JP5046331B2 - Phase modulator - Google Patents

Description

  The present invention relates to a phase modulation apparatus.

  Conventionally, a spatial light modulator (SLM) using LCoS (Liquid Crystal on Silicon) is known. When a voltage is applied to the pixel electrode, the liquid crystal molecules of LCoS rotate on a plane perpendicular to the substrate and change the amount of phase modulation of incident light. However, since the phase modulation amount changes nonlinearly with respect to the voltage applied to the pixel electrode, there is a problem that a desired phase modulation amount cannot be obtained.

Since the LCoS silicon substrate is processed by a semiconductor process, it cannot be made thick and its mechanical strength is weak. Therefore, as shown in FIG. 26, the silicon substrate 21 is distorted by the stress generated by each process of element manufacture, and the flatness of the mirror surface of LCoS is lowered. Further, in LCoS, the thickness of the liquid crystal layer 27 is not uniform. Therefore, the amount of phase modulation in each pixel differs depending on the thickness of the liquid crystal layer 27. That is, there is a problem that the wavefront reflected and output by the LCoS type SLM is greatly distorted due to the variation in the thickness of the liquid crystal layer 27 and the distortion of the reflection surface, and the phase modulation amount varies from pixel to pixel. Specifically, when the x direction and the y direction of the pixel position are represented as (x, y) and the voltage is V, the phase modulation amount Φ (V, x, y) is represented by the following expression.

Thus, the phase modulation amount Φ (V, x, y) is obtained as the sum of φ (V, x, y) that depends on the voltage and Φ ₀ (x, y) that does not depend on the voltage. Here, φ (V, x, y) is expressed by the following equation.

Here, Δn (V) is a birefringence with respect to a polarization component having an electric field that vibrates in a direction parallel to the alignment direction of the liquid crystal. d (x, y) is the thickness of the liquid crystal layer 27 at the positions x and y. That is, φ (V, x, y) depends on the thickness d (x, y) of the liquid crystal layer. φ (V, x, y) has a different value for each pixel. In each pixel, the relationship between the voltage V and φ (V, x, y) is non-linear. On the other hand, Φ ₀ (x, y) is mainly caused by distortion of the LCoS reflecting surface (silicon substrate 21). Hereinafter, the nonlinearity between the voltage and the phase modulation amount and the variation of the phase modulation amount for each pixel due to the variation of d (x, y) are collectively referred to as voltage-dependent phase modulation characteristics. In other words, the voltage-dependent phase modulation characteristic indicates the property of φ (V, x, y) out of the phase modulation amount Φ (V, x, y). Also, the variation in the amount of phase modulation for each position x and y due to the distortion of the LCoS reflecting surface indicated by Φ ₀ (x, y) is called voltage-independent distortion. A method for correcting this voltage-dependent phase modulation characteristic and voltage-independent distortion has been proposed (for example, Non-Patent Document 1).

  Further, although not an LCoS type SLM, there has been proposed a method of correcting voltage-dependent phase modulation characteristics using a phase modulation type SLM (for example, Non-Patent Document 2).

In addition, a method has been proposed in which distortion of an output wavefront is measured with a two-beam interferometer in a phase modulation SLM, and voltage-independent distortion is corrected using a pattern for canceling the distortion. (For example, Patent Document 1)
Phasecalibration of spatially nonuniform spatial lightmodulator (Applied Opt., Vol.43, No. 35, Dec. 2004) Highlystable wavefront control using a hybrid liquid-crystalspatial light modulator (Proc. SPIE, volume 6306, Apr.2006) International Publication WO2003 / 036368

  In Non-Patent Documents 1 and 2, in the LCoS SLM, the distortion of the output light wavefront is measured and corrected using a two-beam interferometer. However, the measurement with the two-beam interferometer has a problem that the voltage-dependent phase modulation characteristic and the voltage-independent distortion are mixed and measured. Further, in Non-Patent Document 1, regarding correction of voltage-independent distortion, only a relatively linear portion is extracted from nonlinear characteristics, and accurate correction cannot be performed.

  In Non-Patent Document 2, voltage-dependent distortion is measured with a polarization interferometer in a phase modulation SLM. Based on the measurement result, 4 × 4 pixels adjacent to each other are taken as one block, a lookup table is created for each block, and the voltage-dependent phase modulation characteristics are corrected by using the lookup table.

An object of the present invention is to provide a phase modulation apparatus capable of efficiently correcting the phase modulation characteristics of all pixels with a small amount of data .

In order to achieve the above object, the present invention includes a plurality of pixels arranged two-dimensionally adjacent to each other, and each pixel performs phase modulation on input light in response to application of a drive voltage. An optical modulator, input value setting means for setting an input value for each pixel, and a plurality of groups based on a phase modulation characteristic indicating a phase modulation amount when a voltage is applied to each pixel. A reference data map set to be associated with one of the plurality of reference data, a plurality of reference data provided in a one-to-one correspondence with the plurality of groups, and the reference data map. On the other hand, the specifying means for specifying the reference data corresponding to the group to which the pixel is associated, the input value input for each pixel, the control value with reference to the reference data specified by the specifying means Conversion means to convert to Converting the control value to the voltage value, provides a phase modulation device, characterized in that each pixel consists of a driving means for driving by the driving voltage of the voltage value.

  According to such a phase modulation apparatus, all pixels are allocated to a plurality of groups based on the phase modulation characteristics, and the same reference data is used for all the pixels in one group. For this reason, it is not necessary to have reference data for each pixel, and the phase modulation characteristics of all the pixels can be corrected efficiently with a small amount of data.

Each pixel can be driven with a voltage value within an operating voltage range, and the driving means converts the control value into a voltage value within a predetermined voltage range set within the operating voltage range. And driving each pixel with the driving voltage having the voltage value, and the predetermined voltage range is at least when a voltage having a voltage value in a predetermined range is applied to at least one of the plurality of pixels. It is preferably set based on voltage-dependent phase modulation characteristics indicating the phase modulation amount in one pixel . Thereby, since the drive voltage is controlled within a predetermined voltage range based on the voltage-dependent phase modulation characteristic within the operating voltage range, the drive voltage can be controlled with high accuracy.

  Each of the plurality of reference data is a lookup table, and the lookup table includes a plurality of first values that can be taken by an input value, the plurality of first values, and voltage-dependent phase modulation characteristics. It is preferable that a plurality of second values to be taken by the control value are stored in a one-to-one relationship because the relationship with the phase modulation amount shown is a predetermined linear relationship. As a result, the voltage-dependent phase modulation characteristic can be corrected so that the relationship between the input value and the phase modulation quantity indicating the voltage-dependent phase modulation characteristic becomes a predetermined linear relationship, and thus a desired phase modulation quantity is obtained. be able to.

  The reference data map preferably associates pixel position information with reference data corresponding to a group associated with the pixel. This makes it possible to perform phase modulation by reliably selecting reference data that matches the characteristics of the pixel.

Further, the reference data map includes a voltage-dependent phase modulation characteristic indicating a phase modulation amount in each pixel when a voltage having a voltage value in a predetermined range is applied to each pixel , and a predetermined range of the reference pixel. Based on the value indicating the difference from the voltage-dependent phase modulation characteristic indicating the phase modulation amount in the reference pixel when the voltage of the voltage value is applied , the position information of the pixel and the group associated with the pixel are it is preferably set to so that associates the corresponding reference data. Thereby, all the pixels can be grouped according to the voltage-dependent phase modulation characteristics.

In addition, the pixel serving as the reference, a voltage-dependent phase modulation characteristic indicating a phase modulation amount at the respective pixel when a voltage is applied to the voltage value of a predetermined range to each pixel, a predetermined in each of the plurality of pixels Selected from the plurality of pixels based on a value indicating a difference from an average of voltage-dependent phase modulation characteristics indicating a phase modulation amount in each pixel when a voltage having a voltage value in a range of is applied. Preferably there is. Thereby, all the pixels can be grouped according to the voltage-dependent phase modulation characteristics.

  The spatial light modulator is an LCoS spatial light modulator, the LCoS spatial light modulator has a glass substrate and a silicon substrate, and the reference data map is a silicon substrate of the LCoS spatial light modulator. It is preferable that the pixel position information and the reference data corresponding to the group associated with the pixel are associated with each other based on the value indicating the distortion of the pixel. Thereby, since the grouping of pixels is performed based on the value indicating the distortion of the silicon substrate, the grouping reflecting the phase modulation characteristics can be performed.

  The reference data map preferably associates pixel position information with reference data corresponding to a group associated with the pixel based on voltage-independent distortion caused by distortion of the silicon substrate. . Thereby, all the pixels can be grouped according to the voltage-independent phase modulation characteristic indicating the distortion of the silicon substrate.

  The reference data map associates pixel position information with reference data corresponding to a group associated with the pixel based on the inclination of the glass substrate and the silicon substrate of the LCoS spatial light modulator. Preferably it is. Thereby, all the pixels can be grouped according to the inclination of the silicon substrate.

  According to the phase modulation apparatus of the present invention, all pixels are allocated to a plurality of groups based on the phase modulation characteristics, and the same reference data is used for all the pixels in one group. For this reason, it is not necessary to have reference data for each pixel, and the phase modulation characteristics of all the pixels can be corrected efficiently with a small amount of data.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. As shown in FIG. 1, an LCoS type phase modulation device 1 includes an LCoS type spatial light modulator 2, a driving device 3 that drives the LCoS type spatial light modulator 2 with a voltage, and a desired pattern 13 that will be described later on the driving device 3. And the like.

  As shown in FIG. 2, the LCoS spatial light modulator 2 includes a silicon substrate 21 and a glass substrate 25 bonded to the silicon substrate 21 via a spacer 26. A liquid crystal layer 27 composed of liquid crystal molecules 28 is filled between the silicon substrate 21 and the glass substrate 25. A plurality of pixel electrodes 22 and a circuit (not shown) for controlling a voltage applied to each pixel electrode 22 are formed on the silicon substrate 21, and an alignment film 23 is formed on the pixel electrode 22. The glass substrate 25 includes a counter electrode 24 and an alignment film 23. The counter electrode 24 is opposed to the pixel electrode 22 through the liquid crystal layer 27. The liquid crystal molecules 28 of the liquid crystal layer 27 are formed so as to be parallel alignment, vertical alignment, or hybrid alignment. In the LCoS spatial light modulator 2, the pixel electrode 22 is made of aluminum, and also functions as a mirror that reflects incident light. One pixel electrode 22 corresponds to one pixel when performing phase modulation.

