Figure 4 - uploaded by Pierre Boher
Content may be subject to copyright.
Source publication
In this paper we apply viewing angle and imaging polarimetry to characterize a polarization based stereoscopic 3D display. Viewing angle multispectral polarization analysis is first performed using a Fourier optics instrument. A vertical polarization modulation is observed and explained. Main imperfections come from a polarization degree that reach...
Contexts in source publication
Context 1
... 3D displays have the same intrinsic method to provide 3D perception: the main idea is always to provide two different images in the left and right eye of the observer to create the depth perception. The two most common solutions are auto-stereoscopic 3D displays and polarization based 3D displays. Up to now few studies have been devoted to the optical characterization of such displays. Recently we have proposed a characterization method for auto-stereoscopic 3D displays based on ultrahigh angular resolution Fourier optics measurement and computation in the observer space [1-3]. We have also been interested in polarization based 3D displays and we have used our viewing angle measurement instruments to characterize such displays [4-8]. A common approach has been proposed to analyze both types of display computing the same types of parameters and in particular the 3D contrast [4]. We are more interested here in understanding the main artifacts of a polarization based stereoscopic 3D display both versus angle and in terms of homogeneity. In depth analysis with viewing angle multispectral instrument allows analyzing the origin of the troubles. Imaging polarization is also used to verify the predictions of viewing angle instrument in terms of 3D contrast and best observer location positions. In addition, “polarization Mura” defects related to the phase difference films are observed. We use the Fourier optics instrument EZContrastMS that provides the full viewing angle cone in one measurement rapidly and accurately. Radiance full viewing angle patterns at different wavelengths using 31 band-pass filters from 400nm to 700nm every 10nm are obtained. The system includes also three polarizers at different orientations (0, 45 and 90°) and two wave- plates at different orientations (45 and 135°). Seven measurements with the various polarization configurations are made automatically for each band-pass filter from which the Stokes vectors are derived for each incidence angle, azimuth angle and wavelength. Multispectral viewing angle polarization measurements have been performed for right view ON and left view OFF and the opposite on our polarization based 3D display. On the viewing angle pattern measured at 549nm (cf. figure 1), we notice a vertical modulation of the ellipticity and of the polarization degree. This vertical modulation is due to a perspective effect induced by the location of the retarder film with regards to the liquid crystal cell. In this display, the film is located on the display top surface (cf. figure 2). The thickness of the top glass modulates the polarization of one view along vertical from right circular state to left circular state. So only one restricted part of the space in front the display is available for correct 3D perception. It is of course possible to suppress this modulation by including the phase difference film directly inside the liquid crystal cell structure. The vertical modulation of the polarization restricts the space in front of the display from which stereoscopic images are seen correctly. For the best working distance it is then necessary to adjust the light emission of the display to ensure that each region of the display will be seen correctly from the same location. In the display under investigation a slight vertical shift of the region circularly polarized is applied from the top to the bottom of the display. The polarization state varies strongly with wavelength. In particular, the polarization degree does not exceed ~90% even at the best angles and the best wavelength (cf. figure 3). This is especially important since the unpolarized component cannot be stopped by the circular polarizers included in the glasses and the 3D contrast will be therefore strongly reduced. Another source of imperfection is the ellipticity which never reaches its best value (circular state at ±45°)(cf. figure 4). It is in addition extremely dependent of the wavelength as reported in figure 4. The polarization state is nearly circular at 530nm but drop down rapidly in the blue and red regions. This is due to the phase shift behavior in 1/ λ of a simple single layer retarder sheet. It can be improved using a more complex stack of birefrigent layers but increases once again the complexity of the display structure. The two views are also not completely symmetric. The degree of polarization of the light at the working incidence angle (~9°) is slightly better for right eye than for left eye. The origin of the effect is probably due to the phase difference film quality. Microretarder films are generally made from two homogeneous retardation films with a uniform half wave phase retardation state. Through laser exposure or thermal printing, the original stretching direction of one film become random and its polarization state changed from half wave to zero but parasitic diffusion is also occurring. Luminance viewing angle measurements are also performed with an ELDIM EZContrastXL88 Fourier optics instrument including left and right eyes filter glasses (GL and GR). Because of the vertical polarization modulation, three locations are measured on the display surface (center, top center, bottom center). In each case and for each filter four measurements are realized: left view ON and right view OFF (LWRK), left view OFF and right view ON (LKRW), left and right views ON (LWRW) and left and right views OFF (LKRK). The stereoscopic quality of the 3D display for an observer is directly related to the capacity to see clearly the correct images in his right and left eyes. In case of stereoscopic displays, the two contrasts