figure, the value of Ec ′ changes greatly according to the change in SNR. In the optical disk device, the shape of the light spot and the SNR of the photoelectric conversion amplifier change for each individual or according to environmental conditions such as temperature. For a storage device such as a hard disk device in which the disk medium cannot be exchanged, it is best to adjust the recording conditions so that the SbER (or bit error rate) is minimized in the drive device. However, it can be said that it is not sufficient that the SbER (or bit error rate) of only the drive is minimized in a storage system with medium exchange like an optical disk. In view of the above-mentioned required performance (2), it can be said that there is room for improvement in this method as a method for adjusting recording conditions under high-density recording conditions.

  Further, it will be described that there is room for improvement in the light of the required performance (1). The bit string used for the evaluation of Tsfp (3s, 2m) is as described above. On the other hand, as described in Patent Document 4, the next evaluation bit string is also used for calculation of SbER.

This is a case where the space following the 2T mark of interest is 2T. As for the left shift bit string, the Hamming distance is 1 and the Euclidean distance is 14 with respect to the positive bit string as described above. On the other hand, in the case of the right shift bit string, the Hamming distance is 3 and the Euclidean distance are 12 with respect to the positive bit string, and the Hamming distance is different from the above. It is desired that the correlation between the recording adjustment evaluation index and the reproduction signal quality evaluation index SbER (or the bit error rate) is sufficiently large depending on the required performance (1). Therefore, also in the evaluation index for recording adjustment, it is necessary that the evaluation bit string conforms to the evaluation index of the reproduction signal quality. In an evaluation index using a target signal in which a target edge is shifted left and right, a left shift is a Hamming distance 1 and a right shift is a Hamming distance 2 and 3 as shown in this example. There was no description about the solution. In this regard, it can be said that there is room for improvement in this method.
(4) Method by Combination of Prior Art In Non-Patent Document 2, based on the technique described in Patent Document 1, paying attention to edge shift, the average value of the distribution of Euclidean distance difference becomes the ideal Euclidean distance. The technique to adjust is described. A method of adjusting this so that the average value of each distribution becomes the ideal Euclidean distance by applying this to the evaluation bit string shown in “FIG. 3 of Patent Document 5” can be easily inferred. However, as shown in FIG. 5, when the recording density is increased, the average value of each distribution is shifted in the direction of decreasing from the ideal Euclidean distance. Similarly, the average value of each distribution changes depending on the SNR. FIG. 8 shows the result of experimental confirmation of this phenomenon. This is an experimental result obtained by performing a reproduction experiment while changing the reproduction power at L0 of the above-described prototype three-layer disc. The horizontal axis in the figure represents the reproduction power of 1.2 mW as 100%. Since the reproduction signal amplitude is proportional to the reproduction power, but the noise (amplifier noise) of the photodetector is constant, this experiment is a result of changing the SNR of the reproduction signal by changing the reproduction power. As can be seen from the figure, the average value of each distribution is smaller than the ideal Euclidean distance (= 1) and decreases as the reproduction power decreases. Also in this method, it is obvious that the difference in SNR depending on the state of the drive device affects the index used for recording adjustment.
(5) Method for minimizing SbER As shown in FIG. 4, SbER shows a good correlation with the bit error rate in the 33 GB / plane experiment. Therefore, a method is conceivable in which, without using an evaluation index for recording adjustment, a condition for performing recording and reproduction for all combinations of recording conditions and obtaining a minimum SbER is selected. However, when the size of a recording adjustment area (trial writing area) is limited as in an optical disk medium, it is practical to search for a condition that minimizes SbER while randomly changing the recording condition. Is impossible. This is because information on the direction for bringing the edge of the mark to be recorded closer to the ideal state cannot be obtained. A method capable of applying test writing to an optical disc apparatus unless the method can quantify the deviation from the target value independently according to each parameter of the recording pulse as in the conventional technique shown above. It will not be. Also, it is desirable to complete the adjustment of the recording conditions in a short time even when the performance of the disk is improved while repeating the trial manufacture of the disk. Also in this sense, a new index for recording adjustment that satisfies the above-mentioned required performances (1) and (2) and can be independently adjusted according to the recording parameters and an adjustment method thereof have been desired. .

  Next, a problem for obtaining good recording quality in a high-density optical disk having a recording capacity of 30 GB or more using a conventional BD recording strategy will be described from the viewpoint of recording pulses.

  In FIG. 44, random data is recorded on the optical disc medium while changing the time width TLP of the last pulse of the pulse train corresponding to the 4T mark, and the value of the edge shift of the trailing edge of the 4T mark is calculated based on the reproduction signal waveform of the portion. It is the result of measurement. Here, the conventional BD N-1 recording strategy was used as the recording strategy, and the TLP adjustment unit was set to 1/16 of the channel bit period. As for TLP, no classification based on the subsequent space length was performed, and a uniform value was designated for all 4T mark recording pulses. In addition, the edge shift of V-SEAT disclosed in Patent Document 2 was used as the edge shift. The conditions of this experiment are the same as those in the experiment using the prototype disk described above.

  According to FIG. 44, when the conventional BD recording strategy is used, the edge shift changes by about 16 points when the last pulse end position is changed by one unit (1/16 of the channel bit period). In order to obtain good recording quality, it is desirable that the edge shift value is within ± 5%. However, with the edge shift change rate when the conventional BD recording strategy is used, the edge shift value is always ± 5%. It is difficult to keep within%.

  When the conventional BD recording pulse waveform is used, the edge shift change rate of the trailing edge of the recording mark is increased because only the time width TLP of the last pulse is a parameter relating to the formation of the trailing edge of the recording mark. This is because changing the last pulse end position also changes the time width of the last pulse. In other words, it is considered that the rate of change in edge shift increases because the thermal energy injected into the trailing edge forming portion of the recording mark changes greatly as the time width of the last pulse changes.

  In a high-density optical disc having a recording capacity of 30 GB or more, the bit length of data recorded on the disc is smaller than that of a conventional BD, so that it is more susceptible to heat when recording front and rear marks. For this reason, a recording pulse used for a high-density optical disc is required to be able to control the edge position of a recording mark with higher accuracy than a conventional BD.

  To deal with this problem, it is conceivable that the recording pulse adjustment unit is a unit smaller than 1/16 of the channel bit period, for example, 1/32 of the channel bit period. In this case, the recording pulse is stored in the control information area of the disc. Since the data amount of the power pulse parameter increases, it is desirable that the trailing edge position of the recording mark can be controlled with high accuracy while the recording pulse adjustment unit remains 1/16 of the channel bit period.

  As described above, regarding the adjustment of the recording condition corresponding to the high density recording condition such that the capacity becomes 30 GB / surface or more based on the BD system, the conventional technology has both the adjustment performance and the guarantee of the medium compatibility. There was a problem that it was not enough. The problem to be solved by the present invention is to provide a new evaluation index and method for recording adjustment that solves these problems, and to provide an optical disk apparatus using the same. Further, the present invention provides a recording method capable of controlling the mark edge position with high accuracy even under the condition that the recording pulse adjustment unit is 1/16 of the channel bit period.

  Since the present invention aims to increase the capacity of 30 GB or more on the basis of the BD system, the following description will be made assuming that the shortest run length of the modulation code is 2T. Further, as described above, SbER that handles 2T continuous numbers up to 2 from the experimental results agrees well with the bit error rate, so that the recording signal adjustment according to the present invention is performed on the premise of SbER as an index for evaluating the reproduction signal quality. The evaluation index will be described. Similar to SbER, if the index for statistically evaluating the playback signal quality based on the Euclidean distance between the target signal and the playback signal, or the index for directly evaluating the bit error rate, etc., the result of adjusting the recording conditions according to the present invention Gives good results.

The above issues can be summarized as follows.
(Problem 1) There is a need for an evaluation index and an adjustment method in which the adjustment target point is constant without depending on the change in SNR for the reproduction compatibility of data recorded based on the adjustment result.
(Problem 2) In order to ensure that the SbER is sufficiently small with respect to the quality of the data recorded based on the adjustment result, the evaluation bit string having at least two consecutive 2Ts matches the evaluation bit string of SbER. Or it is necessary to substantially match.
(Problem 3) Regarding realization of recording adjustment in a short time, it is necessary to provide an evaluation index and an adjustment method that can be independently evaluated in accordance with the recording pulse condition or each parameter of the adaptive recording pulse.
(Problem 4) Regarding the high accuracy of the trailing edge position control of the mark, it is necessary to control the trailing edge position of the mark with high precision even under the condition that the recording pulse adjustment unit is 1/16 of the channel bit period.

  The basic concept constituting the problem solving means of the present invention is to separate the component corresponding to the shift of the target edge and the component depending on the SNR in the evaluation index according to the difference in Euclidean distance between the two target signals and the reproduced signal. To evaluate it. In order to facilitate understanding of the present invention, the definition of an evaluation index that satisfies these problems is shown first, and then that the problems are satisfied.

  Hereinafter, the reproduced signal is W, the target signal of the binarized bit string obtained from the reproduced signal is T, the target edge of the binarized bit string is shifted to the left by one bit, and the target signal of the bit string that satisfies the run length restriction is L, the target edge of the binarized bit string is shifted to the right by 1 bit, and the target signal of the bit string that satisfies the run length restriction is R. The Euclidean distance between W, T, R, and L is expressed as ED (W, T) and ED (W, R). The evaluation value for an error in which the edge of interest shifts to the left is xL, and the evaluation value for an error in which the edge is shifted to the right is xR.