  A circuit (not shown) for controlling the voltage of each pixel electrode 22 is, for example, an active matrix circuit. In the active matrix circuit, a transistor and a capacitor are arranged on each pixel electrode 22, and further, a gate signal line extending in the row direction for selecting the pixel electrode 22 and a column direction for supplying an analog voltage signal to the transistor. The data signal line extended to is connected. The voltage of the pixel electrode is controlled by recording the analog voltage signal applied to the data signal line on the capacitor of the pixel electrode 22 selected by applying the Hi signal to the gate signal line. A predetermined voltage can be input to all the pixel electrodes 22 by sequentially switching the data line and the gate line to be selected.

  As shown in FIGS. 3A-3C, an arbitrary voltage is applied to the pixel electrode 22 to rotate the liquid crystal molecules 28. FIG. 3A shows the state of the liquid crystal molecules 28 when there is no potential difference between the pixel electrode 22 and the counter electrode 24. FIG. 3B illustrates a state where the potential difference is low, and FIG. 3C illustrates a state where the potential difference is large. Since the refractive index with respect to the polarization component changes depending on the voltage, the phase of the light component is modulated. As will be described later with reference to FIG. 8, the voltage range in which the pixel electrode 22 can operate is PS, but in the present embodiment, the pixel electrode 22 has a working voltage range in the operating voltage range PS. A voltage is applied within the QR.

  In order to modulate the phase of light using the LCoS type spatial light modulator 2, linearly polarized light parallel to the alignment direction of the liquid crystal is incident from the glass substrate 25 side. Light enters from the glass substrate 25, propagates through the liquid crystal layer 27, is reflected by the pixel electrode 22, propagates again through the liquid crystal layer 27, and exits from the glass substrate 25. Light undergoes phase modulation while propagating through the liquid crystal layer 27. The phase distribution of light can be controlled by modulating the phase with each pixel electrode 22. Therefore, the LCoS spatial light modulator 2 can control the wavefront.

  Each pixel has a unique voltage-dependent phase modulation characteristic and a voltage-independent phase modulation characteristic. In this embodiment, each of all T pixels belongs to one of r groups according to its voltage-dependent phase modulation characteristics. (Here, T and r are positive integers satisfying T> 0, r> 0, T> r. For example, r is 20.) Therefore, each group has a pixel whose phase modulation amount approximates. Belongs to.

  As shown in FIG. 1, the control device 4 is a personal computer, for example, and includes a central processing unit 41, a communication device 42, a memory 43, and an HDD 44. The HDD 44 stores the desired pattern 13. The central processing unit 41 is for judging the entire LCoS type phase modulation device 1.

  The desired pattern 13 has pixel position information and a value indicating a desired phase modulation amount desired to be achieved in the pixel (hereinafter referred to as a pixel input value) for all pixels. The value indicating the desired phase modulation amount is a digital signal with all the gradation numbers N (0 to N−1), and N = 256 in the present embodiment. The pixel input values of all N gradations from 0 to N−1 indicate the phase modulation amount for one cycle from 0 to 2π.

  When the LCoS type phase modulation apparatus 1 performs phase modulation, the central processing unit 41 reads the desired pattern 13 from the HDD 44 to the memory 43. The central processing unit 41 transmits the desired pattern 13 as input data to the driving device 3 via the communication device 42.

  The driving device 3 includes a communication device 33, a processing device 31, an adding device 35, an LUT processing device 36, a pixel position detecting device 37, a D / A (digital analog) circuit 32, a RAM 38, and a RAM 39. Have. The D / A circuit 32 includes a drive unit 321. The RAM 38 stores the correction pattern 12.

  The driving device 3 stores a program shown in a flowchart of FIG. 9 described later in a ROM (not shown). The processing device 31 reads this program from a ROM (not shown) of the driving device 3 and executes it to determine the entire LCoS type phase modulation device 1 and execute the phase modulation processing.

  The correction pattern 12 is for correcting voltage-independent distortion. The correction pattern 12 has pixel position information and a value (hereinafter referred to as a pixel correction value) to be added to the pixel input value in the pixel for all the pixels. The pixel correction value is also a digital signal in which all gradations are N (0 to N-1). The pixel correction values for all N gradations from 0 to N−1 indicate the phase correction amount for one cycle from 0 to 2π.

  The RAM 39 stores one LUT map 15 and r (r is a positive integer) LUTs 11. The LUT map 15 indicates which group of the r groups each pixel belongs to. The r LUTs 11 correspond to all r groups on a one-to-one basis. Each LUT 11 is for correcting the voltage-dependent phase modulation characteristics of the pixels belonging to the corresponding group. By correcting the voltage-dependent phase modulation characteristic of each pixel by the LUT 11 corresponding to the group to which the pixel belongs, the nonlinearity of the voltage-dependent phase modulation characteristic of each pixel can be linearly corrected, and the voltage dependency It is possible to correct the variation in the phase modulation characteristics for each pixel.

  The communication device 33 receives data such as the desired pattern 13 from the control device 4 and transfers it to the processing device 31. The processing device 31 generates a digital control signal including a vertical synchronization signal and a horizontal synchronization signal necessary for driving the LCoS spatial light modulator 2 based on the desired pattern 13. In parallel, the processing device 31 transfers the desired pattern 13 to the adding device 35. In parallel, the processing device 31 outputs pixel position information in the desired pattern 13 to the pixel position detection device 37.

  The adder 35 adds the pixel input value of the desired pattern 13 and the pixel correction value of the correction pattern 12 for each pixel, and sets the addition result as the control input value A corresponding to the pixel. As described above, each of the pixel input value and the pixel correction value is a digital signal having all the gradation numbers N (0 to N−1), and N = 256 in the present embodiment. Here, when the value of the addition result exceeds N, the addition result is further subjected to a phase folding process, and the result is set as a control input value A. That is, the control input value A corresponds to the phase modulation amount, and the control input value A for all N gradations from 0 to N−1 indicates one cycle (2π (rad)) of the phase modulation amount. Therefore, when the addition result becomes a negative value or a value of 255 or more in the phase folding process of the control input value A, the adder 35 obtains the remainder obtained by dividing the addition result by 256 as the control input value A. And set. For example, when the addition result is 512, the control input value A is 0. When the addition result is 384, the control input value A is 128. In addition, in order to obtain the remainder when the addition result is obtained by dividing the negative value by 256, first, the absolute value of the negative value is obtained, and then the sum of the absolute value and the result is 256. The minimum positive value among the numbers that are integer multiples may be used as the control input value A. For example, when the addition result is −64, the control input value A is 192. The adding device 35 transmits the control input value A of each pixel to the LUT processing device 36 together with the position information of the pixel.

  The pixel position detection device 37 refers to the LUT map 15 based on the pixel position information included in the desired pattern 13 and identifies the group number of each pixel group. The pixel position detection device 37 transfers the position information for each pixel and the LUT 11 corresponding to the specified group number (that is, the LUT 11 corresponding to the position information of the pixel) to the LUT processing device 36.

  For each pixel, the LUT processing device 36 refers to the LUT 11 that has received the position information of the pixel, and converts the control input value A received together with the position information of the pixel into a DA input value B. Here, the DA input value B is a digital signal having all the gradation numbers M (0 to M−1). Here, M is an integer satisfying M> N, and M = 4096 in the present embodiment. The LUT processing device 36 transmits the DA input value B of each pixel to the driving unit 321 together with the position information of the pixel.

  For each pixel, the driving unit 321 converts the DA input value B into an analog signal C indicating a voltage value within a predetermined usable voltage range (QR) that can be operated, and the LCoS spatial light modulator 2. These pixels are driven with a driving voltage having a voltage value indicated by the analog signal C.

  Here, as shown in FIG. 8, DA input values B of all 4096 gradations from 0 to 4095 are linearly assigned to the working voltage range QR. The drive unit 321 converts the DA input value B, which is any value from 0 to 4095, into an analog signal C that indicates a drive voltage value within the use voltage range QR (minimum value Q to maximum value R).

  The LUT processing device 36 converts the control input value A into the DA input value B in the LUT 11 for each pixel, and the drive unit 321 converts the DA input value B into a voltage value within the working voltage range QR. A voltage is applied to the LCoS spatial light modulator 2 after conversion to the analog signal C shown.

  The LUT map 15 is created according to the characteristics of the LCoS type spatial light modulator 2 provided in the LCoS type phase modulator 1 by a method described later. FIG. 4-6 is an example of the LUT map 15. For easy understanding, FIG. 4 shows an example where r = 4, FIG. 5 shows r = 8, and FIG. 6 shows r = 5.

  In the example of the LUT map 15 shown in FIGS. 4, 5, and 6, the thick line corresponds to the pixel area including all pixels, and one area divided by the thin line corresponds to one pixel. In FIG. 4, any one of the group numbers A to D is assigned to each pixel. In FIG. 5, any one of AH group numbers is assigned to each pixel. In FIG. 6, one of the group numbers A to E is assigned to each pixel. In FIGS. 5 and 6, the same group number is assigned to the pixels located in the area surrounded by the broken line.

FIG. 7 shows an example of one of the r LUTs 11. As shown in FIG. 7, LUT11 the control input value A and can take the value t _a (first value), the value _{t b} (second to take the DA input values B for the control input value A And the corresponding relationship.

In FIG. 7, the value t _{b taken} by the DA input value B is converted into a corresponding voltage value by the driving unit 321, and the LUT 11 is applied to the pixel belonging to the corresponding group, thereby achieving the pixel. The average value φ _ave of the phase modulation amount φ is also shown. However, the value related to φ _ave in FIG. 7 indicates the average value of the phase modulation amount when the measurement is actually performed using the LUT 11, and the LUT 11 does not have data corresponding to φ _ave . The average value phi _ave value t _a and the phase modulation amount DA input value A takes has a linear relationship. Moreover, in all the r LUT11, average phi _ave of the phase modulation amount for each values of t _a control input value A can take is to be substantially equal to each other, to be taken by the DA input value B value t _b is defined. Specifically, the value t _b that the DA input value should take is determined such that t _a = 0, φ _ave = 1.500, t _a = 1, φ _ave = 1.508, and so on.