associated to each eye must be calculated using the two sets of measurements obtained with GL and GR filters. We use the following equations: ( θ R , φ R ) and ( θ L , φ L ) are the right and left eye positions in polar coordinates with regards to the measurement location. Y GLKLRW and Y GRKLRW are the luminance for white view on right eye and black view on left eye using GL and GR filters respectively. Y GLLWRK and Y GRLWRK are the luminances for black view on right eye and white view on left eye using GL and GR filters respectively. Y GLRKLK and Y GRRKLK are the luminance for black view on both eyes. It is then possible to combine the two contrasts to get a combined contrast for the observer using: We take the product and not the sum because a good 3D quality requires a good contrast for left and right eyes simultaneously. The square maintains the dimensionally of the quantity as a contrast. These calculations are very similar to those applied to auto-stereoscopic 3D displays but each contrast is obtained separately using the two GL and GR filters [1-3]. When different locations are analyzed simultaneously, the different contrasts for each eye are calculated, and the minimum value of these contrasts for the different locations is taken using the previous equation and: M is the number of measurement locations. 3D Contrast is calculated in the same way as previously. Angular contrasts for central location are reported in figure 5. These diagrams show sharp emissive horizontal bands at an angle around 9° to the top. The maximum of contrast is around 13 for right eye and 9 for left eye. These values are low compared to those generally measured on auto-stereoscopic 3D displays. In addition the contrast is not symmetric and clearly worse on one eye than on the other in agreement with the multispectral measurements reported previously. The 3D contrast has been calculated using 3 measurement locations for a given volume in front of the display (1000x1000x2000mm in figure 6). The Qualified Binocular Viewing Space (QBVS) is composed of one horizontal flat volume titled by about 9° vertically as shown in the cross section along vertical plane. The best working distance is around 650mm as expected for a 19’ computer display. Viewing angle measurements are useful to compute the 3D and standard properties of a display at given locations on its surface. They cannot be used to detect local fluctuations of the light emission on the display surface. To study these inhomogeneities we use a video-colorimeter UMaster that is also capable to measure the polarization state of light at some wavelengths. UMaster includes three polarizers at different orientations (0, 45 and 90°) and two wave-plates at different orientation (45 and 135°). The system makes automatically seven measurements with different polarization configurations and computes the polarization parameters and the Stokes vectors. The wavelength is defined by a band pass filter (550nm in the following measurements). Imaging polarization is made for left view ON and right view OFF and the opposite. One measurement for left and right eye view is reported in figure 7. UMaster is located at 102cm from the display surface to measure the entire display surface and about 9° from the normal incidence. The light polarization state is nearly circular ...
Context 2
... 3D displays have the same intrinsic method to provide 3D perception: the main idea is always to provide two different images in the left and right eye of the observer to create the depth perception. The two most common solutions are auto-stereoscopic 3D displays and polarization based 3D displays. Up to now few studies have been devoted to the optical characterization of such displays. Recently we have proposed a characterization method for auto-stereoscopic 3D displays based on ultrahigh angular resolution Fourier optics measurement and computation in the observer space [1-3]. We have also been interested in polarization based 3D displays and we have used our viewing angle measurement instruments to characterize such displays [4-8]. A common approach has been proposed to analyze both types of display computing the same types of parameters and in particular the 3D contrast [4]. We are more interested here in understanding the main artifacts of a polarization based stereoscopic 3D display both versus angle and in terms of homogeneity. In depth analysis with viewing angle multispectral instrument allows analyzing the origin of the troubles. Imaging polarization is also used to verify the predictions of viewing angle instrument in terms of 3D contrast and best observer location positions. In addition, “polarization Mura” defects related to the phase difference films are observed. We use the Fourier optics instrument EZContrastMS that provides the full viewing angle cone in one measurement rapidly and accurately. Radiance full viewing angle patterns at different wavelengths using 31 band-pass filters from 400nm to 700nm every 10nm are obtained. The system includes also three polarizers at different orientations (0, 45 and 90°) and two wave- plates at different orientations (45 and 135°). Seven measurements with the various polarization configurations are made automatically for each band-pass filter from which the Stokes vectors are derived for each incidence angle, azimuth angle and wavelength. Multispectral viewing angle polarization measurements have been performed for right view ON and left view OFF and the opposite on our polarization based 3D display. On the viewing angle pattern measured at 549nm (cf. figure 1), we notice a vertical modulation of the ellipticity and of the polarization degree. This vertical modulation is due to a perspective effect induced by the location of the retarder film with regards to the liquid crystal cell. In this display, the film is located on the display top surface (cf. figure 2). The thickness of the top glass modulates the polarization of one view along vertical from right circular state to left circular state. So only one restricted part of the space in front the display is available