  The edge shift amount of the edge of interest is called an extended edge shift D and is defined below.

  The correction amount corresponding to the error probability of the edge of interest is called SNR factor S and is defined below.

  The edge shift amount used for recording adjustment is defined as the statistical average value Δ of D for the group of edges recorded under the same recording pulse conditions, that is, the edge of interest, the mark length, and the preceding (or subsequent) space length are the same. It is defined below.

Where N is the total number of measured edges and D _n is the extended edge shift of the nth edge.

  Further, a jitter value corresponding to the error probability of the edge of interest is defined as σ and is defined below.

Where S _n is the SNR factor of the nth edge.

The following metrics of the present invention as defined by the equation (D1) (Formula D6), referred to as L-SEAT (run-length- L imited S equence E rror for A daptive T arget), defined in (Equation D5) Δ is called L-SEAT shift, and σ defined in (Equation D6) is called L-SEAT jitter. According to the recording condition adjusting method of the present invention, recording / reproduction is performed while changing the condition of the recording pulse, and the recording pulse is such that the absolute value of the L-SEAT shift and the value of the L-SEAT jitter are minimized with respect to the corresponding edge. The condition is to select.

  Hereinafter, it will be described that the recording condition adjusting method of the present invention satisfies the above (Problem 1) to (Problem 3). As described in Patent Documents 1 to 5, in the PRML method, the error margin is represented by a Euclidean distance difference. Hereinafter, for the sake of simplicity, dEDL and dEDR are defined as follows as values obtained by normalizing the Euclidean distance difference related to an error in which the target edge shifts to the left and right with the ideal Euclidean distance, respectively.

(Problem 1) Reproduction compatibility of data recorded based on the adjustment result As described above, the evaluation index for recording adjustment needs to have a constant edge shift evaluation index without depending on the change in SNR. The average value of the distribution of the Euclidean distance difference varies depending on the SNR. Since W, T, L, and R are signal levels for a plurality of times t (t = t ₀ +1, t ₀ +2, t ₀ +3, t ₀ +4, t ₀ +5), these are used as coordinates in a multidimensional space. I'll think about it. For simplicity, when considering the right shift error of Hamming distance 1, in the PR ( ₁ , ₂ , ₂ , ₂ , ₁ ) system, T (T ₁ , T ₂ , T ₃ , T ₄ , T ₅ ), W ( T ₁ + δ ₁ , T ₂ + δ ₂ , T ₃ + δ ₃ , T ₄ + δ ₄ , T ₅ + δ ₅ ), R (T ₁ +1, T ₂ +2, T ₃ +2, T ₄ +2, T ₅ +1) be able to. Further, considering a coordinate system in which the origin of this five-dimensional space is T, if W and R position vectors (= coordinates) are again W and R, then W (δ ₁ , δ ₂ , δ ₃ , δ ₄ , δ ₅ ), R (1, 2, 2, 2, 1). These relationships in a plane including T, L, and R are schematically shown in FIG. In the figure, the x-axis is taken in the direction of the line segment TR, and is normalized so that the point R is 1. It should be noted that since the y-axis is perpendicular to the x-axis, the y-axis changes depending on the value of W and does not indicate a certain direction. The Euclidean distance related to W, T, and R has the following relationship.

  That is, the sum of the Euclidean distance from T to W and the Euclidean distance from R to W does not necessarily match the Euclidean distance from T to R.

  FIG. 9 schematically shows the measurement of the edge shift of a physically recorded mark. In this case, when the distance from the target value T (origin) to the edge of the recording mark measured is x, the distance from the target value R shifted to the right by 1T to the edge of the recording mark is (1−x). Is always 1 (= 1T, T is the detection window width). The edge control by the recording pulse is generally a shift control in the time direction, and is based on the concept of linear measurement regarding the edge shift of the physically recorded mark.

  Therefore, in the definition of the Euclidean distance (the square of the length of the line segment) in PRML, if the mapping component on the x-axis of the vector TW is xR, the mapping component on the x-axis of the vector RW is ( 1-xR), and the sum of the two can be 1. xR can be calculated as follows using the Euclidean distance between T, R, and W as the inner product value of the vector TR and the vector TW.

  This is the meaning of the equivalent edge shift xR defined in (Formula D2). The equivalent edge shift can be calculated in the same manner when the Hamming distance is 2 or 3. The second term of (Expression 6) is obtained by normalizing the Euclidean distance difference shown in FIG. 5 with the ideal Euclidean distance. xR is a component value of W in the TR direction and at the same time includes a term relating to the error probability of PRML. By natural expansion, the equivalent edge shift xL can be calculated by (Equation D1) using the target L shifted to the left.

  On the other hand, since the coordinates of W change according to the SNR value, the value of the equivalent edge shift changes for each edge to be measured. However, as described above, since the equivalent edge shift is linearly added on the TR line segment, it is possible to evaluate the edge shift of the recording mark without depending on the SNR by obtaining the average value.

Next, a method for dealing with a problem in which the average value of the Euclidean distance difference changes according to the SNR will be described. As described above, the cause of this phenomenon is considered to be caused by the frequency characteristic of the filter obtained by the automatic equalizer being limited by the sampling theorem. Therefore, the change in average value that appears when the edge of interest shifts to the left and right is the same. As shown in FIG. 5, the change in the average value of the distribution can be inferred from the fact that it can be classified according to the number of continuous 2Ts, that is, Hamming distance. If the average values of the normalized Euclidean distance differences dEDL and dEDR are M _L and M _R , the deviation from these ideal Euclidean distances is dM, and the edge shift amount to be measured is Δ ₂ , the following relationship Is established.

  On the other hand, V-SEAT disclosed in Patent Document 2 discloses a technique for calculating a normalization sequence error by focusing only on edge shift (Hamming distance 1), adding a code corresponding to left and right shift, and performing addition averaging. is doing. For example, it is natural that the equivalent edge shift in the right direction is positive and the transparent edge shift in the left direction is negative. Applying this, calculating the left and right equivalent edge shifts for the target edge, and adding and averaging the codes corresponding to the shift direction as the evaluation value, the distribution of the Euclidean distance difference depending on the SNR The change dM in the average value can be canceled out.

Similarly, it can be seen that, as a measurement value for one edge of interest, the extended edge shift D defined by (Expression D3) is an evaluation value of the edge shift from which the influence depending on the SNR is removed. The L-SEAT edge shift Δ defined by (Expression D5) is statistically equivalent to the difference Δ ₂ between the average values of the respective distributions defined by (Expression 9).

FIG. 10 is a schematic diagram showing equivalent edge shifts xL and xR with L and R as target signals. In the figure, L and R for the six dimensions of t (t = t ₀ , t ₀ +1, t ₀ +2, t ₀ +3, t ₀ +4, t ₀ +5) are taken into account in consideration of the shift of L and R for one time. , W coordinates are represented with T as the origin. The edge shift x of the physically recorded mark is x = {(1 using the distance from the point R on the right side of 1T (= 1-x) and the distance from the point L on the left side of 1T (= 1 + x). -X) + (1 + x)} / 2. (Equation 9) means this and measurement. On the other hand, regarding the Euclidean distance between W, T, L, and R, since there is a time lag between L and R, the line segment TR and the line segment TL do not ride on the geometric line. The angle θ between the two is obtained by using the inner product of two vectors, and when both are errors of edge shift (Hamming distance 1) as shown in the figure, cos θ is as follows.

Here, vector (T, L) and vector (T, R) represent the position vectors of L and R, respectively, and the “·” operator represents the inner product. If T is the most probable target signal and L and T are the second most probable (highest error probability) target signals, the extended edge shift D is set to zero from the viewpoint of the error rate in the PRML system. Thus, it can be said that it is reasonable to adjust the recording conditions. It can be said that the fact that the two target signals do not lie on a geometric line is a feature of edge shift measurement in the PRML system. FIG. 11 summarizes the relationship between L, R, and cos θ when the number of 2T repetitions is 2 (Hamming distance 1, 2, 3). In the figure, when the hamming distance of L is 1 and the hamming distance of R is 3, cos θ> 0 and the angle between L and R is geometrically less than 90 degrees, but L and R are the most error. If a target signal with high probability is selected, the shift of the edge of interest can be measured by the average value Δ of the extended edge shift D or the difference Δ ₂ between the average values of the L and R distributions.