  Therefore, for the pixels belonging to each group, if the control input value A is converted into the DA input value B by the corresponding LUT 11, and further the DA input value B is converted into the analog signal C and a voltage is applied, the voltage is applied to each group. The phase modulation amount φ obtained in the pixel to which it belongs is substantially linear with respect to the control input value A and has a small variation for each group.

  The correction pattern 12, the LUT 11, and the LUT map 15 are read from the ROM (not shown) of the driving device 3 to the RAM 38 and the RAM 39, respectively, when the LCoS phase modulation device 1 is activated. Alternatively, the correction pattern 12, the LUT 11, and the LUT map 15 are stored in the HDD 44 of the control device 4, and are transferred to the drive device 3 when the LCoS type phase modulation device 1 is activated, and are stored in the RAM 38 and the RAM 39, respectively. You may make it hold | maintain. Alternatively, the RAMs 38 and 39 may be integrated into a single RAM, and the single RAM may hold the correction pattern 12, the LUT map 15, and the LUT 11.

  The LCoS type phase modulation apparatus 1 having the above configuration operates as shown in FIG. 9 to perform phase modulation. First, in step 1, the communication device 33 receives the desired pattern 13 from the control device 4 and transfers it to the processing device 31. In step 2, the processing device 31 transmits the position information of each pixel to the pixel position detection device 37. In step 3, the pixel position detection device 37 refers to the LUT map 15 based on the position information of each pixel, and identifies the group number of the group to which each pixel belongs. In step 4, the pixel position detecting device 37 transmits the position information of each pixel and the LUT 11 corresponding to the group number specified for each pixel to the LUT processing device 36.

  Further, in parallel with step 2, the control device 31 transmits the desired pattern 13 to the adding device 35 in step 5. In step 6, the adder 35 adds the pixel input value in the desired pattern 13 and the correction input value in the correction pattern 12 for each pixel, and performs phase folding of the added value as necessary. The value obtained in this way is set as the control input value A corresponding to the position information of the pixel. In step 7, the LUT processing device 36 converts the control input value A into the DA input value B with reference to the LUT 11 received from the pixel position detection device 37 in step 4 for each pixel. In step 8, the drive unit 321 converts the DA input value B into an analog signal C and outputs the analog signal C to the LCoS spatial light modulator 2.

  In parallel with Steps 1 and 5, the processing device 31 generates a digital signal necessary for driving the LCoS spatial light modulator 2 in Step 9.

  In step 10, the LCoS spatial light modulator 2 modulates the phase of the incident light based on the analog signal C received from the drive unit 321 and the digital signal received from the control device 31 in step 8.

  When creating the LCoS type phase modulation device 1, the drive unit 321, the LUT map 15, the LUT 11, and the correction pattern 12 are set corresponding to the LCoS type spatial light modulator 2 provided in the LCoS type phase modulation device 1. The setting method will be described below. As the order of setting, first, the minimum and maximum voltages Q and R of the working voltage range QR of the D / A circuit 32 are set, then the LUT map 15 is created, and then the LUT map 15 is set. Based on this, an LUT 11 is created, and finally a correction pattern 12 is created.

ここで、Ｉ_ｍａｘは、ＬＣｏＳ型空間光変調器２に印加する電圧を動作電圧範囲内で変化させて測定される強度情報の最大値であり、Ｉ_ｍｉｎは、ＬＣｏＳ型空間光変調器２に印加する電圧を動作電圧範囲内で変化させて測定される強度情報の最小値である。

A method for setting the minimum value Q and the maximum value R of the operating voltage will be described with reference to FIG. First, in step 21, voltage-dependent phase modulation characteristics are measured for a plurality of pixels (for example, five pixels) arbitrarily selected from all the pixels using the polarization interferometer 60 shown in FIG. . The polarization interferometer 60 includes a xenon lamp 61, a collimator lens 62, a polarizer 63, a beam splitter 64, an LCoS phase modulator 1, an analyzer 65, image lenses 66 and 67, a bandpass filter 68, And an image sensor 69. Here, the driving unit 321, as shown in FIG. 12, the DA input values 0-4095, to assign linearly with applied possible operating voltage range LCoS spatial light modulator 2 (P-S) Is set. The light phase-modulated by the LCoS type spatial light modulator 2 is measured by the image sensor 69. The polarization direction of the polarizer 63 is shifted by 45 ° with respect to the alignment direction of the liquid crystal molecules of the LCoS type spatial light modulator 2. For this reason, the light (incident light) incident on the LCoS spatial light modulator 2 is shifted by 45 ° with respect to the alignment direction of the liquid crystal molecules 28. When the incident light passes through the liquid crystal layer 27, a phase difference is generated between the phase-modulated component of the incident light (a component parallel to the alignment direction of the liquid crystal molecules 28) and the component that is not phase-modulated. Therefore, the polarization direction of the light (reflected light) reflected by the LCoS spatial light modulator 2 depends on the phase modulation amount of the phase-modulated component of the incident light. Further, the orientation direction of the analyzer 65 is shifted by 90 ° with respect to the polarizer 63, and the intensity of light transmitted through the analyzer 65 depends on the polarization direction of the reflected light. Dependent phase modulation characteristics are measured as intensity information I. From the intensity information I measured by the image sensor 69 in a certain pixel, for example, the phase modulation amount φ is obtained using the following equation.

Here, I _max is the maximum value of intensity information measured by changing the voltage applied to the LCoS spatial light modulator 2 within the operating voltage range, and I _min is applied to the LCoS spatial light modulator 2. This is the minimum value of the intensity information measured by changing the applied voltage within the operating voltage range.

  In step 22, the DA input value-voltage dependent phase modulation characteristic is obtained for each pixel based on the measurement result of the image sensor 69. FIG. 12 is a graph showing the relationship between the DA input value-voltage-dependent phase modulation amount obtained for five pixels. From FIG. 12, the following (A)-(D) can be confirmed. (A) The amount of phase modulation is 2π (rad) or more. (B) There is a region where the amount of phase modulation hardly changes even when the voltage changes (a range where the DA input value is 0 to 800). (C) The amount of phase modulation at five pixels is different. (D) The phase modulation amount is non-linear with respect to the DA input value.

  In the LCoS spatial light modulator 2, if the phase modulation amount is 2π (rad), a phase modulation amount of 2π (rad) or more can be realized by performing phase folding processing. Therefore, the drive range of the voltage applied to the liquid crystal is sufficient if the phase modulation amount is 2π (rad). However, in practice, when distortion correction is performed, a certain amount of margin is required in consideration of variations in the phase modulation amount of each pixel. Therefore, a value that can secure a phase modulation amount of 2π (rad) or more is set as this value. In the embodiment, it is set to 3.5π (rad). Here, the phase folding process is to replace the phase with a value obtained by dividing the phase by 2π (rad) when the phase is equal to or greater than 2π (rad) or smaller than 0, similarly to the phase folding of the control input value.

  Specifically, in step 23, the minimum voltage Q of the working voltage applied to the LCoS type spatial light modulator 2 is equal to or higher than the threshold voltage at which the liquid crystal operates, and the maximum voltage R is a saturation voltage at which the operation of the liquid crystal is saturated. The phase modulation range between the minimum voltage Q and the maximum voltage R of the operating voltage is set to be approximately 3.5π (FIG. 12). The DA input value B is made to correspond to 4096 gradations in the range of the minimum voltage Q and the maximum voltage R of the use voltage set in this way. FIG. 13 shows the relationship between the DA input value B, the phase modulation amount, and the operating voltage range (QR) for the above five points when the minimum voltage Q and the maximum voltage R are set under such conditions. In the case of FIG. 12 in which the entire operable voltage range of the LCoS type spatial light modulator 2 is used, the DA input value B is about 1100-1800 for a range where the phase modulation amount is up to 0.5π-4π (rad). It was about 700 gradations. On the other hand, in FIG. 13, voltage control in 4096 steps is possible for a range up to 0.5π-4π (rad) where the phase modulation amount is the same. Therefore, the DA input value B has about five times the gradation for the same phase modulation amount range, and the voltage can be controlled with high accuracy. In other words, by setting the minimum and maximum voltages Q and R, the scale conversion of the working voltage range with respect to the DA input value B is performed. Thus, the drive unit 321 is set to linearly convert the DA input value B of 0-4095 into an analog signal C indicating the voltage value in the working voltage range QR.

  Next, a method for creating the LUT map 15 will be described with reference to FIG. The LUT map 15 is created after setting the minimum and maximum voltages Q and R of the use voltage range QR of the D / A circuit 32.

  In step 31, the polarization interferometer 60 shown in FIG. 11 obtains the relationship between the DA input value B and the voltage-dependent phase modulation amount for each pixel of the LCoS spatial light modulator 2. Specifically, the DA input value B having the same value is applied to all the pixels, and the phase modulation amount of each pixel is measured. That is, the DA input value B having the same value for all the pixels is converted into an analog signal C by the drive unit 321, and the LCoS spatial light modulator 2 is driven by the analog signal C, and the phase modulation amount is set. measure. This DA input value B is changed from 0 to 4095 and measurement is repeated. Alternatively, the DA input value B may be measured at intervals rather than for all values from 0 to 4095.

In step 32, based on the DA input value-phase modulation characteristic obtained in step 31 for each pixel, the relationship between the phase modulation amount (φ) and the DA input value (t _b ) is determined using a least square method or the like. Approximate with polynomial. For example, the DA input value is t _b
When the phase modulation amount is φ and a K-th order power polynomial is used as the polynomial, the polynomial is expressed as shown in Equation (3).

By obtaining the equation (3), the relationship between the DA input value (t _b ) and the phase modulation amount (φ) can be obtained while reducing the influence of measurement noise caused by a light source, an image sensor, or the like. Further, when the DA input value B is not measured for all values but is measured at intervals, the phase modulation amount for the DA input value B that is not measured can be estimated from the equation (3). In this way, the relationship (3) between the DA input value B and the phase modulation amount φ is obtained for all the pixels.

In step 33, a relationship between the DA input value B and a value obtained by averaging the phase modulation amount φ obtained in each pixel by the input of the DA input value B for all the pixels is obtained. Specifically, first, an average value of the phase modulation amounts for all the pixels is obtained for each DA input value B. Thereby, the relationship between the DA input value B and the average value φ _ave of the phase modulation amount is obtained. This relationship is obtained by an approximate expression. For example, when it is obtained by a K-th order polynomial, the DA input value B is obtained as t _{b, ave} (φ) as shown in equation (4).

