Home Theater and Projectors

More 3D Crosstalk Testing – JVC and Epson

This Blog is a follow-up to my two earlier blogs of December 14th and December 19th on 3D crosstalk (i.e., 3D ghosting).  In the second of those previous blogs I presented some results for 3D crosstalk measurements taken for my own JVC DLA-RS40 projector.  I have now taken one critical additional measurement for that projector plus also present results from my testing of an Epson Home Cinema 5010 projector.  This is physically the same Epson projector Art used for his review HERE at Projector Reviews.  I have done a visual comparison of the 3D crosstalk levels using full resolution 1080p 3D material (input to the projector at 24Hz using the Blu-ray 3D standard frame packing signal format) and also half resolution 3D programming via DirecTV (provided from my DirecTV HD-DVR in 1080i/60 3D side-by-side format).   For my 3D crosstalk measurements I used a Blu-ray 3D test disc which I have created (i.e., in full resolution 1080p at 24Hz in the frame packing format, as per the Blu-ray 3D standard).

Having now owned my JVC DLA-RS40 for one year and using both Blu-ray 3D and the handful of half resolution 3D channels on DirecTV, I now have a fairly good understanding how frequently visible 3D crosstalk is on this JVC projector.  The recently released DLA-RS45 has replaced the DLA-RS40, but reports from owners of the RS45 that have directly compared this new model with last year’s RS40 suggest similar levels of 3D crosstalk.  However, some unit-to-unit variations have been noted.  As background my DLA-RS40 has approx. 100 hours on the lamp and the lumens from the projector has only deceased slightly since I measured it at 50 hours on the lamp.  The Epson projector was the evaluation unit provided by Epson and is physically the same projector that was evaluated by Art for his review (HERE) and therefore I do not intend to cover the ground already discussed in depth by Art.   For this blog my focus will be 3D crosstalk (ghosting) and a direct comparison with the crosstalk performance of my DLA-RS40.


Measurement Approach

I first calibrated each projector for reference black and reference white level.  I used a Panasonic Blu-ray 3D disc player connected via HDMI to a HDMI input on the projector.  Both projectors were run in 3D mode for at least 30 minutes before the measurements were performed and I used two test patterns from my Blu-ray 3D test disc for the following measurements.

White-on-Black Test Pattern:  The first pattern uses a solid black screen for the left eye view and vertical white bar on a black background for the right eye view.  The ideal projector with no crosstalk would project an image that appears (or measures) uniformly black when viewed through the left lens of the 3D active shutter glasses.  Any visible crosstalk with this test pattern will appear as a dark grey vertical bar on a black background when viewed through the left lens of the 3D active shutter glasses.

White-on-Black 3D Test Pattern


Black-on-White Test Pattern:  This second test pattern uses a solid white screen for the left eye view and a vertical black bar on a white background for the right eye view.  The ideal projector with no crosstalk would project an image that appears (or measures) uniformly white when viewed through the left lens of the 3D active shutter glasses.  Any visible crosstalk with this test pattern will appear as a light grey vertical bar on a white background when viewed through the left lens of the 3D active shutter glasses.

Black-on-White - Right Test Image

As noted in my earlier blogs, in order for a 3D projector to provide a 3D image that is free of visible crosstalk for most viewers on most of the time, the crosstalk level should be kept below 0.5% and in the ideal world crosstalk should be kept below 0.1% to be considered “crosstalk free.”  Now on to the measurement results for the JVC and Epson projectors.


JVC DLA-RS40 Crosstalk Levels

All of the JVC 3D projectors display 24 Hz 3D source material (such the test patterns) at 96 Hz total, or 48 Hz per eye.  The measurement results for the White-on-Black test pattern was previously presented in my earlier blog of Dec. 19th (HERE).  For this test pattern the RS40 showed a crosstalk level of approx. 0.7%.  This level of crosstalk does result in 3D crosstalk/ghosting that is occasionally visible where bright objects appear on a dark background.  However, visually the JVC 3D projectors seem to have more problems in handling dark objects against a light background.  The Black-on-White test pattern is intended to show issues in this latter case.  When I displayed the Black-on-White test pattern on my DLA-RS40, the black bar (that should ideally be visible only to the right eye) was clearly visible as light grey bar to the left eye.  My measurement of the light level (measured with a lux meter using the light directly from the projector and not off the screen) showed as surprisingly high (approx.) 17% level of crosstalk.  By this I mean the projected image when measured through the left lens of the JVC 3D active shutter glasses was 17% dimmer in the location of where the black bar was being displayed for the right eye.  Given this high level of 3D crosstalk it is no wonder that the DLA-RS40 has visible crosstalk in scenes with dark objects appearing against a light color background.