  FIG. 12 is a simulation result showing the relationship between dEDL and dEDR. The simulation conditions are as described above, and are the results when a recording mark having a predetermined length is ideally recorded at a recording density corresponding to 33 GB / surface. Here, the SNR was 20 dB. In the figure, 4 of (a) Tsfp (2s, 2m), (b) Tsfp (2s, 3m), (c) Tsfp (3s, 2m), and (d) Tsfp (3s, 3m) for the front edge of the recording mark. Results for 1000 edges are shown for one case. Here, as the L and R target signals, the Hamming distances are (a) (2, 2), (b) (2, 1) and (3, 1), (c) (1, 2) and (1 , 3) and (d) (1, 1) were used. The broken lines in the figure indicate dEDL + dEDR = 2, that is, a relationship equivalent to the measured value storage relationship regarding the physical recording mark shown in FIG. As can be seen in the figure, each plot point shows a correlation substantially along the broken line, and it can be seen that the fluctuation of the reproduction signal due to the influence of noise is generally symmetric with respect to the left shift and the right shift. In detail, as shown in FIGS. 12B and 12C, when the hamming distances of the left and right target signals are not equal, it can be seen that the distribution has a slightly different slope from that of the broken line. This corresponds to the fact that the probability of error occurrence by the PRML method differs depending on the left and right shifts, and is the difference between the measurement related to the physical recording mark and the measurement along the error margin of the PRML method. The evaluation of edge shift by V-SEAT disclosed in Patent Document 2 uses only a target signal having a Hamming distance of 1, and therefore, the broken line is also along the broken line in the cases of FIGS. 12B and 12C. Only the measurement of the relationship can be performed. Therefore, the present invention can detect the difference in error rate according to the left and right edge shifts more accurately than in the prior art. This is the first effect of the present invention.

  FIG. 13 is a simulation result showing the relationship between the average value of dEDL and dEDR and the extended edge shift D. The simulation conditions are the same as in FIG. Again, each of the four cases of (a) Tsfp (2s, 2m), (b) Tsfp (2s, 3m), (c) Tsfp (3s, 2m), and (d) Tsfp (3s, 3m) Edge results are shown. In the figure, the distribution of the average values of dEDL and dEDR is (1) the spread is greatly different for each edge pattern, and (2) the ideal Euclidean distance difference is shifted to a side smaller than 1 as a result of FIG. Is reflected. In contrast, the distribution of the extended edge shift D does not depend on the edge pattern, (1) the spread of the distribution is almost uniform, and (2) the center of the distribution is almost zero. In the figure, these differences are represented by a schematic distribution shape. The two effects obtained by the introduction of the extended edge shift D are (1) that the shift of the reproduction signal is calculated as an inner product value with the vector TR or the vector TL by the equivalent edge shift, and (2 This is because the left and right equivalent edge shifts are averaged by adding a sign.

  A summary of the effects of the present invention is shown in FIG. This figure is obtained by adding the average value Δ of the extended edge shift defined in (Equation D5) to the relationship between Ec ′ (the method of Patent Document 5) and SNR shown in FIG. Since an optical disk device is a device that handles portable media, it is necessary to be able to record on an unspecified number of media under similar recording conditions. However, the SNR of the information actually recorded varies depending on the operating environment of the optical disc apparatus and the machine condition. In the case of the conventional method, as shown in FIG. 14, the Ec ′ value changes greatly according to the change of the SNR, which is an ideal recording signal and a recording signal that is actually recorded on the recording medium. (In FIG. 14, evaluation is performed using the edge shift amount as an index) means that the difference varies depending on the operation condition (SNR) of the optical disc apparatus. On the other hand, the Δ value of the present invention does not depend on the change in SNR and is almost zero and constant, which means that recording can be performed under the same conditions even if the operation condition of the optical disk fluctuates. As described above, in this simulation, random noise is added to a signal when a recording mark of a predetermined length is ideally recorded. With this condition, the edge shift evaluation value Δ is almost equal. The measurement result of zero is very excellent from the viewpoint of reproduction compatibility of recorded data. This is the second effect of the present invention.

(Problem 2) As a result of adjusting the recording condition according to the present invention for the quality of data recorded based on the adjustment result, SbER needs to be sufficiently small. In order to realize this, it is necessary that dEDL and dEDR are minimized by adjusting the recording pulse, and that the T, L, and R evaluation bit strings and the SbER evaluation bit string are substantially equivalent.

First, the former will be explained. As described above, the target signals T, L, and R do not ride on the geometric line due to the difference in Hamming distance and the time shift. As a result, the absolute value of the equivalent edge shift differs for the left and right shifts. This is a feature of the edge shift measurement according to the present invention. Now, when evaluated of N edges, the value of n-th edge of dEDL and dEDR as dEDL _n and dEDR _n, when the recent average of 1 and a standard deviation sigma _L, sigma _R is the following formula It is represented by

The bit error rate is evaluated by these combined standard deviations σ _LR . Therefore,

  Thus, the right side is a value twice as large as the L-SEAT jitter shown in (Equation D6). The coefficient 2 shown here is not essential. In the distribution of dEDL and dEDR, the error margin is 1 (ideal Euclidean distance = 1), whereas in L-SEAT, the jitter caused by a conventional time interval analyzer is used. Similar to the measurement, this is a coefficient generated when the error margin is ± 1 / 2T. The error rate values by the error function when both are Gaussian distributions are equal. It can be seen that the L-SEAT jitter represents the combined standard deviation when the Euclidean distance difference distributions shown in FIG. Therefore, it can be said that L-SEAT jitter is an evaluation index having a good correlation with SbER and bit error rate. More specifically, as can be seen from (Equation D4), the SNR factor by definition has the same value as the amount by which the distribution of the Euclidean distance difference deviates from the ideal value 1 depending on the SNR and recording density. Therefore, the contribution of the SNR factor in the L-SEAT jitter defined in (Equation D6) takes into account the deviation of the average value of the Euclidean distance difference distribution. As described above, in the L-SEAT of the present invention, it is possible to perform evaluation by separating the component corresponding to the shift of the edge of interest (extended edge shift) and the component dependent on SNR (SNR factor). As a result, it is possible to simultaneously provide two functions such as shift adjustment excellent in reproduction compatibility performance independent of the SNR of individual drive devices and guaranteeing the minimum conditions of SbER and bit error rate. In other words, the L-SEAT of the present invention is a conventional reproduction signal quality evaluation index such as jitter and SbER compared to the conventional signal adjustment index for recording adjustment including V-SEAT disclosed in Patent Document 2. It is an excellent signal evaluation index. Therefore, the present invention provides an unprecedented superior signal evaluation index, which is the third effect obtained by the present invention. Experimental verification on this point will be described later along with experimental results.

Next, the affinity with an evaluation bit string used for evaluating the quality of a reproduced signal such as SbER will be described. The reproduction signal evaluation techniques described in Patent Documents 1, 3, 4, etc. have different configurations, but as a common technique, the most probable first evaluation bit string from among the binary bit strings output from the PRML decoder The process of searching and extracting is included. The length M of the evaluation bit string can be generalized to M = 2N-1 + 2N _2T using the constraint length N of the PRML method and the continuous number N _2T of 2T patterns included in the evaluation bit string. Where N _2T is 0, 1, 2,. . . Is an integer. N _2T = 0, 1, and 2 correspond to edge shift, 2T shift, and 2T ball shift, respectively, according to the above-described notation. When N _2T is 0, 1, ₂ , 3, 4, 5, and 6, the Hamming distances are 1, 2, 3, 4, 5, 6, and 7, and the evaluation bit strings of pattern A and pattern B The Hamming distance between is (N _2T +1). The evaluation bit string is the relationship between the most probable first evaluation bit string and the second evaluation bit string corresponding to the target signal that minimizes the Euclidean distance from the target signal of the first evaluation bit string from among ^2M bit strings. Can be easily enumerated by the mechanical operation of extracting.

  FIG. 15 shows an example of an evaluation bit string corresponding to the PR (1, 2, 2, 2, 1) system with a constraint length of 5, and the same is described in Patent Document 4. As shown in the figure, when the PRML system with a constraint length of 5 is used and the evaluation bit string is searched for and extracted from the binarized bit string of the PRML decoder and the quality evaluation of the reproduced signal is performed, there are 18 sets for each Hamming distance. A total of 54 sets, that is, 108 evaluation bit strings are listed. When evaluating the reproduction signal, it is necessary to perform the search / extraction processing of these evaluation bit strings in parallel.

  FIG. 16 shows the common terms extracted from the evaluation bit string corresponding to PR (1, 2, 2, 2, 1) having a constraint length of 5 shown in FIG. As seen in the figure, the 108 evaluation bit strings corresponding to the Hamming distances 1, 2, and 3 are a main bit string having bit lengths 5, 7, and 9, respectively, and a 2-bit sub-bit string XX, YY can be expressed. Here, when the Hamming distance is 1, the main bit string is “00011”, “00111”, “11100”, and “11000”, and when the Hamming distance is 2, “0001100”, “0011000”, “1110011”. , And “1100111”, and when the Hamming distance is 3, “000110011”, “001100111”, “111001100”, and “110011000”, and the sub-bit string AA is “00”, “10”, or “11”, and the sub-bit string BB is “00”, “01”, or “11”. The section of the main bit string defined here is a section for calculating the Euclidean distance between the target signal and the reproduction signal. The sub-bit string is necessary only for calculating the target signal level at the end of the main bit string, and is not related to the Euclidean distance between the plurality of target signals. In this sense, the sub-bit string can be considered as defining a boundary condition for determining the level of the end of the target signal.