In step 34, the least square error (RMS) value ε ₁ (DA) between the DA input value t _{b, ave} (φ) and the DA input value t _b (φ) for the averaged phase modulation amount obtained in (4). Hereinafter, the first least square error value) is obtained for each pixel from Equation (5).

Next, a pixel (maximum value pixel) having the largest _first least square error (RMS) value ε ₁ among all the pixels is found. It can be determined that the pixel having the largest least square error (RMS) value ε ₁ is the pixel whose phase modulation amount φ is farthest from the average value of all the pixels of the phase modulation amount.

In step 35, the least square error (RMS) value ε ₂ (hereinafter referred to as _second minimum 2) between the DA input value (hereinafter referred to as t _MAX (φ)) of the maximum value pixel and the DA input value t _b (φ). (Multiplier error value) is obtained for each pixel using equation (6).

In step 36, it obtains the second minimum square error (RMS) value epsilon ₂ of the maximum value obtained for all the pixels. On the other hand, the minimum value of the second least square error (RMS) value ε ₂ obtained for all pixels is zero. This is because t _b (φ) = t _MAX (φ) in the maximum value pixel. Then, the interval between the maximum value and the minimum value of the second least square error (RMS) value ε ₂ is divided into r intervals at equal intervals. Next, for each section, the pixels having the least square error value ε ₂ in the section are collected as one group. In this way, one group is formed for one section, and all pixels are distributed to 20 groups. The LUT map 15 stores the relationship between the pixel thus configured and the group to which the pixel belongs.

As described above, the LUT map 15 is configured by grouping pixels in which the amount ε ₂ indicating the voltage-dependent phase modulation amount characteristic exists in the same section as one group. Therefore, pixels having approximate voltage-dependent phase modulation characteristics can be grouped into one group.

  FIG. 4 is an example of the LUT map 15 created for the LCoS spatial light modulator 2 in which pixels having voltage-dependent phase modulation characteristics that are similar to each other are distributed substantially uniformly in the entire pixel region. . Pixels belonging to group A-D are distributed substantially uniformly in the entire pixel region. FIG. 5 is an example of the LUT map 15 created for the LCoS spatial light modulator 2 in which adjacent pixels have approximate characteristics. Since the pixels in the area surrounded by the broken line have characteristics close to each other, they belong to the same group.

  Further, depending on the phase modulation characteristics of the LCoS spatial light modulator 2, not only adjacent pixels are included in the same group as in groups A, B, and C shown in FIG. It may be included in the same group.

  Note that the above grouping method may be changed as shown in [1]-[5] below.

[1] When the pixels are grouped by the above method, depending on the characteristics of the LCoS spatial light modulator 2, all the pixels may not be uniformly distributed to the r groups. That is, the number of pixels belonging to each group may deviate greatly from T / r. On the other hand, according to the present method [1], all the pixels can be distributed to the r groups substantially uniformly. That is, the number of pixels belonging to each group can be approximately T / r. Specifically, the grouping in step 36 is changed as follows. First, the second least square error (RMS) values ε ₂ obtained for all the pixels are arranged in ascending order (or descending order). That is, a column in which the least square error (RMS) values ε ₂ for all pixels are arranged is created. This column is divided at substantially constant intervals so that r sections are formed as a whole. As a result, the number of pixels included in one section is approximately equal to T / r and is approximately equal to each other.

[2] A reference value t _MAX (φ) may be determined in advance. In this case, steps 33 and 34 are not executed.

[3] In the manufacturing process of the LCoS type phase modulator 2, when it is known that the performance of a predetermined pixel is significantly different from that of other pixels, the predetermined pixel is the first least square. with the highest value among the error (RMS) value epsilon _1. In that case, the equation (3) obtained for the pixel may be set as the reference value t _MAX (φ). Also in this case, steps 33 and 34 are not executed.

  [4] In step 31, only the phase modulation amount φ with respect to the DA input value B of one value (for example, the minimum value 0) may be obtained for each pixel. The processing from step 32 to step 35 is not performed, and grouping is performed in step 36 based on the measured phase modulation amount φ. For example, the phase modulation amounts φ of all the pixels are arranged in ascending order (or descending order). That is, a column in which the phase modulation amounts φ of all the pixels are arranged is created. This column of phase modulation amounts φ is divided at regular intervals to form r sections. Therefore, only T / r phase modulation amounts φ are arranged in one section. Pixels that achieve the phase modulation amount φ included in the same section are assumed to be in the same group. Therefore, T / r pixels that have achieved phase modulation amounts approximate to each other with respect to the same DA input value B can belong to the same group. A substantially equal number of pixels can be distributed to each group.

  Further, instead of arranging the phase modulation amounts φ of all the pixels in ascending order or descending order, the maximum value and the minimum value of the phase modulation amounts of all the pixels may be divided into r sections having the same length. . Pixels having values in the same section of the phase modulation amount φ are configured as the same group. In this case, the number of pixels belonging to one group may deviate greatly from T / r.

[5] In step 32, instead of using the DA input value t _{b, ave} (φ) for the averaged phase modulation amount _, the DA input value t _{b, 0} (φ) of a predetermined specific pixel is used. The first least square error (RMS) value ε ₁ ′ may be obtained from the following equation. At this time, step 33 is not performed.

  In the above embodiment, grouping is performed using scalar quantization, but the grouping method is not limited to this. For example, after obtaining Equation (3) for all the pixels in step 31, other methods using scalar quantization other than the above embodiment, or those having similar characteristics using vector quantization are r pieces. You may make it divide into groups.

  A method of creating the LUT 11 will be described with reference to FIG. First, in step 41, as in step 31, the relationship between the DA input value B and the voltage-dependent phase modulation amount is applied to each pixel of the LCoS spatial light modulator 2 by the polarization interferometer 60 shown in FIG. 11. Ask for it. In step 42, the DA input value-voltage dependent phase modulation characteristic is obtained for each pixel based on the measurement value obtained in step 41. The result is the same as in FIG. 13 described above, has non-linearity, and varies from pixel to pixel.

In step 43, a pixel-specific LUT is created for each pixel based on the obtained DA input value-voltage dependent phase modulation characteristic. That is, the equation (7) is obtained in the same manner as the equation (3).

  Here, the subscript (1) in the expression represents a value in the approximate polynomial obtained based on the first measurement. Other than that, it is the same as Formula (3).

In this way, an approximate expression indicating the relationship between the DA input value B and the phase modulation amount is obtained for all pixels. On the other hand, since the relationship between the control input value A and the phase modulation amount is linear, and 0.0−2.0π (rad) is represented by the 256-step control input value A, the control input value is represented by ta _{(1 )} And the relationship with the phase modulation amount φ is expressed as follows.

Here, ta ₍₁₎ is an integer value from 0 to 255, and const is an offset value. This offset value is set to the same value that can realize Equation (8) for all pixels. Equation (8) determining the relationship between the expression (7) assigned to control input values _{t a (1)} and _{t b.} In this case, rounding for t _b is an integer (or, rounded down, up) it is necessary to make. When representing an operation of rounding in ROUND, the relationship between t _b and t a ₍₁₎ is as follows.

A pixel-specific LUT for one pixel is created by associating the value of t _b obtained in (9) with the value 0-255 of t _{a (1)} . Such pixel-specific LUTs are obtained for all pixels.

  In step 44, the pixel-specific LUT created as described above is stored in the HDD 44. The pixel-specific LUT is temporarily used to obtain the LUT 11 for each group. The pixel-specific LUT stored in the HDD 44 is read out from the HDD 44 and transferred to the RAM 39 of the drive device 3 in the following processing, corresponding to the pixel position specified by the pixel position detection device 37. The pixel-specific LUT obtains the phase by calculation from the interference intensity output of the interferometer. At this time, the maximum value and the minimum value of the measured interference intensity are used, but these values may include an error. In steps 45-47, the extent of this error is evaluated.

Specifically, in step 45, similarly to the step 41, to measure the relationship between the DA input value t _b and the phase modulation amount φ for all pixels. However, in step 45, the LUT processing unit 36 converts the control input value A (0-255) into the DA input value B based on the pixel-specific LUT for each pixel obtained in the immediately preceding step 44. The drive unit 321 converts the DA input value B into an analog signal C, and drives the corresponding pixel of the LCoS spatial light modulator 2. In this way, the relationship between the control input value A (t _a ) and the voltage-dependent phase modulation amount φ is measured for all pixels. In step 46, a control input value-phase modulation characteristic is obtained based on the result of step 45. In step 47, it is determined from the result of step 46 whether the voltage-dependent phase modulation characteristic is corrected with a desired accuracy by the pixel-specific LUT. For example, a method is used in which it is determined that a desired accuracy is obtained if the control input value-voltage-dependent phase modulation characteristic is close to linear. The determination method is not limited to this example. If it is determined in step 47 that the desired accuracy is not obtained, step 43 is repeated to update the pixel-specific LUT based on the result of step 46, and the voltage-dependent phase modulation characteristics by the pixel-specific LUT Improve the accuracy of correction.

In step 43 performed for the second time, the relationship between the control input value A (t _{b (1)} ) and the phase modulation amount φ is approximated as follows.

The new control input value ta ₍₂₎ obtained by this processing is also expressed by 256 gradations and has a linear relationship with the phase modulation amount. Therefore, the following equation holds.

The relationship between the previous control input value A (ta ₍₁₎ ) and the current control input value A (ta ₍₂₎ ) can be expressed as follows from (10) and (11).

（１３）式により新たな制御入力値Ａ（ｔ_ａ（２））とＤＡ入力値Ｂ（ｔ_ｂ）との関係が求められる。尚、ステップ４３をＪ回（Ｊは、Ｊ＞２を満たす自然数）行なった場合には、以下のような、ｔ_ｂとｔ_ａ（Ｊ）の関係が得られる。

By substituting equation (12) into equation (9 ₎ , the relationship between t _b and t _{a (2)} is as follows.

The relationship between the new control input value A (t _{a (2)} ) and the DA input value B (t _b ) is obtained from equation (13). When step 43 is performed J times (J is a natural number satisfying J> 2), the following relationship between t _b and ta _(J) is obtained.