for correct 3D perception. It is of course possible to suppress this modulation by including the phase difference film directly inside the liquid crystal cell structure. The vertical modulation of the polarization restricts the space in front of the display from which stereoscopic images are seen correctly. For the best working distance it is then necessary to adjust the light emission of the display to ensure that each region of the display will be seen correctly from the same location. In the display under investigation a slight vertical shift of the region circularly polarized is applied from the top to the bottom of the display. The polarization state varies strongly with wavelength. In particular, the polarization degree does not exceed ~90% even at the best angles and the best wavelength (cf. figure 3). This is especially important since the unpolarized component cannot be stopped by the circular polarizers included in the glasses and the 3D contrast will be therefore strongly reduced. Another source of imperfection is the ellipticity which never reaches its best value (circular state at ±45°)(cf. figure 4). It is in addition extremely dependent of the wavelength as reported in figure 4. The polarization state is nearly circular at 530nm but drop down rapidly in the blue and red regions. This is due to the phase shift behavior in 1/ λ of a simple single layer retarder sheet. It can be improved using a more complex stack of birefrigent layers but increases once again the complexity of the display structure. The two views are also not completely symmetric. The degree of polarization of the light at the working incidence angle (~9°) is slightly better for right eye than for left eye. The origin of the effect is probably due to the phase difference film quality. Microretarder films are generally made from two homogeneous retardation films with a uniform half wave phase retardation state. Through laser exposure or thermal printing, the original stretching direction of one film become random and its polarization state changed from half wave to zero but parasitic diffusion is also occurring. Luminance viewing angle measurements are also performed with an ELDIM EZContrastXL88 Fourier optics instrument including left and right eyes filter glasses (GL and GR). Because of the vertical polarization modulation, three locations are measured on the display surface (center, top center, bottom center). In each case and for each filter four measurements are realized: left view ON and right view OFF (LWRK), left view OFF and right view ON (LKRW), left and right views ON (LWRW) and left and right views OFF (LKRK). The stereoscopic quality of the 3D display for an observer is directly related to the capacity to see clearly the correct images in his right and left eyes. In case of stereoscopic displays, the two contrasts associated to each eye must be calculated using the two sets of measurements obtained with GL and GR filters. We use the following equations: ( θ R , φ R ) and ( θ L , φ L ) are the right and left eye positions in polar coordinates with regards to the measurement location. Y GLKLRW and Y GRKLRW are the luminance for white view on right eye and black view on left eye using GL and GR filters respectively. Y GLLWRK and Y GRLWRK are the luminances for black view on right eye and white view on left eye using GL and GR filters respectively. Y GLRKLK and Y GRRKLK are the luminance for black view on both eyes. It is then possible to combine the two contrasts to get a combined contrast for the observer using: We take the product and not the sum because a good 3D quality requires a good contrast for left and right eyes simultaneously. The square maintains the dimensionally of the quantity as a contrast. These calculations are very similar to those applied to auto-stereoscopic 3D displays but each contrast is obtained separately using the two GL and GR filters [1-3]. When different locations are analyzed simultaneously, the different contrasts for each eye are calculated, and the minimum value of these contrasts for the different locations is taken using the previous equation and: M is the number of measurement locations. 3D Contrast is calculated in the same way as previously. Angular contrasts for central location are reported in figure 5. These diagrams show sharp emissive horizontal bands at an angle around 9° to the top. The maximum of contrast is around 13 for right eye and 9 for left eye. These values are low compared to those generally measured on auto-stereoscopic 3D displays. In addition the contrast is not symmetric and clearly worse on one eye than on the other in agreement with the multispectral measurements reported previously. The 3D contrast has been calculated using 3 measurement locations for a given volume in front of the display (1000x1000x2000mm in figure 6). The Qualified Binocular Viewing Space (QBVS) is composed of one horizontal flat volume titled by about 9° vertically as shown in the cross section along vertical plane. The best working distance is around 650mm as expected for a 19’ computer display. Viewing angle measurements are useful to compute the 3D and standard properties of a display at given locations on its surface. They cannot be used to detect local fluctuations of the light emission on the display surface. To study these inhomogeneities we use a video-colorimeter UMaster that is also capable to measure the polarization state of light at some wavelengths. UMaster includes three polarizers at different orientations (0, 45 and 90°) and two wave-plates at different orientation (45 and 135°). The system makes automatically seven measurements with different polarization configurations and computes the polarization parameters and the Stokes vectors. The wavelength is defined by a band pass filter (550nm in the following measurements). Imaging polarization is made for left view ON and right view OFF and the opposite. One measurement for left and right eye view is reported in figure 7. UMaster is located at 102cm from the display surface to measure the entire display surface and ...