The main bit string is determined without depending on the constraint length of the PRML system. The reason for this will be explained. When the shortest run length m is 2T, in order to express that one bit changes due to edge shift, the shortest length of the bit string is a value obtained by doubling the shortest run length and adding 1, that is, 2m + 1 = 5 bits. This is the main bit string. Similarly, when generalized using the number N _2T of consecutive 2Ts included in the evaluation bit string, the length of the main bit string is (2m + 1 + 2N _2T ). Thus, the main bit string means the shortest bit string that is determined according to the number of consecutive 2Ts included in the evaluation bit string. On the other hand, as described above, the length of the bit string necessary for calculating the Euclidean distance from the reproduction signal is (2N-1 + 2N _2T ) using the constraint length N of the PRML method. The difference between the lengths of both bit strings is (2N-1 + 2N _2T ) − (2m + 1 + 2N _2T ) = 2 (Nm−1), and it can be seen that this is always an even number. When the shortest run length m = 2, this value is 2 (N−3).

  As described above, the evaluation bit string can be organized and expressed by using the main bit string that does not depend on the PRML constraint length N and the sub-bit string of the length (N-3) added to both ends of the main bit string. Is possible.

  Thus, by arranging and describing the evaluation bit string, the relationship between the reproduction signal quality evaluation index and the present invention can be simplified, and the circuit scale can also be reduced in the present invention. Become.

  In FIG. 16, the evaluation bit string is described as a set of A and B in accordance with the contents described in Patent Document 4. A first evaluation bit string (an evaluation bit string corresponding to the target signal T) is searched from the bit string obtained by binarizing the reproduction signal, and the second most likely second evaluation bit string (target signal L or R) is searched. It is advantageous to reduce the circuit scale by generating and using an evaluation bit string corresponding to Since the Hamming distance between the first and second evaluation bit strings is predetermined, a bit string having the same number of “1” as the Hamming distance is used as a generated bit string, and the first evaluation bit string (XOR) operation is performed by an exclusive OR (XOR) operation. By applying to T), the second evaluation bit string can be generated. FIG. 17 summarizes the main bit strings corresponding to the Hamming distances from 1 to 7. In the figure, the main bit string shown above is listed in the main bit string column. Here, the main bit string number of the combination of the Hamming distance and the number from 1 to 4 is determined and arranged in the main bit string. As shown in the figure, the operation for generating the second main bit string can be obtained by performing an XOR operation on the generated bit string determined for each Hamming distance. The main bit string number of the second main bit string is also shown.

  As described above, considering the main bit string, it is possible to explain the affinity between the SbER evaluation bit string and the evaluation bit string according to the method of the present invention.

  First, FIG. 18 shows a table listing the main bit strings of edge evaluation according to the present invention when the number of consecutive 2Ts is 2 or less. When L and S targets are generated simultaneously by L-SEAT, the length of the main bit string is 1T longer than the above, and is 6, 7, and 8 for the Hamming distances 1, 2, and 3, respectively. . Here, as in FIG. 17, the main bit string included in the bit string obtained by binarizing the reproduction signal and the generated bit string for generating the main bits for L and R by performing XOR operation on this bit string are listed. did. The total number of main bit strings is 12, and the underlined bits in each main bit string represent the edge of interest. The selection rule for the main bit string and L, R adopted here is such that the edge of interest is shifted by 1 bit to the left and right, the run length limit is satisfied, and the Hamming distance is minimized (bit inversion number is minimized). The target is selected as L and R. Also, the record mark is described as “1” and the space as “0”. In the case of a so-called High To Low type medium in which the amount of reflected light from the recording mark is small compared to the space, if the PR class is (1, 2, 2, 2, 1), the recording mark is set to “0”, It is only necessary to invert “1” and “0” of the main bit string so that the space is “1”. Alternatively, if the PR class is set to (-1, -2, -2, -2, -1) and the direction of the impulse response is reversed, FIG. 18 can be used as it is. Hereinafter, in the description of the present invention, unless otherwise specified, a recording mark is treated as “1” and a space as “0”.

Hereinafter, when the maximum value of N _2T is 2, the relationship between the SbER calculation main bit string shown in FIG. 17 and the L-SEAT evaluation main bit string shown in FIG. 18 will be described. FIG. 19 is a comparison between N _2T = 0, that is, when the Hamming distance is 1. This is an evaluation regarding the front edge of a mark of 3T or more. For SbER, an evaluation time t is added, and for L-SEAT, an edge type is added. As can be seen from the figure, there is one edge included in the main bitstream. Both SbER and L-SEAT evaluate two Hamming distances per edge, and the main bit strings are the same. That is, both evaluation bit strings including the sub-bit string match. In the figure, only the front edge is shown, but if “1” and “0” are inverted, they can be treated as the rear edge, and even in this case, the evaluation bit string matches clearly.

FIG. 20 is a comparison between N _2T = 1, that is, when the Hamming distance is 2. The main bit string includes two edges. Both SbER and L-SEAT evaluate two Hamming distances per edge, and the main bit strings are the same. Looking at the transition of evaluation with respect to time, as shown by the dashed arrows in the figure, in the case of SbER, evaluation is performed in the order of L, L, R, and R, and L-SEAT is evaluated in the order of L, R, L, and R. I know what to do. Similarly, the evaluation bit strings coincide with each other for patterns in which “1” and “0” included in the main bit string are inverted.

FIG. 21 shows a comparison between N _2T = 2 and the case where the Hamming distance is 3. There are three edges included in the main bit string. Similar to the above example, both SbER and L-SEAT are evaluated for two hamming distances per edge, and the main bit strings are the same. Looking at the transition of evaluation with respect to time, as shown by broken line arrows in the figure, in the case of SbER, evaluation is performed in the order of L, L, L, R, R, R, and L-SEAT is L, R, L, R , L and R in this order. Similarly, the evaluation bit strings coincide with each other for patterns in which “1” and “0” included in the main bit string are inverted.

From the above examination, it was found that when N _2T is 2 or less, the evaluation bit string for SbER calculation matches the evaluation main bit string shown in FIG. Similarly, when N _2T is 3 or more, the run length restriction is satisfied for the main bit string up to the Hamming distance where the maximum value is the same as SbER as the main bit string for L-SEAT calculation, By selecting the main bit string that minimizes the Hamming distance as the L and R main bit strings, the evaluation main bit strings can be matched in the same manner. A specific example when N _2T is 3 will be described later. As shown in FIG. 5, the basic concept of L-SEAT is a target in which the edge to be evaluated focusing on symmetry is shifted left and right when the average value of the distribution of the Euclidean distance difference is different from the difference of the ideal Hamming distance. The edge shift is evaluated from the difference between the average values of the distributions using the signal. According to this concept, a method for evaluating the extended edge shift at each time (from (formula D1) to (formula D6)) or a method for evaluating the average value of the Euclidean distance difference distribution calculated independently (from (formula 7)) Equation 13)) evaluates the edge shift. The evaluation main bit string is not limited to that shown in FIG. 18, and various variations including the case where N _2T is 3 can be used.

  As described above, by performing the evaluation based on the equivalent edge shift for the evaluation main bit string shown in FIG. 18, the correlation with SbER and an index common to the concept is also performed from the viewpoint of the affinity of the evaluation main bit string. L-SEAT can be provided as an evaluation index with improved performance. This is the fourth improvement in the present invention.

(Problem 3) There is a need to provide an evaluation index and an adjustment method that can be independently evaluated in correspondence with recording pulse conditions or adaptive recording pulse parameters for realizing recording adjustment in a short time. As a general optical disk apparatus, not only a single standard but also a high-density optical disk based on CD, DVD, BD, or BD must be similarly supported. Depending on these standards, adaptive recording pulses are different. Further, as an evaluation index for recording adjustment, it is preferable to use an index suitable for each of edge shift and jitter in the time axis direction measured by the time interval analyzer, jitter, V-SEAT, L-SEAT of the present invention, and the like. desired. In order to realize this, first, a parameter table for recording adjustment may be selected. In addition, a comprehensive response can be achieved by arranging a circuit for calculating an evaluation value such as an edge shift and an SNR factor for each edge of the reproduction signal. An example of the block configuration of the circuit for adjusting the recording condition is shown in FIG. In the figure, a reproduction signal 51 reproduced from an optical disk medium and subjected to analog filter processing (not shown) is converted into 6 to 8-bit digital data by an A / D converter 21 and equalized by an automatic equalizer 22. After that, it is binarized by the PRML decoder 23 and a binarized bit string 52 is output. The signal quality evaluation circuit 30 for recording condition adjustment includes edge quality evaluation circuits 40, 41, and 42, a selector 60, a recording pulse quality evaluation table 35, and a timing adjuster 36. The edge quality evaluation circuit 40 evaluates edge shift in the time axis direction for CD / DVD for each edge, the edge quality evaluation circuit 41 evaluates V-SEAT for BD, and the edge quality evaluation circuit 42 has a high density. Evaluate L-SEAT for BD. Each edge quality evaluation circuit calculates an edge shift amount, an extended edge shift, or an SNR factor for each edge. The selector 60 selects the output of the edge quality evaluation circuit in accordance with the type of disc to be recorded / reproduced. In the recording pulse quality evaluation table 35, the binarized bit string 52 and the edge evaluation index output from the edge quality evaluation circuit are synchronized, and patterns are classified according to adaptive recording pulses, and separated into 4 × 4 tables and the like. Then, the average value and standard deviation are calculated for each table element. The CPU 140 adjusts each parameter of the adaptive recording pulse while referring to this result. With the above configuration, it is possible to adjust the parameters of adaptive recording pulses corresponding to a plurality of optical disk media. Such a configuration makes it possible to adjust the parameters of a plurality of adaptive recording pulses in parallel, and within a short time and in comparison with a method using a single reproduction signal quality evaluation index. The recording pulse conditions can be adjusted using the trial writing area.