  Based on these values, a new pixel-specific LUT is created, and in step 44, the pixel-specific LUT is overwritten and stored in the HDD 44. Note that the data of the equations (7) to (14) are temporarily stored in the HDD 44 because they are used in step 48 described later. On the other hand, if it is determined in step 47 that the desired accuracy has been obtained, or if it is determined that no improvement in accuracy can be obtained compared to the pixel-specific LUT before update, the process proceeds to step 48.

  In step 48, the LUT 11 for each group is created based on the pixel-specific LUT obtained for all pixels.

Specifically, first, for each group, an average value of phase modulation amounts φ (hereinafter referred to as phase modulation amount average value φ _ave ) obtained for all pixels in the group is set for each DA input value (t _b ). Ask for. This is determined based on the pixel-specific LUT obtained for all the pixels belonging to the group. However, if there is a pixel having an extremely different DA input value-phase modulation characteristic in the group, the phase modulation amount average value φ _ave is obtained for the pixels excluding the pixel.

Then determine the relationship between the DA input value t _b or control input value t _a and the phase modulation amount average value phi _ave in approximation. Based on such approximate expression determines the LUT11 of each group showing the relationship between the control input values _{t a} and DA input value _{t b.} The LUT 11 for each group thus obtained is stored in a ROM (not shown) of the driving device 3. Also, the pixel-specific LUT stored in the HDD 44 is deleted.

Hereinafter, the approximate expression indicating the relationship between the DA input value t _b or control input value t _a and the phase modulation amount average value phi _ave, and the relationship between the control input values t _a and DA input value t _b on the basis of the approximate expression The method for obtaining the value will be specifically described in the following three cases (1) to (3).
(1) This is a case where the process of returning from step 47 to step 43 is not performed, and step 48 is reached without updating the pixel-specific LUT obtained based on the measurement result of step 42. In this case, the pixel-specific LUT is obtained based on the result of the first measurement (that is, the measurement in step 42).
(2) The case where the process of returning from step 47 to step 43 is performed once, and after step 45 is performed once, step 48 is reached. In this case, the pixel-specific LUT is obtained by updating based on the second measurement (that is, the first measurement performed in step 45).
(3) The case where the process of returning from step 47 to step 43 is performed twice or more, and after step 45 is performed twice or more, step 48 is reached. In this case, the pixel-specific LUT is obtained by updating based on the measurement performed in the Mth time (M is a natural number of 3 or more) (that is, the measurement in step 45 performed in the (M−1) th time). It is a thing.
<In the case of (1)>

第１回目の測定における制御入力値ｔ_ａと位相変調量平均値φ_ａｖｅとの関係が線形で、かつ、０．０−２．０π（rad）を２５６段階の制御入力値Ａで表わすために、制御入力値をｔ_ａ（Ｍ）とし、位相変調量の平均値φ_ａｖｅとの関係を以下のように表す。なお、ここで、Ｍ＝１である。

First, an approximate expression indicating the relationship between the DA input value t _b obtained by the first measurement and the phase modulation amount average value φ _ave is obtained as follows.

A linear relationship between the first control input values _{t a} and the phase modulation amount average value phi _ave in the measurement, and to represent 0.0-2.0π a (rad) in 256 steps of the control input value A The control input value is t _{a (M)} , and the relationship with the average value φ _ave of the phase modulation amount is expressed as follows. Here, M = 1.

t _{a (M)} is an integer value from 0 to 255, and const is the same offset value for all groups.

（１７）式に対して四捨五入を行なうことにより、以下の（１８―１）式が得られる。

By substituting equation (16) into equation (15), the following relationship (17) is obtained.

By rounding off the equation (17), the following equation (18-1) is obtained.

Equation (18-1) shows the relationship between the DA input value (t _{b (1)} ) and the control input value (ta ₍₁₎ ). Therefore, the LUT 11 is created based on the relationship between the control input value ta ₍₁₎ indicated by the equation (18-1) and the DA input value tb ₍₁₎ .
<In the case of (2)>

Ｍ＝２において、式（１６）を式（１９）に代入することで、以下の関係（２０）が得られる。

（２０）式を（１８−１）式に代入することにより、以下の（１８―２）式が得られる。

First, an approximate expression indicating the relationship between the previous control input values t _a and the current phase modulation amount average value phi _ave as follows. Note that M = 2.

By substituting equation (16) into equation (19) at M = 2, the following relationship (20) is obtained.

By substituting the equation (20) into the equation (18-1), the following equation (18-2) is obtained.

Expression (18-2) shows the relationship between the DA input value (t _{b (2)} ) and the control input value (ta ₍₂₎ ). Therefore, the LUT 11 is created based on the relationship between the control input value ta ₍₂₎ indicated by the equation (18-2) and the DA input value tb ₍₂₎ .
<In the case of (3)>
The following equation (18-3) is obtained by the same method as in the case of (2).

Expression (18-3) shows the relationship between the DA input value (t _{b (M)} ) and the control input value (ta _(M) ). Therefore, the LUT 11 is created based on the relationship between the control input value ta _(M) indicated by the equation (18-3) and the DA input value tb _(M) .

Instead of obtaining the average value φ _ave of the phase modulation amount φ, a value that minimizes the dispersion of the phase modulation amount φ in the group may be obtained and the LUT 11 may be created based on the value .

The LUT 11 shown in FIG. 7 stores the relationship between t _a and t _b and the phase modulation amount φ _ave obtained for a certain group by the above processing. By referring to the LUT 11 and performing the conversion in step 7 (FIG. 9), the relationship between the control input value A and the phase modulation amount φ becomes linear in the pixels belonging to the group. By using the LUT 11, the variation in the phase modulation amount for each pixel in the group is corrected, and the correction is realized so that the relationship between the control input value A and the phase modulation amount is substantially linear. Moreover, since the same equation (18) is used when creating the LUT 11 for each group, the variation in the phase modulation amount for each pixel is corrected over all groups, and the control input value A and the phase are corrected. Correction is realized such that the relationship with the modulation amount is substantially linear.

As described above, after the LUT 11 is created for each group, the correction pattern 12 is created. The voltage-independent distortion cannot normally be measured by itself, but can be measured by measuring the output wavefront of the LCoS type phase modulation device 1 with the voltage-dependent phase modulation characteristic corrected using the LUT 11. It is. The measurement of the light wavefront including voltage-independent distortion is measured using a two-beam interferometer. In this embodiment, a Michelson interferometer 80 shown in FIG. 16 is used as a two-beam interferometer. The Michelson interferometer 80 includes a laser light source 81, a spatial filter 82, a collimator lens 83, a polarizer 84, a beam splitter 85, an LCoS type phase modulator 1, a mirror 86, and image lenses 87 and 88. And CCD89. The polarization direction of the polarizer 84 is parallel to the polarization direction of the liquid crystal. Interference fringes generated by interference between the wavefront reflected by the mirror 86 and the wavefront reflected by the LCoS spatial light modulator 2 in the LCoS type phase modulator 1 are measured, and an analysis method shown in the following document is used. By using it, the output wavefront of the LCoS type phase modulation device 1 can be obtained from the measured interference fringes. That is, a voltage-independent distortion pattern is formed on the wavefront reflected by the LCoS spatial light modulator 2, and the wavefront reflected by the mirror 86 is a plane. By removing the carrier component, voltage-independent distortion can be obtained.
References: M. Takeda, H. Ina, and S. Kobayashi, "Fourier-transform method of fringe pattern analysis for computer-based topography and interferometry", J. Opt. Soc. Am., Vol. 72, 156-160 (1982).

  With reference to FIG. 17, a method of creating the correction pattern 12 for correcting the voltage-independent distortion will be described. First, in step 51, in the driving device 3, a pattern in which all the pixel values are 0 is stored in the RAM 38 as the initial correction pattern 12. In step 52, the central processing unit 41 sets a phase image in which all the pixel values are any one of 0 to 255 and equal to each other as the desired pattern 13, and transmits the desired pattern 13 to the driving device 3. In the driving device 3, the received desired pattern 13 is transmitted to the adding device 35, and pixel position information in the desired pattern 13 is transmitted to the pixel position detecting device 37. Note that the pixel position detection device 37 specifies the corresponding LUT 11 based on the position information of the pixel. In step 53, the adding device 35 adds the desired pattern 13 and the correction pattern 12, and uses the result of phase folding as the control input value A. In step 54, the LUT processing device 36 converts the control input value A into the DA input value B based on the specified LUT 11 and transfers it to the driving equipment value 3. In step 55, the drive unit 321 generates an analog signal C based on the DA input value B, and applies a use voltage to the LCoS spatial light modulator 2. In step 56, the output wavefront of the LCoS type phase modulator 1 is measured based on the output result of the CCD 89 of the Michelson interferometer 80. Since the voltage-dependent phase modulation characteristic is corrected using the LUT 11, the output wavefront measured in step 56 includes only voltage-independent distortion. In step 57, the sign of the measured voltage-independent distortion is reversed, and the correction pattern 12 is created. In step 58, the phase value of the correction pattern 12 is folded. In step 59, the phase value (phase modulation amount) of each pixel of the correction pattern 12 is converted into a control input value and re-expressed into 256 gradation values. Such conversion may be performed using, for example, an ideal relational expression between the phase modulation amount and the control input value, such as the equations (8) and (16), or the phase modulation amount and the control input value. This relationship may be held in the LUT 11 and performed based on this relationship. FIG. 18 is an example in which the correction pattern 12 is expressed as a gradation image having 256 gradations. In step 60, the obtained value of the correction pattern 12 expressed in 256 gradations is stored in a ROM (not shown).