Similar publications
Eclipsing binary is an edge-on binary system where the components appear to be eclipsing one another. Eclipsing binaries therefore give a lot of information on the systems’ behavior, one of them from their period change. O-C diagram exhibits the deviation of the observed time of minimum from the calculated ones and its analysis will reveal the resp...
Citations
... 39 Studies of applying viewing angle and viewing polarimetry to characterize a polarization-based S3D display are among the studies performed in such imaging studies. 40 In time-sequential imaging (active S3D imaging); the image pair is sequentially displayed in time-sequential screens. The right and left eyeglasses are darkened in turn by the glasses working simultaneously with the screen. ...
The Ministry of National Education of Turkey equipped all schools with interactive boards (IBs) within the scope of the FATİH project. The need for appropriate materials has arisen for the effective use of IBs in educational activities. Three‐dimensional (3D) technologies are also among these materials. It has been determined in the study that IBs have the necessary hardware to run stereoscopic 3D (S3D) training materials, but the panel has not got an S3D imaging feature. Therefore, only the anaglyph S3D imaging method can be applied to IBs. Thus, an anaglyph S3D training material was prepared for the interaction of the skeletal system and interactive 3D material design for IBs with its effects in education was investigated. Besides, S3D imaging methods with features of IBs were examined. A Likert‐type scale was developed to measure the usability of the training material on IBs and the material was evaluated by 20 experts. The data were analyzed by the SPSS statistical program and the results were interpreted. According to the results, educational material seems to be positive in terms of image characteristics, content, navigation, and ease of use, font sizes were moderate for readability, the feedback process and the help menu were moderately effective. An anaglyph S3D training material was prepared for the interaction of the skeletal system and interactive 3D material design for interactive boards (IBs) with its effects in education was investigated. A Likert‐type scale was developed to measure the usability of the training material on IBs by 20 experts and the results were interpreted. According to the results, educational material seems to be positive in terms of image characteristics, content, navigation, and ease of use, font sizes were moderate for readability, the feedback process and the help menu were moderately effective.
... It can also analyze the polarization state of the light with additional polarizers and wave-plates inside the instrument. This feature has for example being used with great success for the characterization of passive glass 3D displays [17][18][19]. ...
Different types of optical characterizations, have been applied to OLED displays: multi-spectral viewing angle, multi-spectral imaging and response time. The interest of the spectral information is that the exact shape of the light emission for each sub-pixel can be precisely determined and analyzed. Wavelength fluctuations versus angle give direct information on the OLED structure of each sub-pixel. Imaging measurements allow to detect wavelength shift on the surface of the displays probably related to thickness non-uniformities. Local radiance fluctuations from one pixel to the other are more related to driving problems due to the dispersion of the properties of the driving TFTs. A new generation of temporal instrument is also used to measure response times shorter than 10µs. In this case collecting light coming from one pixel or one line of pixels is mandatory to avoid averaging.
... It can also analyze the polarization state of the light with additional polarizers and wave-plates inside the instrument. This feature has for example being used with great success for the characterization of passive glass 3D displays [13][14][15]. ...