(Problem 4) Regarding the improvement of the accuracy of the trailing edge position control of the mark Next, for a high-density optical disk having a recording capacity of 30 GB or more, even when the recording pulse adjustment unit is 1/16 of the channel bit period, A means for controlling the edge position of the recording mark with high accuracy will be described.

  In the conventional BD recording pulse, since the end position of the last pulse is changed by changing the width of the last pulse, the amount of heat energy injected into the trailing edge forming portion of the mark changes. The problem is that the rate of change of the mark end position with respect to the change of position becomes large.

  Therefore, in the present invention, a parameter dTLP for designating the last pulse start position is introduced. FIG. 45 is a diagram showing a recording pulse waveform obtained by adding the dTLP of the present invention to a conventional BD recording strategy. dTLP is defined as shown in FIG. 45 with reference to the NRZI channel bit signal of the recording data.

  If the dTLP is changed while the TLP is fixed, the position of the last pulse can be changed without changing the width of the last pulse. In this case, since the change in the amount of heat energy injected into the trailing edge forming portion of the recording mark is small, the rate of change in the trailing edge position of the mark with respect to the change in the last pulse end position can be reduced. The reason why the trailing edge position of the recording mark is changed by this method is considered to be that the maximum temperature at the mark end forming portion changes due to the change in the size of the gap immediately before the last pulse.

  FIG. 46 is a diagram showing a parameter table of the last pulse start position dTLP introduced in the present invention. The dTLP values are classified and designated in three ways when the recording mark length is 3T, 4T and 5T or more, and the following space length is 2T, 3T, 4T and 5T or more in total 3 × 4 ways.

  Next, the effect of introducing the parameter dTLP for specifying the last pulse start position will be described. FIG. 47 shows the result of measuring the edge shift of the trailing edge of the 4T mark while changing the last pulse end position corresponding to the 4T mark as in the experiment shown in FIG. Here, the last pulse end position was changed by two methods, and the results were plotted respectively. One is based on a conventional BD recording pulse, and the last pulse end position is changed by changing only the last pulse width TLP while the last pulse start position is fixed. Based on the recording pulse of the invention, the last pulse end position is changed by changing the last pulse start position dTLP while the last pulse width TLP is fixed. The last pulse end position adjustment unit was 1/16 of the channel bit period. The experimental conditions such as the used write-once optical disk sample and the recording / reproducing conditions are the same as in the above-described experiment. Here, the time width TLP of the last pulse of the 4T mark recording pulse train is uniformly changed regardless of the subsequent space length. Further, the adjustment unit of the TLP value at this time is 1/16 of the channel bit period according to the conventional BD.

  According to FIG. 47, when the last pulse end position is changed using the conventional BD recording pulse (TLP control), the edge shift of the trailing edge of the recording mark when the last pulse end position is changed by one unit. While the change of about 16 points, it was 6 points when the recording pulse of the present invention (dTLP control) was used. Therefore, by using the dTLP of the present invention, it should be possible to always keep the edge shift value within ± 5% which is the above-mentioned standard of good recording quality.

  As described above, regarding the adjustment of the recording condition corresponding to the high density recording condition such that the capacity becomes 30 GB / surface or more based on the BD system, the above-mentioned problems of the conventional technology are solved, and (1) the adjustment result Based on the adjustment result, the recorded data is excellent in reproduction compatibility, and the quality of the recorded data is measured with an evaluation index of a reproduction signal such as SbER to guarantee a sufficiently good result. (3) Adaptive recording pulse condition adjustment can be performed in a short time, and (4) an evaluation index and adjustment method for adjusting a recording condition capable of controlling the trailing edge position of a mark with high accuracy, and an optical disk using the same The device can be provided.

  The essence of the present invention is that an optical disc using a code with a minimum run length of 2T, such as a BD, uses a target signal for three or more hamming distances (corresponding to N2T = 0, 1, and 2) to generate a reproduction signal. In the evaluation method, the quality of the edge of interest is evaluated by evaluating the extended edge shift at each time, or by evaluating the average value of the Euclidean distance difference distribution calculated independently, and recording based on this A method for adjusting the conditions and an optical disc apparatus mounting the method.

  As described above, the optical disc apparatus that realizes high-density recording equivalent to 30 GB or more in BD can be provided by the recording condition adjusting method using L-SEAT of the present invention as an evaluation index.

The block diagram which shows the structure of the reproduction signal evaluation circuit for implement | achieving the optical disk apparatus of this invention. Experimental results that summarize the relationship between the recording power and the number of bit errors measured using a prototype three-layer recordable optical disc sample. Simulation results summarizing the relationship between SNR and SbER. The experimental result showing the relationship between a bit error rate and SbER. Relationship between recording capacity and Euclidean distance difference distribution. Distribution of Euclidean distance difference used for evaluation of Tsfp (3s, 2m) obtained by simulation. Relationship between SNR and Ec '. Relationship between reproduction power and distribution center shift. FIG. 6 is a schematic diagram showing an equivalent edge shift. FIG. 6 is a schematic diagram showing an equivalent edge shift. Relationship between Hamming distance and cos θ. Relationship between dEDL and dEDR. Relationship between average value of dEDL and dEDR and extended edge shift D. Relationship between SNR, Ec 'and average value of extended edge shift D. PR (1, 2, 2, 2, 1) correspondence evaluation bit string table. Feature-extracted evaluation bit string table for PR (1,2,2,2,1). Main bitstream and second main bitstream generation operation table. Evaluation main bit string (N2T _max = 2). Comparison of evaluation main bit strings. Comparison of evaluation main bit strings. Comparison of evaluation main bit strings. The block diagram of an evaluation circuit. Evaluation main bit string (N2T _max = 3). Another embodiment of the evaluation main bit arrays (N2T _max = 2). Another embodiment of the evaluation main bit string (N2T _max = 3). The figure which shows a response | compatibility with an evaluation main bit stream and a recording pulse table. The figure which shows a response | compatibility with an evaluation main bit stream and a recording pulse table. The figure which shows the edge shift evaluation by L-SEAT. Another figure showing edge shift evaluation by L-SEAT. L-SEAT distribution and SAM distribution. Relationship between reproduction power and L-SEAT evaluation index. The block diagram which shows the structure of a symmetrical automatic equalizer. Results of recording adjustment by L-SEAT. Results of recording adjustment by L-SEAT. Results of recording adjustment by L-SEAT. Results of recording adjustment by L-SEAT. The figure which shows the power margin after recording adjustment. The figure which shows the relationship between a bit error rate and jitter. FIG. 6 is a schematic diagram showing a method for adjusting a recording pulse. 6 is a flowchart of a recording adjustment method of the present invention. The figure which shows the focus adjustment method using the evaluation method of this invention. 1 is a schematic diagram showing a configuration of an optical disc device. The figure which shows the recording pulse waveform of the conventional BD. The figure which shows the relationship between the last pulse end position at the time of using the recording pulse of conventional BD, and the edge shift of the trailing edge of a recording mark. The figure which shows the recording pulse waveform of this invention. The figure which shows the parameter table of the last pulse start position dTLP of this invention. The figure which shows the relationship between the last pulse end position at the time of using the recording pulse of this invention, and the edge shift of the trailing edge of a recording mark. The flowchart of the recording pulse adjustment method of this invention.

  Embodiments of a recording condition adjusting method and an optical disc apparatus according to the present invention will be described below with reference to the drawings.

FIG. 23 shows another embodiment of a table listing edge evaluation main bit strings according to the present invention. Here, the case where N _2T is 3 is shown. The total number of main bit strings is 20, and the underlined bits in each main bit string represent the edge of interest. Main bit string No. 1-12 is the same as that shown in FIG. 13-20 corresponds to the case where the number of consecutive 2Ts is 3. As described above, in the SbER shown in FIG. 15 and FIG. 16, there is one second most likely evaluation bit string, and therefore there are three 2Ts in which the binarized bit string is “0000011001100”. Also, the quality of the reproduced signal is evaluated with “0000110011000” having a Hamming distance of 3 as the second most likely evaluation bit string. On the other hand, this is not evaluated in the evaluation main bit string shown in FIG. Depending on the recording density and the condition of the disk medium, it is necessary to evaluate these bit strings independently, that is, the difference in edge shift between 2T marks to be recorded cannot be ignored when the number of consecutive 2Ts is 2 or 3. In this case, although the circuit scale increases, it is necessary to use the evaluation main bit string of FIG. Further, by using the evaluation main bit string of FIG. 23, as seen in main bit strings No. 15 and 17, when the mark preceding Tsfp (2s, 2m) is 3T or more (No. 15) and 2T (No. 17) can be separated and evaluated. Even when an adaptive recording pulse corresponding to the length of the preceding mark is used as the recording pulse, not only the space preceding the mark to be recorded, but also the one-to-one recording pulse table by the evaluation main bit string of FIG. Can be obtained for recording adjustment corresponding to. As for the 2T continuous number (N _2T ) included in the evaluation main bit string, an appropriate one may be used by judging such a situation. The relationship with the evaluation bit string for the calculation of SbER is a relationship corresponding to 1: 1 as in the case of N _2T = 2 described above. The case where N _2T is 4 or more will not be described because it is redundant, but from the relationship between FIG. 18 and FIG. 23, a general engineer related to the optical disc technology should be able to easily expand it.