  Even in the above processing, there is a possibility that a measurement error due to measuring interference is included as in the case of creating the LUT 11. In steps 61-65, it is evaluated how much this error exists. Specifically, in step 61, the adding device 35 adds the desired pattern 13 and the correction pattern 12 obtained in the immediately preceding step 60 in the same manner as in step 53, and if necessary, adds the result to the addition result. The control input value A is the one subjected to phase folding. Steps 62-64 are the same as Steps 54-56. In Step 62, the LUT processing unit 36 obtains the DA input value B for the control input value A obtained in Step 61. In Step 63, the drive unit 321 is obtained. Converts the DA input value B into an analog signal C and applies a drive voltage to the LCoS spatial light modulator 2. In step 64, the output wavefront is measured based on the output result of the CCD 89. In step 65, based on the measurement result, it is determined whether or not necessary correction has been performed by the correction pattern 12 obtained in the immediately preceding step 60. For example, a method is used in which it is determined that a desired accuracy is obtained if the flatness of the output wavefront is obtained. The determination method is not limited to this example. If it is determined in step 65 that the required accuracy has been obtained with the correction pattern 12, or if it is determined that the accuracy cannot be improved, the correction pattern creation processing is terminated. If the required accuracy cannot be obtained, the process returns to step 57, and the correction pattern 12 is recreated based on the voltage-independent distortion indicated by the result of step 64. Specifically, in step 60, the value of the correction pattern 12 obtained immediately before and the value of the correction pattern 12 obtained by repetition are added for each pixel, and the result of addition is a new correction pattern 12. Is stored in a ROM (not shown). Thus, the creation of the correction pattern 12 is repeated repeatedly. The correction pattern 12 having a desired accuracy is stored in a ROM (not shown) of the driving device 3.

  In the above-described LCoS type phase modulation apparatus 1 according to the present embodiment, all the pixels are allocated to a plurality of groups based on the phase modulation characteristics, and the same LUT 11 is used for all the pixels in one group. For this reason, it is not necessary to have an LUT 11 for each pixel, and the phase modulation characteristics of all the pixels can be efficiently corrected with a small amount of data. Therefore, even when it is difficult to mount a large-capacity memory (RAM) in the device, the LUT 11 can be stored in the drive device 3.

  Further, by storing the LUT 11 in the driving device 3, (1) processing for adding the desired pattern 13 and the correction pattern 12, and performing phase folding processing if necessary (processing in the adding device 35), (2 ) Processing for obtaining pixel position information (processing in the pixel position detection device 37), (3) In the LUT processing device 36, the control input value A is converted to the DA input value B based on the LUT 11, and the LCoS spatial light modulator 2 is performed using dedicated hardware (adder 35, pixel position detector 37, LUT processor 36). The processing time required for the processes (1) to (3) in the driving device 3 can be shortened compared to the case where the above processing is performed by the central processing unit 41 in the control device 4, for example. Therefore, these processes can be completed within one frame time.

  With the LUT map 15, it is possible to grasp the correspondence between the pixel position information and the group number. As a result, it is possible to perform phase modulation by selecting the LUT 11 that surely matches the characteristics of the pixel.

  Further, in the LCoS type phase modulation apparatus 1 according to the present embodiment, the LCoS type spatial light modulator 2 is required to be smaller than the operable voltage range for the DA input value B represented by 4096 gradations. The amount is controlled within the range of working voltage. Therefore, the voltage applied to the LCoS spatial light modulator 2 can be accurately controlled. In addition, the LUT 11 that is optimal for the pixel is identified using the LUT map 15. For this reason, the relationship between the control input value A and the voltage-dependent phase modulation amount is substantially linear by the LUT 11, and variations among pixels due to the voltage dependency are also corrected. Obtainable. Furthermore, by correcting the voltage-independent distortion using the correction pattern 12, more accurate phase modulation can be performed.

When correction is performed using the LUT map 15, LUT 11, and correction pattern 12 in the present embodiment, (1) no correction is performed at all, and (2) single correction is applied to all pixels. It was confirmed that the output wavefront can be measured with higher accuracy than in the case where correction is performed using the LUT 11 and the correction pattern 12. For example, the measured control input value in each condition of the - phase and modulation characteristic, the ideal control input value - the phase modulation characteristics _{(φ (t a) = 2π} / 256) and minimum squared error (RMS) value The following results were obtained.

  FIG. 19A is a diagram in which the phase modulation of the Laguerre Gaussian beam is measured using the LUT 11 and the correction pattern 12. Compared to FIG. 19B without such correction, FIG. 19A shows a concentric pattern as in theory.

  In addition, in the creation of the LUT 11 and the creation of the correction pattern 12, the creation process is repeated until the required accuracy is obtained or improvement in accuracy cannot be obtained. Therefore, the LUT 11 and the correction pattern 12 are obtained with high accuracy, and the voltage-dependent phase modulation characteristic and the voltage-independent distortion can be corrected with high accuracy.

  The phase modulation device and the method for setting the phase modulation device according to the present invention are not limited to the above-described embodiments, and various modifications and improvements can be made within the scope described in the claims. For example, in the LCoS spatial light modulator 2, the pixel electrode 22 also serves as a mirror. However, like the LCoS spatial light modulator 120 shown in FIG. 20, a dielectric mirror 29 is stacked on the pixel electrode 22. A configuration having the configuration may be used instead of the LCoS spatial light modulator 2. In the LCoS spatial light modulator 120, the same components as those in the LCoS spatial light modulator 2 are denoted by the same reference numerals and description thereof is omitted.

  The control device 4 and the drive device 3 are not limited to the above example, and the function of the control device 4 may be incorporated in the drive device 3.

  In the LCoS type phase modulation device 1 in the above embodiment, the correction pattern 12 is stored in the RAM 38 of the driving device 3, and the desired pattern 13 and the correction pattern 12 are added by the adding device 35. However, the LCoS shown in FIG. Like the mold phase modulation apparatus 100, the desired pattern 13 and the correction pattern 12 may be stored in the HDD 44, read out to the memory 43, and these values may be added in the control apparatus 4. At this time, the central processing unit 41 includes input value setting means 47. The input value setting means 47 adds the correction pattern 12 to the desired pattern 13. Further, the driving device 130 is not provided with an adding device and a RAM for storing the correction pattern. That is, the drive device 130 includes a communication device 133, a processing device 131, a pixel position detection device 137, an LUT processing device 136, a D / A circuit 132, and a RAM 139. The RAM 139 stores the LUT map 15 and the LUT 11. The D / A circuit 132 includes a drive unit 321.

  When performing phase modulation, the input value setting means 47 adds the correction pattern 12 to the desired pattern 13 and sets the control input value A by performing phase folding on the added value if necessary. The communication device 42 transmits the control input value A and pixel position information to the driving device 130. The communication device 133 transfers the control input value A and the pixel position information to the processing device 131. The processing device 131 sends the position information of the pixel to which the control input value A is input to the pixel position detection device 137 and sends the value of the control input value A of the pixel to the LUT processing device 136. Thereafter, the LCoS type phase modulation device 100 performs the same processing as the LCoS type phase adjusting device 1 of the embodiment, and the LCoS type spatial light modulator 2 modulates the phase of the incident light. In the above-described LCoS type phase modulation device 100, since there is no need to add the correction pattern 12 and the desired pattern 13 in the driving device 130, the capacity of the RAM mounted on the driving device 3 can be reduced.

  Further, like the LCoS type phase modulation device 200 shown in FIG. 22, the desired pattern 13, the correction pattern 12, the LUT 11, and the LUT map 15 are stored in the HDD 44, and these are read out to the memory 43 to obtain the DA input value. B may be obtained and transmitted to the driving device 230. In this case, the central processing unit 41 includes conversion means 46, input value setting means 47, and pixel position detection means 48. The driving device 230 does not include an adding device, an LUT processing device, a pixel position detecting device, and a ROM. That is, the drive device 230 includes a communication device 233, a processing device 231, and a D / A circuit 232 having a drive unit 321.

  When performing phase modulation, the input value setting means 47 adds the correction pattern 13 to the desired pattern 12 and sets the control input value A by performing phase folding on the added value if necessary. The pixel position detecting unit 48 refers to the LUT map 15 and identifies a group number corresponding to the pixel position information. The conversion means 46 converts the control input value A of the pixel into the DA input value B using the LUT 11 corresponding to the specified group number. The communication device 42 transmits the DA input value B to the driving device 230. The communication device 233 transfers the received DA input value B to the processing device 231. The subsequent processing is the same as in the embodiment, and the LCoS spatial light modulator 2 modulates the phase of the incident light. In the LCoS type phase modulation device 200, it is not necessary to provide the driving device 230 with a RAM that holds the desired pattern 13, the LUT 11, the LUT map 15, and the correction pattern 12. Therefore, the cost of the apparatus can be reduced.

  In the present embodiment, the D / A circuit 32 is provided in the driving device 3, but the D / A circuit is separated from the driving device 3, and the D / A circuit and the DA are provided on the LCoS spatial light modulator 2 side. A configuration may be adopted in which a receiving circuit that receives the input value B is newly provided. In this case, the DA input value B is transmitted from the driving device 3 to the receiving circuit on the LCoS side.

  In the drive device 3 of the present embodiment, the LUT 11 is read by the LUT processing device 36 via the pixel position detection device 37, but the LUT processing device 36 may directly process the LUT 11. In this case, the pixel position detection device 37 transmits information on the LUT 11 specified with reference to the LUT map 15 to the LUT processing device 36. The LUT processing device 36 performs LUT processing (processing for converting the control input value A into the DA input value B) while switching the LUT 11 to be referenced based on the received information on the LUT 11.

  1 has only one D / A circuit 32, a plurality of D / A circuits 32 are provided, and a plurality of analog signals are simultaneously output to the LCoS spatial light modulator 2. An analog signal may be simultaneously written in a plurality of pixels. In this case, the processing circuit of the driving device 3 outputs the DA input value B for a plurality of pixels to each D / A circuit 32 at the same time.

  In the present embodiment, the voltage-dependent phase modulation characteristics are measured for five pixels, and the minimum and maximum voltages Q and R are set based on the measurement results. However, the number of pixels to be measured is at least one. You only have to. In this case, the minimum and maximum voltages may be set based on the measured voltage-dependent phase modulation characteristic for at least one pixel.

Instead of the LUT 11, the approximate polynomial data (coefficients _{ak (I), ave} and I are natural numbers) obtained by the equations (15) and (19) are stored in a ROM (not shown) of the driving device 3, and the phase modulation amount When these are measured, they may be read out to the RAM 39. Similar to the method of creating the LUT 11 in the above-described embodiment (step 48), the formulas (18-1) and (18) are determined according to the number of times the data has returned from step 47 to 43 from the data and formula (16). -2) or (18-3), the relationship between the control input value A and the DA input value B can be obtained.

