For this Blog (and the next) I will focus on the subject of 3D crosstalk, or frequently called 3D ghosting. This undesired effect occurs when portions of the image intended to be seen by one eye become visible to the other eye. This ‘leakage’ of information between the right and left visual channels may occur with virtually any 3D display technology, but certain display/projection technologies are more prone to have a level of 3D crosstalk that rises to the point of being objectionable for most viewers. For this initial blog (i.e., Part 1) on the subject of 3D crosstalk, I will discuss the principle sources for 3D crosstalk for both active 3D and passive 3D projection system and in Part 2 (i.e., my next blog) I will provide, as an example, results from testing 3D crosstalk on my own JVC DLA-RS40 projector.
With 3D projection systems both the right and left image streams are being displayed on the screen and without the use of 3D glasses the viewer is effectively seeing the 100% crosstalk case, i.e., where both eyes are seeing the full intensity images for both the right and left image streams. For both passive and active 3D projection systems the combination of the projector, projection screen and 3D glasses must work together to allow the viewer’s right eye to only receive the right image stream and the viewer’s left eye to only receive the left image stream. In a perfect 3D projector system where this ideal case would be satisfied, there would be 0.0% 3D crosstalk. However, in the real work there will always be some level or 3D crosstalk and the goal should be to keep it to a low enough level that it is not visible under ordinary circumstances such that moderately critical viewers will not find it visible often enough to be distracting. It must also be noted 3D crosstalk seems to be much more noticeable/objectionable to some viewers than to others. The following 3 pics show a real-world example of 3D crosstalk. The first image is the original left image, the second image is the original right image and the third image is a photo from the screen taken through the left lens of the 3D glasses. If there were no crosstalk present the first and third images should have the same content (note that the 3 images below are framed a little different). However, in the third image below (i.e., the screen shot), a trace amount of the vertical white bar from the right image is leaking through and is visible. Note this may not be very visible on your computer’s display depending on how it is adjusted.
Potential Sources for 3D Crosstalk
For Active 3D systems (either a 3D projection system or a 3D direct view TV) the left and right images are displayed sequentially with the right and left images rapidly alternating on the screen. The companion active shutter 3D glasses must be synchronized with the projector to correctly display only the left images to the left eye and only the right images to the right eye. With Active 3D projection systems there are the following five potential sources for 3D crosstalk:
1. Video Source – It is possible that the 3D video source (e.g., Blu-ray disc, satellite, cable TV, etc.) is outputting a 3D signal with crosstalk between the right and left video channels. This could be the result of an issue anywhere in the video chain from the camera, video processing, video recorder, video encoding, video decoding, etc.
Note: Although it is technically possible to have crosstalk introduced in the 3D signal before it is input to your 3D projector/display this is not typically the case. I have not heard of any confirmed case of a specific Blu-ray 3D disc title to have visible crosstalk in the recording itself. Most consumers having 3D projectors (or 3D TVs) have noted more visible crosstalk when viewing certain Blu-ray 3D movies than with certain other Blu-ray 3D movies and some have incorrectly jumped to the conclusion that the fault lies with the 3D recording itself on the titles prone to have visible 3D crosstalk. Typically in such cases, when only certain movies have objectionable crosstalk, this is because those titles include visual scenes with the characteristics (e.g., high contrast) that their specific 3D projector or 3D TV has the most difficulty in cleanly displaying. For more on this see below.
2. 3D Projector’s Imaging Technology Performance – By this I’m talking about the ability of the projector’s imaging chips (e.g., LCD, LCoS or DLP) to cleanly alternative between the display of the right and left image frames without being effected by the contents of the previously image frame (i.e., the frame intended to only be seen by the other eye). If any trace of that previous frame is still present when the next frame is displayed to the viewer the result can be visible 3D cross talk. LCoS technology is used for 3D projectors from such manufacturers as JVC and Sony, LCD technology is used for 3D projectors from Epson and Panasonic and DLP technology issued by Acer, Optoma, Mitsubishi (arriving soon) and Benq (arriving soon).
Note: LCD and LCoS micro-display chips inherently have a response time required to fully replace one image with the next image. Projector manufacturers using these technologies typically insert a black frame or blanking frame in the video stream between right and left images in an attempt to remove the previous image before displaying the next image. This appears to only be moderately successful since all LCD and LCoS based 3D projectors seen/reviewed to date do appear to have some level of 3D crosstalk resulting from the behavior of their micro-display chips.
3. Screen Characteristics – For active 3D systems the projection screen itself cannot be the source 3D crosstalk. However, for some viewers 3D crosstalk may be more visible/objectionable when the overall image brightness is higher and a higher gain screen will provide an overall brighter image as compared to a lower gain screen.
Note: I’m not suggesting using a low gain screen for 3D projection systems as a means to reduce objectionable 3D crosstalk. Rather I’m just pointing out that some viewers will find the same level of 3D crosstalk less noticeable on dim images as compared to bright images. Using a low lamp mode or a user adjustable iris could also be employed as a means to reduce overall image brightness if you really find that desirable for 3D viewing. However, most consumer 3D projection systems available to date already produce a fairly dim 3D image when used with a moderately large screen and having “too bright” a 3D image is rarely the case.
4. Synchronization between 3D Projector and 3D Glasses –3D is being projected as a sequence of alternating right and left images with a blanking interval between each image. The liquid crystal lenses of the active shutter 3D glasses must be synchronized with the projector such that the right lens only becomes transparent when the right images are displayed, the left lens only becomes transparent when the left images are displayed and these lenses become opaque all other times (i.e., when the image intended for the other eye is being displayed and also during the blanking interval between when the fully formed right and left images are being displayed). If a given lens of the active shutter 3D glassed become transparent too soon or become opaque too late the 3D crosstalk will be the result.