FIG. 24 shows another embodiment of a table listing edge evaluation main bit strings according to the present invention. Here, the case where the Hamming distances of L and R are made equal when N _2T is 2 is shown. The total number of main bit strings is 12, and the underlined bits in each main bit string represent the edge of interest. The difference from FIG. 18 is the generated bit string and the Hamming distance for generating the main bits for L and R. When used with the evaluation main bit string of FIG. 18, a 1: 1 correspondence relationship with the evaluation bit string of SbER cannot be obtained, but the deviation of the distribution with respect to the SNR can be canceled in principle, so the SNR shown in FIG. Better SNR dependence (constant in principle) can be obtained than dependence. When the first priority is given to changes in the SNR due to environmental changes in the drive device or medium, such an evaluation main bit string may be used.

FIG. 25 shows another embodiment of a table listing edge evaluation main bit strings according to the present invention. Here, a case where the L and R hamming distances are made equal for N _2T of 3 is shown. The characteristics and evaluation performance of the table for FIG. 23 are the same as in the description of FIG. 24, and good SNR dependency (constant in principle) can be obtained.

  FIG. 26 and FIG. 27 are examples showing the correspondence between the evaluation main bit string and the recording pulse table. It is another Example of the table which enumerated the main bit stream number. FIG. 26 is an evaluation table corresponding 1: 1 to the evaluation main bit string shown in FIG. 18 and the recording pulse of the front and rear edge 4 × 4 table type. As seen in the figure, for example, the evaluation result of Tsfp (2s, 2m) includes the main bit string No. The result of 9 may be used. With reference to this table, if a recording adjustment circuit of the drive device is configured, L-SEAT can be used to adjust each parameter of a 4 × 4 table type recording pulse. FIG. 27 summarizes the L and R Hamming distances in this case. As described above, the edge evaluation result by L-SEAT can be easily developed so as to coincide with the parameter table of the recording pulse. Similarly, the evaluation main bit string shown in FIGS. 23 to 25 can be evaluated corresponding to the parameter table of the recording pulse.

  FIG. 28 and FIG. 29 are simulation results showing an example of edge evaluation by L-SEAT corresponding to the recording pulse parameter table. The simulation conditions are the same as described above, the recording density is equivalent to 33 GB / surface in BD, and the PR class is (1, 2, 2, 2, 1). Here, a simulation result when Tsfp (2s, 2m) is shifted by + 0.2T (moved to the right by 0.2T) is shown. FIG. 28 shows a case where a method (Equation D1) to (Equation D6) for evaluating the extended edge shift at each time is used. As can be seen in FIG. 28B, an edge shift of Tsfp (2s, 2m) is detected, and it can be seen that the corresponding distribution is shifted to the right. A good recording condition can be obtained by adjusting each parameter of the recording pulse so that the edge shift approaches zero. FIG. 29 shows a case where a method (Equation 7) to (Equation 13) for evaluating the average value of the distribution of the Euclidean distance difference calculated independently is used. Here, simulation results are shown for L and R shifts with both Hamming distances being two. As can be seen from FIG. 29A, when the edge shift is zero, the average value of the distribution of L and R is different from the ideal Euclidean distance difference (= 1), but both have the same average value in the error range. . On the other hand, as shown in FIG. 29A, when the edge shift is not zero, the average values of the L and R distributions are separated in the reverse direction. Therefore, good recording conditions can be obtained by adjusting the parameters of the recording pulse so that the average values of the L and R shift distributions coincide. As described above, if the hamming distances of the L and R evaluation main bit strings are made equal, the recording condition can be adjusted without depending on the SNR by using symmetry. As described above, L and R evaluation main bit strings having different Hamming distances can be used.

FIG. 29 compares the L-SEAT distribution obtained by simulation with the SAM distribution. The simulation conditions are the same as described above, the recording density is equivalent to 33 GB / surface in BD, and the PR class is (1, 2, 2, 2, 1). The average value of each SAM distribution shifts to approach zero depending on the decrease in SNR, whereas the average value of the L-SEAT distribution is zero and constant without depending on the SNR. Is confirmed. When N _2T is 3 or more as the evaluation main bit string, it is an extended form of this, and the same result is obtained.

  FIG. 30 shows the result of experimentally confirming the SNR dependency of L-SEAT. This is a result obtained by carrying out a reproduction experiment while changing the reproduction power in L0 of the prototype three-layer disc described above, and corresponds to the result of FIG. 8 according to the prior art. The horizontal axis in the figure represents the reproduction power of 1.2 mW as 100%. Since the reproduction signal amplitude is proportional to the reproduction power, but the noise (amplifier noise) of the photodetector is constant, this experiment is a result of changing the SNR of the reproduction signal by changing the reproduction power. The L-SEAT index was sorted into 4 × 4 tables for the front and rear edges of the recording mark according to the configuration shown in FIG. 22, and the shift and jitter were evaluated. FIG. 31A shows measured values of L-SEAT jitter, and the increase in jitter depending on the decrease in reproduction power reflects the change in SNR. On the other hand, FIG. 31B shows the edge shift evaluation results for Tsfp (2s, 2m). It can be seen that the value of the edge shift is constant regardless of the reproduction power (SNR). This is a feature of the method of the present invention that enables evaluation by separating a margin evaluation index based on a difference in Euclidean distance into an edge shift component and an SNR dependent component by L-SEAT. As a result, it was confirmed that adjustment of recording conditions with high reproduction compatibility can be performed by using this method without depending on SNR changes based on individual differences of drive devices and environmental conditions.

  Here, an automatic equalizer suitable for use in recording adjustment will be described.

FIG. 32 is a block diagram showing the configuration of the symmetric automatic equalizer of the present invention. As described above, when L-SEAT is used, the recording pulse can be adjusted stably with respect to the change in SNR. On the other hand, in an actual drive device, (1) asymmetry of the light spot search direction mainly due to the relative tilt angle (tangential tilt) angle between the disk medium and the optical head, and (2) the automatic equalizer There is asymmetry in the time axis direction of the reproduction signal based on the asymmetry of the tap coefficient. Since the distortion of the reproduction signal in the time axis direction is detected as an edge shift, it can be an obstacle to adjusting the recording condition with excellent reproduction compatibility. For example, even if an edge shift remains in a recording mark, if the internal tap coefficient is asymmetrically learned so that the automatic equalizer compensates for it, the measured edge shift is small and it is determined that the recording is good. The In general, since the structure of the playback system differs depending on the drive manufacturer or model, recording such data that can be easily played back only by the drive is an optical disc that is a medium-replaceable storage system. Is a problem that must be solved. The symmetric automatic equalizer shown here provides a solution. In the figure, a reproduction signal 51 reproduced from an optical disk medium (not shown) is converted into digital data by an A / D converter (not shown), equalized by an automatic equalizer 22, and then binarized by a PRML decoder 23. A binarized bit string 52 is output. The tap coefficients C ₀ , C ₁ , C ₂ ,... Of the automatic equalizer are automatically set so that the RMS error between the target signal based on the binarized bit string 52 and the output signal of the automatic equalizer 22 is minimized. A learning process is performed. This algorithm is generally called an LMS (Least Mean Square) method and is implemented by the LMS circuit 62. In this configuration, the tap coefficients a ₀ , a ₁ , a ₂ ,... Updated by the LMS circuit are temporarily stored in the buffer 64, and the work register 65 used for the actual operation of the FIR filter has a time as shown in FIG. A value averaged between tap coefficients (a combination of a ₀ and a _N−1 , etc.) at positions symmetrical in the axial direction is set. With such a configuration, the automatic equalizer can symmetrize the tap coefficients and prevent the edge shift of the recording mark from being distorted and reproduced. In addition, a circuit-like group delay may remain in the IV conversion amplifier and other filters included in the photodetector. Such group delay can be reduced by mounting the group delay compensator 61 as necessary. The group delay compensator 61 can be realized using an FIR having a predetermined value of asymmetric tap coefficients. Furthermore, by using the circuit of this configuration, a well-recorded reference disk is reproduced, and the tangential tilt amount is adjusted so that SbER or L-SEAT jitter is minimized, so that the time axis direction of the light spot It is possible to reduce asymmetry. With such a configuration, the automatic equalizer can operate only for adjusting the frequency characteristic of the reproduction signal. Not only L-SEAT, but also the symmetrical automatic equalizer according to this configuration can obtain recording conditions with high reproduction compatibility even when combined with a conventional recording adjustment method. Since the result of the LMS circuit 62 can be directly transferred to the buffer 64 by adding a circuit such as a selector, the symmetric automatic equalizer of this configuration operates as a normal (no symmetric type limitation) automatic equalizer. It is also easy to make it.

  The following results are obtained using a symmetric automatic equalizer with 21 taps.