  In the above embodiment, measurement is performed on all pixels when the LUT 11 and the LUT map 15 are created. However, the phase modulation amount may not be measured for all the pixels, and only the representative pixel may be measured. For example, a plurality of adjacent pixels are configured as one block. As a block configuration method, for example, 4 pixels × 4 pixels are configured as one block. One pixel is used as a representative pixel in one block, and measurement is performed on the representative pixel in each block. Based on the measurement result, the blocks (representative pixels) are divided into groups. An LUT map 15 indicating the group created in this way is created. That is, the LUT map 15 shows the relationship between the block and the LUT 11 corresponding to the block. In this case, the same LUT 11 is applied to the pixels in one block.

  Further, as shown in FIG. 23, the value of the correction pattern 12 may be included in the LUT 11. Let t be the control input value A of the pixels in a certain group, and p be the value of the correction pattern 12 corresponding to one representative pixel among the pixels in that group (p = 64 in FIG. 23). In the above-described embodiment, the LUT 11 is applied after both are added. That is, the control input value A when referring to the LUT 11 is t + p. Although the desired image changes from time to time, since p is a fixed value, the reference position is always shifted by p. This is equivalent to shifting the reference start position of the LUT by p. In the correction pattern 12, instead of setting the value of one representative pixel in the group as p, the average value of the pixels in the group may be obtained and the value may be set as p.

Information for correcting voltage-independent distortion can be included in the LUT 11 by shifting the reference position of the LUT 11 of each pixel by a value corresponding to the pixel of the correction pattern. FIG. 23 includes data for correcting voltage-independent distortion included in the data of FIG. 4 with p = 64. For example, in FIG. 7, the value 1050 of _{t b} when _{t a} is 0 is, in FIG. 23 _{t a} has appeared at a position of 64. In this case, the adder 35 and the RAM 38 are unnecessary in the driving device 3 of the LCoS type phase modulation device 1 in FIG. Furthermore, in the phase modulation method described with reference to FIG. 9, step 6 is not necessary, and the processing device 31 transmits the input value sequence to the LUT processing device 36. Further, the LUT 11 used in steps 3, 4, and 7 is the LUT 11 to which the value of the correction pattern 12 as shown in FIG. 23 is added.

As another example, a case where p = 1 and p = −1 will be described. As shown in FIG. 7, the LUT11, the control input value _{t a} is 255,0,1, DA input value _{t b,} respectively, 3036,1050,1055 was supported. In contrast, in the case of p = 1, to the control input value _{t a} is 255,0,1, DA input value _{t b,} respectively, creating a LUT11 so 1050,1055,1058 corresponding do it. Or in the case of p = -1, to the control input value _{t a} is 255,0,1, DA input value _{t b,} respectively, may be creating a LUT11 so 3028,3036,1050 corresponding .

  As described above, by including the correction pattern information in the LUT 11 and converting the control input value A to the DA input value B using the LUT 11, distortion due to voltage independence can also be corrected. Therefore, the adding device 35 for processing the correction pattern 12 can be omitted, and efficient phase modulation is possible.

  In the LCoS type spatial light modulator 2, since the thickness of the glass substrate 25 is extremely thick, for example, about 3 mm, the glass substrate 25 is not distorted. Only the distortion of the silicon substrate 21 becomes a problem as shown in FIG. In FIGS. 24A and 24B (described later), the alignment film 23 and the counter electrode 24 are omitted. In the thickness of the liquid crystal layer 27 in FIG. 24A, as shown by d1 and d2, the strain difference of the silicon substrate 21 becomes the difference in thickness of the liquid crystal layer 27 as it is.

  That is, pixels having the same thickness of the liquid crystal layer 27 and applying an equal voltage have the same phase modulation amount. From the above, when the glass substrate is thick, if the distortion shape of the silicon substrate is known, pixels having the same phase modulation amount can be found. Therefore, instead of the method of creating the LUT map 15 shown in FIG. 14, for example, the LUT map 15 can be obtained by measuring the amount indicating the strain shape of the silicon substrate by one of the following three methods. it can.

  1. This will be described with reference to FIG. First, in step 71, pixel-specific LUTs are created for all pixels. Specifically, in the method of creating the LUT 11 shown in FIG. Next, in step 72, the phase modulation amount Φ is measured using the Michelson interferometer 80 shown in FIG. Specifically, a single control input value A having the same value for all pixels is converted into a DA input value B using a pixel-specific LUT, further converted into an analog signal C, and the analog signal C is applied. . In step 73, a pixel corresponding to the maximum phase modulation amount and a pixel corresponding to the minimum phase modulation amount are obtained from the obtained phase modulation amounts. In step 74, the maximum value and the minimum value of the phase modulation amount are divided into r equal sections. Pixels having a phase modulation amount in the same section are collected as the same group. The relationship between groups and pixels configured in this way is stored in the LUT map 15.

In this method, the Michelson interferometer 80 converts the control input value A to the DA input value B using the pixel-specific LUT and measures the phase modulation amount Φ. Therefore, the voltage dependent phase modulation characteristic is corrected. That is, the term φ depending on the voltage V in the equation (1) is corrected and there is no variation for each pixel. Therefore, the variation of the measured phase modulation amount Φ for each pixel is the variation for each pixel of Φ ₀ itself. Φ ₀ was an amount indicating strain of the silicon substrate. For this reason, in the LUT map 15 created by this method, all the pixels are grouped according to the voltage-independent phase modulation characteristic indicating the distortion of the silicon substrate.

  2. In this modification, the method for creating the LUT map 15 shown in FIG. 14 is changed as follows. In the embodiment, in step 31, the same voltage based on the same DA input value B is applied to each pixel of the LCoS spatial light modulator 2 using the polarization interferometer 60 of FIG. Measured. In addition, measurement was repeated with the DA input value B varied from 0 to 4095. On the other hand, in this modification, the Michelson interferometer 80 of FIG. 16 is used instead of the polarization interferometer 60. Further, the same voltage based on the same DA input value B is applied to each pixel of the LCoS spatial light modulator 2 to measure the phase modulation amount. However, this measurement is performed by applying a voltage value with respect to the DA input value B having a single value from 0 to 4095. Steps 32-35 are not performed. In step 36, the minimum value and the maximum value of the phase modulation amount for all the pixels obtained in step 31 are obtained, and the interval between them is divided into r equal parts. Pixels corresponding to phase modulation amounts having values in equally divided sections are grouped into one group. In this way, the relationship between pixels and groups is obtained and the LUT map 15 is created.

  Also in this method, the phase modulation amount Φ is measured by a Michelson interferometer without using the pixel-specific LUT as in the first modification. However, conversion from the control input value A to the DA input value B by the pixel-specific LUT is not performed. Therefore, the measured phase modulation amount Φ includes an amount Φ that depends on the voltage of equation (1). As shown in the equation (2), φ depends on the thickness d (x, y) of the liquid crystal layer 27. In the LCoS spatial light modulator 2 in which the glass substrate 25 is not distorted, the thickness d (x, y) of the liquid crystal layer 27 is the amount indicating the distortion of the reflecting surface as it is. Therefore, obtaining Φ in equation (1) results in an amount related to the strain of the silicon substrate. Therefore, also in this method, all the pixels can be grouped according to the voltage-independent phase modulation characteristic indicating the distortion of the silicon substrate.

  Similarly to the case of 3.2, in step 31, the phase modulation amount Φ achieved by each pixel is measured by the Michelson interferometer 80 of FIG. Based on this measurement result, a pattern in which the phase modulation amount Φ is a constant value in all pixels is created, and the LUT map 15 is created using the pattern. Specifically, the DA input value B is sequentially set to all values from 0 to 4095, and each pixel is driven by the corresponding analog signal C. Based on the obtained phase modulation amount, a pattern indicating a distribution of DA input values B in which the phase modulation amounts of all the pixels are equal to each other is obtained. Steps 32-35 are not performed. In step 36, the minimum value and the maximum value are obtained from the DA input values B distributed in the pattern obtained in step 31, and r is equally divided between them. Pixels that have achieved the amount of phase modulation in the same section are grouped into one group. In this way, the relationship between pixels and groups is obtained and the LUT map 15 is created. Also in this method, all the pixels can be grouped according to the voltage-independent phase modulation characteristic indicating the distortion of the silicon substrate.

  In the above method 1-3, the voltage-independent distortion is measured by the Michelson interferometer 80, and the pixels are grouped based on the difference in the film thickness d (x, y) of the liquid crystal layer 27. The method is not limited to this. If an amount indicating the difference in film thickness d (x, y) depending on the pixel position can be measured, an LUT map can be created in the same manner as described above based on the result. For example, the film thickness d (x, y) at each pixel position may be directly measured by optical measurement to perform grouping.

ここで、基準点Ｏを、ｄ_ｇ（ｘ，ｙ）＝０となる点とする。Ｌ_ｘ、Ｌ_ｙは、基準点Ｏから画素の位置（ｘ，ｙ）までの、それぞれｘ方向，ｙ方向の距離である。 In the above three cases, when the glass substrate 25 is tilted, the strain difference of the silicon substrate 21 does not become the film thickness difference as shown in FIG. When the glass substrate 25 is tilted in this way, in addition to the film thickness d _s (x, y) resulting from the distortion of the silicon substrate 21, the film thickness d _g (x, y resulting from the tilt of the glass substrate). Grouping is performed in consideration of the difference of y). For example, when the inclinations θ _x and θ _y of the lower surface of the glass substrate 25 with respect to the x and y directions are known, the LUT map 15 can be created by the following method. In FIG. 24B, the reference plane S1 is a plane parallel to the lower surface of the silicon substrate 21. That is, when the glass substrate 25 is not tilted, the lower surface of the glass substrate 25 coincides with the reference plane S1. The film thickness from the upper surface of the silicon substrate 21 to the reference plane S1 is d _s (x, y), the film thickness from the reference plane S1 to the glass substrate is d _g (x, y), and the following equation (21 ). D (x, y) is given by the sum of d _s (x, y) and d _g (x, y).

Here, the reference point O is a point where d _g (x, y) = 0. L _x and L _y are distances from the reference point O to the pixel position (x, y) in the x direction and the y direction, respectively.

ここで、電圧Ｖは、例えば、位相変調量Φ_０（ｘ，ｙ）の測定に使用した値とする。 Therefore, the voltage-dependent phase modulation amount φ _g (V, x, y) resulting from the film thickness difference d _g (x, y) can be calculated from _the following equation.

Here, the voltage V is, for example, a value used for measurement of the phase modulation amount Φ ₀ (x, y).