Epson Home Cinema 5010 Crosstalk Levels

I only required a few minutes of viewing some scenes from a couple of Blu-ray 3D discs plus also a few minutes of viewing a college football game (replay) from DirecTV (using half resolution 3D format) to realize the Epson has much less crosstalk than my JVC DLS-RS40.  The measurement results were only made with the full resolution 1080p/24 3D signal format, but based on my visual observations the different is even more to the advantage of the Epson when the input signal is in the half resolution 1080i/60 side-by-side 3D format.

For the White-on-Black test pattern the Epson 5010 showed a crosstalk level of approx. 0.29%.  I measured the same result in both the “3D Dynamic” and the “3D Cinema” modes.  This level of crosstalk results in 3D ghosting that is only rarely visible where bright objects appear on dark background.  The Black-on-White test pattern is where the JVC projector has major issues.  For this latter test pattern the Epson measured a more respectable approx. 3%, as compared to the JVC’s 17%.  Such a 3% level will result in 3D ghosting that is visible on some scenes where dark objects appear on a light background, but not nearly as often or as noticable as with the JVC DLA-RS40.


Conclusions from the 3D Crosstalk Measurements

3D crosstalk with both LCoS and LCD projectors most typically results from less than ideal response time from the projector’s micro-display chips.  It appears both the JVC (LCoS technology) and Epson (LCD technology) 3D projector’s micro-display chips have a slower response time for pixel transitions going from a full black state to a full white state than for going the other direction (i.e., from full white to full black).  While I have not tested enough LCoS and LCD based 3D projectors to conclusively draw a general conclusion, this characteristic may be inherent with these technologies.  As for the JVC DLA-RS40 vs. Epson 5010 3D performance, clearly the Epson has significantly less 3D crosstalk which results in less visiable 3D ghosting.  One caveat that I must add is to point out that the above results and the following observations were based on a single sample of each of these projectors.  However, I have no reason to believe the measured and observed performance is not representative of these projector models but some unit-to-unit variations are to be expected.


3D Viewing Observations/Comparisons

I started off with viewing 3D programming delivered by DirecTV.  All the 3D material I viewed from DirecTV was provided in the half resolution 1080i side-by-side format.  Both projectors display this type of signal and an effective rate of 60 Hz per eye (or 120 Hz total).  This is true even for the Epson which claims a 480 Hz refresh, but for 3D the same video frame is displayed 3 consecutive times followed by one blanking frame which results in a 480/4 = 120 Hz total effective rate.  The JVC DiLA projectors appears to have a response time for the LCoS micro-displays that is marginal at 96 Hz and in my opinion becomes really a problem for 120 Hz operation.  As a result the visible crosstalk with such DirecTV programming is obvious with many, or even most, scenes depending on the specific program material.  While the Epson 5010 was not totally ghost free with the DirecTV programs (see photos below), the visible 3D ghosting appeared only on the occasional scene and overall was vastly superior to the JVC DLA-RS40 for viewing programming from DirecTV.  The following photos represent some more-or-less worst case examples of 3D ghosting on these JVC and Epson projectors.

JVC DLA-RS40 Ghosting Example for 1080i Side-by-Side 3D Source

Epson 5010 Ghosting Example from 1080i Side-by-Side 3D Source

Epson 5010 - Another example of 3D Ghosting with 1080i SBS Source

For viewing Blu-ray 3D programs I used one movie and a couple of IMAX documentaries.  From my experience some Blu-ray 3D titles overall appear very good on the JVC DLS-RS40 with only very modest 3D ghosting visible on the occasional scene.  The Avatar Blu-ray 3D disc is one example that looks quite good on the JVC 3D projectors.  However, some other Blu-ray 3D titles do result in more frequent cases of visible 3D ghosting with the JVC 3D projectors.  Even though overall the JVC DLA-RS40 is acceptable for displaying Blu-ray 3D discs, the Epson 5010 is even better.  I noticed in some scenes that caused visible 3D ghosting with the JVC DLA-RS40 the Epson 5010 either had noticeably less visible 3D ghosting or in many cases no visible ghosting.

For either the DirecTV or Blu-ray 3D sources, I did not observe any case where there was visible ghosting with the Epson but not with the JVC.  So the bottom line for my observations is while the JVC is acceptable to me for viewing most Blu-ray 3D programs, the Epson is noticeably better, at least in terms of visible 3D ghosting.   With the Epson 5010 I did not find any program, either using DirecTV or Blu-ray as the 3D source, where the level of visible 3D ghosting was to the level that I would consider it a major issue.  However, for the JVC visible 3D ghosting is a major issue with many of the DirecTV programs, while the same JVC projector only displays really objectionable 3D ghosting on occasional scenes from most Blu-ray 3D discs that I have viewed.


For my next Blog I plan to start the discussion on Passive 3D projection systems



News And Comments