FIG. 33 to FIG. 37 are experimental data showing the result of adjusting the condition of the recording pulse using L-SEAT. Here, while changing the four recording pulse parameters Tsfp (2s, 2m), Tsfp (3s, 2m), Tsfp (2s, 3m), and Tsfp (3s, 3m) in L0 of the above-described prototype three-layer disc, It is the result of measuring L-SEAT jitter, L-SEAT shift, and SbER. For SbER, the measurement was performed in the same manner as in normal reproduction, without limiting the symmetry of the tap coefficient of the automatic equalizer. The unit for adjusting the edge of the recording pulse was T / 64, and the linear velocity for recording / reproduction was set such that the data transfer rate was equivalent to twice the speed of BD. As can be seen from these, the zero point of the L-SEAT shift and the bottom conditions of the L-SEAT jitter and the SbER agree with each other with an accuracy of a pulse width of T / 64 or less. In general, since the recording pulse width adjustment unit is about T / 16, it can be confirmed from these results that very good recording condition adjustment can be performed using L-SEAT shift and L-SEAT jitter. It was. As a result of performing such adjustment for all the recording pulse parameters, the SbER value was improved from ³ × 10 ⁻³ to 1 × 10 ⁻⁷ . FIG. 37 shows measurement results of recording power and bit error rate after recording adjustment. A good power margin of about ± 10% can be obtained.

  FIG. 38 shows experimental results showing the relationship between the bit error rate and L-SEAT jitter. Here, the relationship between the L-SEAT jitter and V-SEAT jitter and the bit error rate was measured while changing the recording power, defocus, spherical aberration, tangential tilt and radial tilt of the disk medium. As seen in the figure, it was confirmed that the correlation between the bit error rate and the jitter was improved in L-SEAT compared to V-SEAT. The reason for this is as described above.

  Based on the results of the above experiments and simulations, the recording condition adjusting method of the present invention will be described with reference to the drawings.

  FIG. 39 shows an embodiment of a method for adjusting the adaptive parameter of the recording pulse. Here, a case where the adaptive parameter of the recording pulse is a 4 × 4 type table is shown. The edge shift and jitter measurement results by L-SEAT are sorted into 4 × 4 tables as described above. At this time, the recording pulse condition is changed, recording is performed on the optical disk medium, the corresponding portion is reproduced, the corresponding L-SEAT shift value is evaluated, and the recording pulse parameter is determined so as to minimize it. As a result, good recording pulse conditions can be obtained. As seen in the results of FIGS. 33 to 36, by adjusting not only the L-SEAT shift but also the minimum jitter condition, it is possible to obtain a more stable adjustment result with respect to various fluctuations. It becomes like this. As is clear from this example, since the recording pulse parameter and its evaluation value correspond one-to-one, a plurality of recording pulse parameters can be changed at once to perform recording / reproduction, thereby simultaneously The recording pulse parameters can be optimized in parallel. As a result, the trial writing time in the drive device can be significantly shortened. Specifically, in the method of sequentially determining the recording pulse parameters one by one, the processing time takes about 30 seconds to 1 minute with a double speed drive device, but when parallel processing is performed using this method, You should be able to finish trial writing in about 1 second. In the case of such an adjustment method, stable adjustment can be performed if there is a fixed condition part in the recording pulse condition. In general, it is desirable to apply long mark formation conditions such as Tsfp (5 s, 5 m) and Telp (5 s, 5 m) to this.

  FIG. 40 is a flowchart showing the overall flow of recording pulse adjustment. First, in step S101, the group delay of the reproduction circuit is checked as necessary to determine the group delay correction condition of the reproduction circuit shown in FIG. Next, in step S102, the operation mode of the automatic equalizer is set to the symmetric mode. In step S103, while reproducing the reference data and the like, the defocus amount, the spherical aberration correction amount, and the tilt amount of the disk medium are adjusted so that the reproduction evaluation index such as SbER and L-SEAT jitter is in the best state. As described above, the tangential tilt needs to be adjusted with particular consideration by reproducing a plurality of reference data or adding conditions that provide the best recording sensitivity. In step S104, appropriate basic pulse and power conditions are determined using recording data consisting of marks and spaces of 5T or more, taking into consideration the symmetry of the reproduction signal, the S / N ratio, the amount of crosstalk, and the like. This fixes the long mark recording conditions Tsfp (5 s, 5 m) and Telp (5 s, 5 m) in the 4 × 4 table. Tsfp (5s, 5m) is the pulse condition for the leading edge, and Telp (5s, 5m) is the pulse condition for the trailing edge. In steps S105 and S106, the adjustment is performed until the remaining edge shift becomes a predetermined value (for example, ± 0.1% T) or less while adjusting the adaptive parameter of the recording pulse. In step S107, SbER, the bottom value of the bit error rate and the power margin are evaluated for the obtained recording pulse, and the performance evaluation of the recording pulse is performed to determine whether a predetermined performance can be obtained. If sufficient, the process returns to step S104, and the same adjustment is performed by changing the basic pulse and power. If a predetermined performance is obtained by such a series of flows, the adjustment is finished.

  FIG. 41 shows experimental results showing the relationship between the focus offset amount and SbER. As the automatic equalizer, the symmetrical automatic equalizer of the present invention was used. By using this relationship and minimizing SbER, an appropriate adjustment of the focus offset value can be realized. The same method can be applied to radial tilt, tangential tilt, spherical aberration, and various adjustments. By such a method, step S103 in FIG. 39 can be performed.

  Next, a specific adjustment method for the last pulse start position dTLP introduced in the present invention will be described.

  As described above, the measurement result of edge shift and jitter by L-SEAT is 4 × 4 of the recording mark length and the leading space length for the front edge of the mark, and 4 × 4 of the recording mark length and the trailing space length for the trailing edge of the mark. Sort into each table. The dTLP parameters are sorted in the same manner as the L-SEAT for the trailing edge. However, since the 2T mark recording pulse does not have the last pulse, there are actually three types of dTLP (3T, 4T, and 5T or more) for the recording mark length. , The subsequent space length is sorted into 3 × 4 tables (2T, 3T, 4T and 5T or more).

  At this time, each value of dTLP sorted into 3 × 4 is changed and recorded on the optical disk medium, and the corresponding portion is reproduced to evaluate the L-SEAT shift value of the corresponding pattern, and each shift value is minimized. Thus, by determining the value of each individual dTLP, it is possible to obtain a good recording pulse condition. Here, minimizing the shift value means minimizing the absolute value of the shift value. That is, in the recording pulse adjustment method described above, each dTLP value is determined so that each L-SEAT shift value is closest to zero.

  FIG. 48 is a flowchart showing the overall flow of dTLP adjustment. As a premise at the start of adjustment, it is assumed that the group delay correction condition of the reproducing circuit is determined using the above-described method, and the operation mode of the automatic equalizer is set to a symmetric mode. Further, it is assumed that an initial value of dTLP is set for the recording apparatus.

After processing starts,
In step S201, the defocus amount, the spherical aberration correction amount, the tilt amount of the disk medium, and the like are adjusted so that the reproduction evaluation index is in the best state while reproducing the reference data and the like.
In step S202, random data is recorded on the optical disk medium, and the corresponding portion is reproduced to obtain a reproduction signal waveform.
In step S203, the L-SEAT shift value for the trailing edge of the mark is calculated by being classified into 3 × 4 as described above based on the acquired reproduction signal waveform.
In step S204, it is determined whether the calculated shift value of L-SEAT is the minimum. If the result is “No”, the dTLP value of the non-minimum pattern is updated in step S205. In this way, the process is repeated until the L-SEAT shift values of all 3 × 4 patterns are minimized, and the process is terminated.

  It should be noted that the above adjustment flow can be similarly applied to recording pulse parameters other than dTLP.

  Embodiments relating to the optical disc apparatus of the present invention will be described below.

  FIG. 1 is an embodiment showing the configuration of a reproduction signal evaluation circuit for realizing the optical disk apparatus of the present invention. In the figure, a reproduction signal 51 reproduced from an optical disk medium and subjected to analog filter processing (not shown) is converted into 6 to 8 bit digital data by an A / D converter 21 and equalized by an automatic equalizer 22. Thereafter, the signal is binarized by the PRML decoder 23 and a binarized signal 52 is output. The reproduction signal quality evaluation circuit 30 for calculating L-SEAT comprises a main bit discriminating circuit 31, an evaluation bit string generation circuit 32, an Euclidean distance calculation circuit 33, a recording pulse corresponding pattern classifier 34, and an evaluation value totaling circuit 35. Is done. The main bit string discriminating circuit 31 stores predetermined main bit string data, and determines whether or not the binarized signal 52 includes the main bit string. When the binarized signal 52 includes a main bit string, the evaluation bit string generation circuit 32 performs the XOR process described in FIG. 18 and the like to generate L and R evaluation bit strings. The Euclidean distance calculation circuit 33 calculates the Euclidean distance between the target signal of the T, L, and R evaluation bit strings and the equalized reproduction signal 53 output from the automatic equalizer 22. The recording pulse corresponding pattern sorter 34 uses the value of the Euclidean distance difference to evaluate the extended edge shift at each time (from (Equation D1) to (Equation D6)), or the distribution of the Euclidean distance difference calculated independently. According to the method of evaluating the average value ((Expression 7) to (Expression 13)), each value is statistically processed in a format according to the recording pulse adaptation table. The evaluation value totaling circuit 35 obtains the table shown in FIG. The CPU 140 refers to this, controls the recording pulse setting circuit (not shown), changes the parameters of the recording pulse, and adjusts the parameters of the recording pulse according to the method shown in FIG.