When the phase modulation amount is measured by the Michelson interferometer 80 of FIG. 16 using the LCoS type spatial light modulator 2 in which the glass substrate 25 is tilted in this way, the obtained phase modulation amount is expressed by the equation (1). Φ ₀ (V, x, y), and φ _g (V, x, y) does not affect Φ ₀ (V, x, y).

That is, the measured phase modulation amount Φ ₀ (V, x, y) is a pattern based on the difference in film thickness d _s (x, y), and φ _g (V, x, y) obtained by calculation. Is a pattern based on the difference in film thickness d _g (x, y). Therefore, the phase modulation amount Φ ₀ (V, x, y) measured by the Michelson interferometer 80 in FIG. 16 and φ _g (V, x, y) obtained by calculation from the equation (20) are added. Based on the phase modulation amount that has been subjected to the phase folding process (hereinafter referred to as the phase modulation amount including the effect of tilt), for example, grouping is performed as in Modification 2 described above. That is, in the grouping, the maximum value and the minimum value are specified among the phase modulation amounts including the tilt effect, and the interval between the maximum value and the minimum value is divided into r equal sections. Pixels having a phase modulation amount including a tilt effect in the same section are collected as the same group. Thereby, the LUT map 15 is created.

  By the above method 1-3 and the method in the case where the glass substrate 25 is tilted, the grouping of pixels can be performed by reflecting the amount indicating the distortion of the silicon substrate 21.

Instead of the LCoS type spatial light modulator 2, a general phase modulation type spatial light modulator, for example, an optical address type phase modulator, a MEMS type phase modulator, a variable mirror, or an analog type magneto-optical element is used. May be. As an optical address type phase modulator, for example, described in Yasunori Igasaki et al. “High Efficiency Electrially-Addressable Phase-Only Spatial Light Modulator”, Optical Review, Vol. 6, No. 4, pp. 339-344, 1999 Can be used. As a MEMS type phase modulator, for example, it was described in M. Friedrichs et al. “One Megapixel SLM with high optical fill factor and low creep actuators”, Optical MEMS and Their Applications Conference 2006, IEEE / LEOS International Conference on. Things can be used. As an analog type magneto-optical element, for example, it is described in Mitsuteru Inoue et al. “Magnophophicnic crystals-a novel magneto-optic material with artificial periodic structures ”, J. Mater. Chem. Vol. 16, pp678-684, 2006. Can be used.

When a MEMS SLM is used, voltage-independent distortion appears as wavefront distortion obtained when no voltage is applied. If V = 0 in equation (1), then φ (V, x, y) = 0, and Φ ₀ = Φ (0, x, y). For this reason, in the Michelson interferometer 80 of FIG. 16, by performing measurement without applying a voltage, it is possible to determine Φ ₀ due to the distortion of the reflecting surface as it is. To create a correction pattern 12 on the basis of [Phi _0. In addition, the voltage-dependent phase modulation characteristics appear as variations in the amount of phase modulation for each pixel when a voltage is applied. Such voltage-dependent phase modulation characteristics can be corrected by the LUT 11 created by the same method as the present embodiment.

  When a voltage is applied to the analog type magneto-optical element, the polarization direction of incident light is rotated. The voltage-independent phase modulation characteristic shows the variation of the rotation of the polarization direction of light measured for each pixel without applying a voltage with the Michelson interferometer 80 of FIG. Further, the voltage-dependent phase modulation characteristic indicates the variation of each pixel in the rotation amount in the polarization direction of the light measured by applying a voltage with the Michelson interferometer 80 in FIG. Therefore, in the Michelson interferometer 80, the correction pattern 12 is created based on the rotation of the polarization direction of the light measured without applying a voltage. Further, the LUT 11 may be created based on the amount of rotation in the polarization direction of light measured by applying a voltage with the Michelson interferometer 80.

  The phase modulation apparatus of the present invention is suitable for use in laser processing, optical tweezers, adaptive optics, various imaging optical systems, optical communication, aspheric lens inspection, pulse waveform control of a short pulse laser, optical memory, and the like.

It is a block diagram which shows the structure of the LCoS type | mold spatial phase modulation apparatus by embodiment of this invention. It is a figure which shows the structure of a LCoS type | mold spatial light modulator. It is a figure which shows the state of a liquid crystal molecule when there is no electric potential difference of the pixel electrode and counter electrode of a LCoS type | mold spatial light modulator. It is a figure which shows the state of a liquid crystal molecule in case the potential difference of the pixel electrode of a LCoS type | mold spatial light modulator and a counter electrode is small. It is a figure which shows the state of a liquid crystal molecule in case the potential difference of the pixel electrode and counter electrode of a LCoS type | mold spatial light modulator is large. It is the figure which showed the example of the LUT map. It is the figure which showed another example of the LUT map. It is the figure which showed another example of the LUT map. It is a figure which shows LUT. It is explanatory drawing explaining the conversion by a D / A circuit. It is the flowchart which showed the method of the phase modulation by the LCoS type | mold phase modulation apparatus of this Embodiment. It is the flowchart which showed the method of setting the minimum value and the maximum value of the voltage driven to a LCoS type | mold spatial light modulator. It is a block diagram which shows the structure of a polarization interferometer. It is a graph which shows the relationship between DA input value and phase modulation amount. It is a graph which shows the relationship between DA input value after setting the minimum value and the maximum value of a use voltage, and a phase modulation amount. It is the flowchart which showed the method of producing a LUT map. It is the flowchart which showed the method of producing LUT. It is a block diagram which shows the structure of a Michelson interferometer. It is the flowchart which showed the method of producing a correction pattern. It is an example of the correction pattern produced based on the procedure shown in FIG. It is a figure which shows the result of having applied phase modulation by applying a LUT and a correction pattern. It is a figure which shows the result of having performed phase modulation, without applying LUT and a correction pattern. It is a figure which shows the structure of the LCoS type | mold spatial light modulator of the example of a change. It is a block diagram which shows the structure of the LCoS type | mold phase modulation apparatus of another modification. It is a block diagram which shows the structure of the LCoS type | mold phase modulation apparatus in another modification. It is a figure which shows LUT containing the information of a correction pattern. It is a figure explaining the thickness of the liquid crystal layer of a LCoS type spatial light modulator. It is a figure explaining the thickness of the liquid crystal layer of a LCoS type space light device, and the inclination of a glass substrate. It is the flowchart which showed the method of producing the LUT map in the example of a change. It is a figure which shows the distortion of the reflective surface of LCoS in a prior art example.

1,100,200 LCoS type phase modulation device 2,120 LCoS type spatial light modulator 31 processing device 35 adder device 36,136 LUT processing device 37 pixel position detection device 32,132 DA circuit 321 drive unit 46 conversion means 48 pixel position Detection means 47 Input value setting means 11 Look-up table 12 Correction pattern 15 LUT map

Claims (9)

A spatial light modulator comprising a plurality of pixels arranged two-dimensionally adjacent to each other, each pixel performing phase modulation on input light in response to application of a drive voltage;
Input value setting means for setting an input value for each pixel;
A reference data map set to associate each pixel with one of a plurality of groups based on a phase modulation characteristic indicating a phase modulation amount in each pixel when a voltage is applied to the pixel ;
A plurality of reference data provided in a one-to-one correspondence with the plurality of groups;
Using the reference data map, for each pixel, specifying means for specifying reference data corresponding to a group associated with the pixel;
Conversion means for converting an input value input to each pixel into a control value with reference to the reference data specified by the specifying means;
A phase modulation apparatus comprising: a driving unit that converts the control value into a voltage value and drives each pixel with the driving voltage having the voltage value. Each pixel can be driven with a voltage value within the operating voltage range,
The drive means converts the control value into a voltage value within a predetermined voltage range set within the operating voltage range, and drives each pixel with a drive voltage of the voltage value,
The predetermined voltage range is a voltage dependency indicating a relationship between a voltage and a phase modulation amount in at least one pixel when a voltage having a voltage value in a predetermined range is applied to at least one pixel of the plurality of pixels. The phase modulation apparatus according to claim 1, wherein the phase modulation apparatus is set based on a characteristic phase modulation characteristic.   Each of the plurality of reference data is a look-up table, and the look-up table includes a plurality of first values that can be taken by an input value, and a phase that indicates the plurality of first values and voltage-dependent phase modulation characteristics. 2. The phase modulation apparatus according to claim 1, wherein a plurality of second values to be taken by the control value are stored in a one-to-one relationship so that the relationship with the modulation amount is a predetermined linear relationship.   The phase modulation apparatus according to claim 1, wherein the reference data map associates pixel position information with reference data corresponding to a group associated with the pixel. The reference data map includes a voltage-dependent phase modulation characteristic indicating a relationship between a voltage and a phase modulation amount in each pixel when a voltage having a voltage value in a predetermined range is applied to each pixel , and a predetermined pixel for a reference pixel . Based on the value indicating the difference between the voltage and the phase modulation amount indicating the relationship between the voltage and the phase modulation amount in the reference pixel when the voltage of the voltage value in the range is applied , the pixel position information, phase modulation device according to claim 1, wherein it is set to so that associates the reference data corresponding to the group associated with the pixel. The reference pixel includes a voltage-dependent phase modulation characteristic indicating a relationship between a voltage and a phase modulation amount in each pixel when a voltage having a voltage value in a predetermined range is applied to each pixel, and the plurality of pixels . The plurality of pixels based on a value indicating a difference between an average of voltage-dependent phase modulation characteristics indicating a relationship between a voltage and a phase modulation amount in each pixel when a voltage having a voltage value in a predetermined range is applied to each pixel. The phase modulation apparatus according to claim 5 , wherein the phase modulation apparatus is selected from among the above. The spatial light modulator is an LCoS spatial light modulator, and the LCoS spatial light modulator has a glass substrate and a silicon substrate,
The reference data map associates pixel position information and reference data corresponding to a group associated with the pixel based on a value indicating distortion of the silicon substrate of the LCoS spatial light modulator. The phase modulation device according to claim 1.   The reference data map is characterized by associating pixel position information with reference data corresponding to a group associated with the pixel based on voltage-independent distortion caused by distortion of the silicon substrate. The phase modulation device according to claim 7.   The reference data map associates pixel position information with reference data corresponding to a group associated with the pixel based on the inclination between the glass substrate and the silicon substrate of the LCoS spatial light modulator. The phase modulation apparatus according to claim 7.

Images