Note: With the increasing availability of 3rd party universal active shutter 3D glasses, differences in the ability of the glasses to synchronize optimally with the projector can be observed. Some such 3D glasses have a means to adjust the synchronization timing parameters to allow the user to select the best setting for use with their specific 3D projector or 3D TV.
5. On/Off ratio of the Active Shutter 3D Glasses – In the ideal world when the shutter lens of the 3D glasses is supposed to be transparent it would be perfectly clear and when it is supposed to be opaque it would pass no light at all. In the real world however the liquid crystal lenses never become perfectly transparent nor opaque. It is the inability of a lens to become full opaque that can lead to 3D crosstalk when even a very small percentage of the light is still being transmitted through the lens.
Note: It appears that 3D glasses being sold by the major projector manufacturers are improving in their ability to block light in the opaque state. Some inexpensive third party 3D glasses may or may not be adequate in this characteristic.
For Passive 3D projection systems there are two different technologies commonly used to visually separate the right and left image streams being seen by the viewers. I will be getting into passive 3D projection technologies and systems in a future blog. The more common technique for passive 3D projection uses a different orientation for the polarization of the projected light associated with the right and left images streams (e.g., used by IMAX and RealD in commercial cinemas). The only consumer oriented single projector using polarization for 3D is the LG CF3D (review is HERE). High-end dual projector setups using polarization for 3D are available from such manufacturers as Runco.
The less common passive 3D technique uses narrow bandpass color filtering for the primary colors associated with the right and left image streams (e.g., used by Dolby in commercial cinemas). This 3D technique is sometimes referred to as “Interference Filter Technology”. In either case, both the right and left images streams are typically (for consumer systems) displayed simultaneously, rather than sequentially, and the passive 3D glasses have lenses that must pass the desired image stream and effectively block the unwanted image stream. The following 4 potential sources for 3D crosstalk apply as noted below to one or both of these passive 3D approaches:
1. Video Source – same as described above for active 3D systems.
2. 3D Projector’s Filter Technology/Performance – Passive 3D projection systems must use either polarizing filters or color bandpass/interference filters. In both cases there are of multiple versions being used by different manufacturers. In the case of polarizing filters, these can be either linear or circular polarizing and in the case of color bandpass/interference filters, the number of bands and the specific set of wavelengths used can differ between manufacturers.
Note: None of these passive 3D technologies can be considered best or worst. Each have their positives and negatives. For example, in the commercial cinemas the IMAX 3D systems use linear polarization while RealD uses circular polarization. IMAX claims linear is better as it offers better rejection of the unwanted images (i.e., less crosstalk), but this is only true if you are viewing with the glasses perfectly level with the screen. If you tilt your head, or the glass, the crosstalk will increase. On the other hand with circular polarized projectors/glasses the crosstalk is not increased if the viewers fail to hold their head perfectly upright. With any passive technology the characteristics of the 3D related filters used by the projector(s) can have an impact on the level of 3D crosstalk.
3. Screen Characteristics – Of all 3D projection technologies, those passive 3D projection systems using polarization as the means of separating the right and left video streams are by far the most demanding on the projection screen characteristics.
Note: This is because the projection screen must retain the vast majority of the polarization in the image it reflects. Projection screens intended for use with such polarized passive 3D projection systems typically use a silver surface and such screens from the major screen manufacturers (such as Stewart and Da-lite) are specified to retain greater than 99% of the polarization. For passive 3D projection systems using bandpass/inference color filters, crosstalk levels are normally not especially sensitive to screen characteristics.
4. Passive 3D Glasses Characteristics – Passive 3D glasses are typically very low priced and that is one of the selling points as compared to the much more complex 3D active shutter glasses. With passive 3D glasses each lens is a filter and the quality of that filter can determine how well it will perform. The differences between mediocre and outstanding optical performance from such 3D glasses will usually be in terms of light loss and rejection of the unwanted images (i.e., 3D crosstalk levels). Quality 3D glasses using either linear or circular polarization are probably possible at a lower price point than those using the color bandpass/interference filters.
Ghostbusting – How Good is Good Enough
As noted in the above discussion the level of 3D crosstalk, or 3D Ghosting, what viewers find acceptable will vary from viewer to viewer. Some people have compared this to the sensitivity of individual viewers to the so called “rainbow effect” (RBE) seen on single chip DLP projectors. Some viewers complain about the excessive RBE even with those DLP projectors having 6X color wheels (i.e., this highest speed currently available) while other viewers say can do not see RBE even on DLP projectors with slower 4X or even 2X color wheels. There is some published information that indicates that the 3D crosstalk level should be always kept to less than 1% of the peak white level of the primary image (i.e., the ghost image should be no more than 1% as bright as the main image). With a 1% crosstalk level, 3D ghosting will be visible to most viewers during some higher contrast scenes and some viewers will consider this level of crosstalk too high. A 0.5% crosstalk level has been suggested by some as the minimum target level for any consumer 3D TV or projection system as most viewers will find the visible 3D crosstalk acceptable for casual viewing. With 0.5% crosstalk some viewers may not notice 3D ghosting, but more critical viewers will likely see some ghosting at least occasionally in most programs. More ideally the crosstalk level would be 0.1% or less. In this case the peak white level of the primary image would be at least 1000 times brighter than any ghost image. By comparison it appears that the better commercial cinemas using either linear or circular polarization 3D passive technology can achieve crosstalk levels of approximately 0.5% (some are worse or even much worse mainly due to use of less-than-ideal projection screen materials).
This discussion will continue with Part 2 in my next blog where I will include results from testing of my own JVC DLA-RS40.