  FIG. 42 is a schematic diagram showing an example of the configuration of an optical disc apparatus equipped with the reproduction signal evaluation method of the present invention. The optical disk medium 100 mounted on the apparatus is rotated by a spindle motor 160. At the time of reproduction, the laser power / pulse controller 120 controls the current flowing to the semiconductor laser 112 via the laser driver 116 in the optical head 110 so that the light intensity instructed by the CPU 140 is generated, and the laser light 114 is generated. The laser beam 114 is condensed by the objective lens 111 to form the light spot 101 on the optical disc medium 100. The reflected light 115 from the light spot 101 is detected by the photodetector 113 via the objective lens 111. The photodetector is composed of a plurality of photodetecting elements. The reproduction signal processing circuit 130 reproduces information recorded on the optical disc medium 100 using the signal detected by the optical head 110. The present invention is built in the reproduction signal processing circuit 130 as the circuit block shown in FIG. With such a configuration, the optical disc apparatus of the present invention is an apparatus that realizes a BD of 30 GB or more, and can optimize recording pulse conditions by trial writing to ensure a good system margin and reproduction compatibility.

  The present invention relates to a method for adjusting recording conditions of a large-capacity optical disc and an optical disc apparatus.

21 A / D converter 22 Automatic equalizer 23 PRML decoder 30 Reproduction signal evaluation circuit 31 Main bit string discrimination circuit 32 Evaluation bit string generation circuit 33 Euclidean distance calculation circuit 34 Recording pulse corresponding pattern sorter 35 Evaluation value totaling circuit 51 Reproduction signal 52 Binary signal 53 Equalized reproduction signal 100 Optical disk 101 Optical spot 110 Optical head 111 Objective lens 112 Semiconductor laser 113 Photo detector 114 Laser light 115 Reflected light 116 Laser driver 120 Laser power / pulse controller 130 Reproduction signal processor 140 CPU
160 Spindle motor

Claims (9)

A method for adjusting a recording condition in an optical disc in which information is recorded using a code having a shortest run length of 2T, and the information is reproduced using an adaptive equalization method and a PRML method,
As the adaptive equalization method, the tap coefficient Cn is a value averaged between positions symmetrical in the time axis direction of the tap coefficient an updated by the LMS method,
Binarizing a reproduction signal waveform obtained from the optical disc by the PRML method to obtain a first binarized bit string;
Generating second and third binarized bit strings from the first binarized bit string as bit strings obtained by shifting a target edge by 1T to the left and right;
Generating first to third target signal waveforms corresponding to the first to third binarized bit strings;
The reproduction signal waveform is W, the first target signal waveform is T, the second target signal waveform is L, the third target signal waveform is R, and the Euclidean distance between W, T, R, and L is ED (a, b) (where a and b are either W, T, R, or L), and the evaluation value for an error in which the edge of interest shifts to the left is xL, and an error for an error that shifts to the right Assume that the evaluation value is xR, the edge shift amount of the edge of interest is the extended edge shift D,
xL

,
xR

Represented by
The extended edge shift D is

Represented by
Calculating a shift evaluation value that is a statistical average value Δ of the extended edge shift D;
The correction amount corresponding to the error probability of the edge of interest is SNR factor S

Represented by
Using the SNR factor S to calculate a jitter value corresponding to the error probability of the edge of interest;
In order to form a recording mark having a length NT, where N is an integer, a pulse train composed of (N−1) pulses is used, and the edge shift evaluation value and the jitter value are minimized. And adjusting the start position of the last pulse, which is the last pulse in the pulse train. Based on the reproduction signal waveform, the edge shift evaluation value corresponding to the trailing edge of the recording mark is divided into three cases when the recording mark length is 3T, 4T and 5T or more, and the subsequent space length is 2T, 3T, 4T and A process of classifying and calculating in a total of 3 × 4 patterns of 4 types in the case of 5T or more;
Determining whether each of the edge shift evaluation values calculated by classification is minimum;
Have
2. The recording condition adjusting method according to claim 1, wherein the last pulse start position is adjusted until the edge shift evaluation values calculated by the classification are minimized.   3. The recording condition adjusting method according to claim 2, wherein an adjustment unit of the last pulse start position is 1/16 of a channel bit period.   In an optical disc apparatus having a function of recording information on an optical disc medium using a code having a shortest run length of 2T and reproducing the information using an adaptive equalization method and a PRML method.
  As the adaptive equalization method, the tap coefficient Cn is a value averaged between positions symmetrical in the time axis direction of the tap coefficient an updated by the LMS method,
  Means for binarizing a reproduced signal waveform obtained from the optical disc by the PRML method to obtain a first binarized bit string;
  Means for generating second and third binarized bit strings from the first binarized bit string as bit strings obtained by shifting a target edge by 1T to the left and right;
  Means for generating first to third target signal waveforms corresponding to the first to third binarized bit strings;
  The reproduction signal waveform is W, the first target signal waveform is T, the second target signal waveform is L, the third target signal waveform is R, and the Euclidean distance between W, T, R, and L is ED (a, b) (where a and b are either W, T, R, or L), and the evaluation value for an error in which the edge of interest shifts to the left is xL, and an error for an error that shifts to the right Assume that the evaluation value is xR, the edge shift amount of the edge of interest is the extended edge shift D,
  xL

,
  xR

Represented by
  The extended edge shift D is

Represented by
  Means for calculating a shift evaluation value which is a statistical average value Δ of the extended edge shift D;
  The correction amount corresponding to the error probability of the edge of interest is SNR factor S

Represented by
  Means for calculating a jitter value corresponding to an error probability of an edge of interest using the SNR factor S;
  In order to form a recording mark having a length NT, where N is an integer, a pulse train composed of (N−1) pulses is used, and the edge shift evaluation value and the jitter value are minimized. An optical disc apparatus comprising: means for adjusting a start position of a last pulse which is a final pulse in a pulse train.   Based on the reproduction signal waveform, the edge shift evaluation value corresponding to the trailing edge of the recording mark is divided into three cases when the recording mark length is 3T, 4T and 5T or more, and the subsequent space length is 2T, 3T, 4T and Means for classifying and calculating in a total of 3 × 4 patterns of 4 patterns for 5T or more,
  Means for determining whether each of the edge shift evaluation values calculated by classification is minimum,
  5. The optical disc apparatus according to claim 4, wherein the last pulse start position is adjusted until the edge shift evaluation values calculated by the classification are minimized.   6. The optical disc apparatus according to claim 5, wherein the last pulse start position adjustment unit is 1/16 of a channel bit period.   As an adaptive equalization method, using a method in which the tap coefficient is limited to be centrosymmetric with respect to the direction corresponding to the time axis;
  As the adaptive equalization method, the tap coefficient Cn is a value averaged between positions symmetrical in the time axis direction of the tap coefficient an updated by the LMS method,
  Binarizing a reproduction signal waveform obtained from an optical disc by the PRML method to obtain a first binarized bit string;
  Generating second and third binarized bit strings from the first binarized bit string as bit strings obtained by shifting a target edge by 1T to the left and right;
  Generating first to third target signal waveforms corresponding to the first to third binarized bit strings;
  The reproduction signal waveform is W, the first target signal waveform is T, the second target signal waveform is L, the third target signal waveform is R, and the Euclidean distance between W, T, R, and L is ED (a, b) (where a and b are either W, T, R, or L), and the evaluation value for an error in which the edge of interest shifts to the left is xL, and an error for an error that shifts to the right Assume that the evaluation value is xR, the edge shift amount of the edge of interest is the extended edge shift D,
  xL

,
  xR

Represented by
  The extended edge shift D is

Represented by
  Calculating a shift evaluation value that is a statistical average value Δ of the extended edge shift D;
  The correction amount corresponding to the error probability of the edge of interest is SNR factor S

Represented by
  Using the SNR factor S to calculate a jitter value corresponding to the error probability of the edge of interest;
  In order to form a recording mark having a length NT, where N is an integer, a pulse train composed of (N−1) pulses is used, and the edge shift evaluation value and the jitter value are minimized. Adjusting the start position of the last pulse, which is the last pulse in the pulse train;
  An information recording method, wherein information is recorded using a code having a shortest run length of 2T based on the adjusted recording condition.   Based on the reproduction signal waveform, the edge shift evaluation value corresponding to the trailing edge of the recording mark is divided into three cases when the recording mark length is 3T, 4T and 5T or more, and the subsequent space length is 2T, 3T, 4T and A process of classifying and calculating in a total of 3 × 4 patterns of 4 types in the case of 5T or more;
  Determining whether each of the edge shift evaluation values calculated by classification is minimum;
  Have
  8. The information recording method according to claim 7, wherein the last pulse start position is adjusted until the edge shift evaluation values calculated by the classification are minimized.   9. The information recording method according to claim 8, wherein an adjustment unit of the last pulse start position is 1/16 of a channel bit period.

Images