OLED displays exhibit luminance fluctuations and color shifts that can be sensitive to human eye in particular conditions. Using viewing angle and imaging multispectral measurements we show that color shifts are generally related to the multilayered structure of each sub-pixel. Interference fringes result in angular variations while thickness variations result in surface non-uniformities.
... The new system has a spot size of 1mm and a maximum angular aperture of 80°. The spectral information is obtained using different automated wheels with 31 band pass filters regularly distributed in the visible range near the imaging sensor in the same way as the laboratory version introduced in 2008 [13] and used with great success for the characterization of passive glass 3D displays [14][15][16]. Near the sensor, the optical design is quasi-telecentric in order to avoid incidence angle variations within the field of view. This is especially important since band-pass filters are highly dependent on the incidence angle. ...
A new generation of Fourier optics multispectral instruments that allow rapid full diffused or collimated beam spectral BRDF measurements is presented. Light detection is made simultaneously at all angular locations including the illumination direction. Backscattering effect in the fields of cosmetics and parasitic reflection of mobile displays are reported as examples
... We have shown that this type of measurement is especially helpful for LCD characterization56 and in particular for passive glass 3D displays78. Polarization imaging has also been introduced recently by ELDIM for the characterization of displays and their components9101112131415. In the present paper, after a presentation of the different measurement instruments and their capacities, we have selected some experimental examples to enlight the interest of measuring the polarization state in addition to standard light measurements. ...
Optical characterization of all types of displays is mandatory to quantify the different imperfections and compare the different technologies. Standard characterization is focused on luminance and color measurements of the light emitted by the displays. Viewing angle measurements devoted to the angular characterization and homogeneity measurements are the most frequently used. In the present paper we demonstrate that the same type of measurements extended to the polarization analysis of the light emitted by the display can be a powerful tool in order to understand the origin of many imperfections. The present paper is composed of two main parts: first the different measurement tools for viewing angle and imaging characterization are presented and the method used to analyze the polarization state of the light is explained. The second part gives details on some examples of applications for the characterization of display components (brightness enhancement film, backlight), the measurement of the crystal cell rotation inside a standard LCD display, and the characterization of a passive 3D TV
Optical characterization of all types of displays is mandatory to quantify the different imperfections and compare the different technologies. Standard characterization is focused on luminance and color measurements of the light emitted by the displays. Viewing angle measurements devoted to the angular characterization and homogeneity measurements are the most frequently used. In the present paper we demonstrate that the same type of measurements extended to the polarization analysis of the light emitted by the display can be a powerful tool in order to understand the origin of many imperfections. The present paper is composed of two main parts: first the different measurement tools for viewing angle and imaging characterization are presented and the method used to analyze the polarization state of the light is explained. The second part gives details on some examples of applications for the characterization of display components (brightness enhancement film, backlight), the measurement of the crystal cell rotation inside a standard LCD display, and the characterization of a passive 3D TV.
The optical properties in the patterned retarder system are discussed and calculated. The general equation is derived which helped to build a platform simulating the optical performance of patterned retarder 3D display. With this platform, the designers can optimize parameters to improve viewing quality and drastically save time and cost.
Optical characterization of all types of displays is mandatory to quantify the different imperfections and compare the different technologies. Standard characterization is focused on luminance and color measurements of the light emitted by the displays. Viewing angle measurements devoted to the angular characterization and homogeneity measurements are the most frequently used. In the present paper we demonstrate that the same type of measurements extended to the polarization analysis of the light emitted by the display can be a powerful tool in order to understand the origin of many imperfections. The present paper is composed of two main parts: first the different measurement tools for viewing angle and imaging characterization are presented and the method used to analyze the polarization state of the light is explained. The second part gives details on some examples of applications for the characterization of display components (brightness enhancement film), the measurement of the crystal cell rotation inside a standard LCD display, and the characterization of a passive 3D TV
One commercial passive glass 3D TV is characterized using imaging video-luminance-meter across one of the glass filters. Crosstalk homogeneity on the entire surface of the display is deduced. Results are correlated to those obtained locally using Fourier optics viewing angle instrument. We demonstrate in particular that the local crosstalk imperfections are linked to an imperfect alignment of the quarter wave plate film with regards to the pixel structure
