Warranty: 3 years parts and labor
Summary: Pixel perfect 1080p24 for the videophile performance enthusiast but there's a catch
Review Note: Two different firmware versions of the product were reviewed and tested. This review reflects the current version and differences between them are noted when applicable. To update a projector to the current firmware requires an exchange by BenQ. The new version replaces USER 2 and USER 3 presets with ISF Day and ISF Night on the remote and is noted in some portions of the technical section. After the final conclusion, you will find a quick synopsis of the main differences between the two versions.
1080p DLP front projection is slowly starting to dribble into the market but pricing has been high. BenQ is providing the W10000 for the street price of just under $6000, providing a full 1920x1080 DLP Dark Chip3 DMD far closer in price range with other recently released 1080p technology at $5k and below.
As noted, this is a full 1920x1080 chip and does not use wobulation. The advantage here is the potential for a pixel perfect response. Bear in mind this also means pixels are going to be more visible but whether or not that is a problem is a matter of viewing distance. This is a single chip display utilizing an 8 segment color wheel, which also means rainbows are possible. The "rainbow effect", as its known, is a byproduct of using a color wheel to rapidly flash alternating red, green, and blue picture elements to the screen. For a small segment of the population, this can create a visible "rainbow" artifact when projected content features bright objects on a dark background (movie credits, for example). While greatly reduced over the years by adding more sections and increasing rotational speed, they are still perceptible by some viewers.
Pictures are taken with a digital camera, which has limitations of its own inducing artifacts that are not there. The main purpose of the pictures is to provide a reference for the review, regardless of quality, and provide a fair impression of actual performance as compared to other pictures to which they may be compared.
The focus uniformity problem was brought to the attention of BenQ and the projector was exchanged. The replacement had near perfect focus uniformity with an even response left to right. Extreme left and right were ever so slightly de-focused but even with a PC source, this was a case of splitting hairs.
Back to the zoom for a moment; the 1.15 range is quite small indeed and could be a limitation for some applications. I therefore suggest that you see this as more of a tweak adjustment to set size and overscan for your application, rather than a means for convenient placement or use in a zoomed 2.35 aspect application. If you are considering the W10000 as a replacement projector, then check the online owners manual for screen size and installation logistics to make sure it can fit wherever your present mount is.
At 2.8 screen heights none could be found with HD material or from the PC. Even when setting up the projector to fill out my 2.35 aspect screen, none where there. Testing reveals that at about 2 screen heights, pixels begin to become noticeable, so eagle eyed viewers should be safe at 2.5.
The W10000 provided a reference response for this test for both analog component video and HDMI, maintaining similar levels of output for black and white along with color regardless of two-pixel or single-pixel response. If you were here with me you would note a pixel perfect response that can't quite be captured with a camera. While the image obscures the red single pixel response in real life, they are just as bright as the two pixel response. Observed during this test was the perception of the black lines being ever so slightly wider than the white ones, but that was due to the fill factor between pixels adding to the size of black pixels.
The BenQ also has frequency and phase adjustment, either auto or manual, for the analog video inputs for 1:1 pixel mapping. Whether you choose auto or manual adjustment, you need a 1080 pattern that provides a single pixel response like the burst pattern I use to set it properly. On that note I found the auto adjust just as precise as doing it manually. If not set properly, you lose the pixel mapping and get a banding artifact covered in the next section.
If the above is confusing, you are in good company since the centered pixel mapped output of the REAL setting is unique for a consumer display. For the video perfectionist, the REAL aspect ratio setting offers 1:1 pixel mapping. This feature is great for seeing both 1080p and 720p as a straight shot from your source without scaling. What that means is 1080p and 1080i will be full screen but 720p will only occupy the center of the screen using only the 1280X720 pixels within the 1920x1080 panel. Naturally this means the image size decreases as the pixel matrix decreases. Ultimately this makes sense because if you increase the size of the image you also increase the size of the pixels and at some point they will become quite clear. You are then seeing the technology as well as the image, and that is an artifact. With 480p or SD signals this has been mapped as 4:3 640x480 rather than the 16:9 DVD standard of 720x480 which means you lose your 1:1 pixel mapping for 16:9 DVD but retain it for 4:3. It would have been nice to have both options available considering the videophile performance this feature is intended to provide.
Back to ANAMORPHIC. Naturally the projector provides the ability to watch any input scan rate with the screen filled out using ANAMORPHIC, with the penalty of a 4.5% overscan. The W10000 provides a unique opportunity for me to show you just how destructive it is to veer away from 1:1 pixel mapping using 720p as the example. I have included a 720p burst from the Sencore VP403 comparing REAL, 1:1 pixel mapping, with ANAMORPHIC.
TOP: 720p 1:1 pixel mapping, BOTTOM 720p scaled to 1080p
This was not the best camera shot, and pixel detail is blurred creating the misperception of thick white lines and thinner black ones. Nonetheless, a comparison makes it quite clear how the 1280 response of 720p is riddled with artifacts when remapped to 1920. The perfect transition of one pixel off and one pixel on creating a nice hard edge is replaced by a transition of in-between pixels; intermediate pixels in between peak white and black. Note that numerous white lines in the scaled image for 1280 don't even reach peak white, another clear artifact, and one of the lines in the 640 section is not the same size, being off by one pixel with a different luminance response compared to the lines on either side. At the viewing position where you are seeing the forest rather then the trees, this shows up as banding. All of these errors combined clearly degrade detail response and have the same affect on vertical pixel mapping and detail for the same reasons.
An example of banding is shown and explained in the Zenith DVD DVB318 review in our HD Library. Scroll down to the high frequency burst image from the DVE calibration DVD; note the text above and the related waveform images.
I love the REAL aspect feature and BenQ should keep that feature intact! If BenQ would replace ANAMORPHIC with the conventional method employed by other manufacturers for a16:9 aspect ratio, it would be a great step forward for future products.
Out of the box, the W10000 doesn't do anything wacky with gamma to sell it, as shown by the near-perfect similar response.
Delta C Pre-calibration
D65 RGB Chart Pre-calibration
Delta C Post-calibration
D65 RGB Chart Post-calibration
Out of the box, pre-calibration, the W10000 has a Delta C error of 2.5 to 5.5 and the RGB chart shows this error is driven towards cyan by increasing both green and blue output over red. Red is the Achilles Heel of all arc lamp based displays as it is the primary with the least amount of light output. All display products have a weak primary so no big deal. What this means for an arc lamp light source is to increase light output for sales and marketing you turn up green and blue since they have more light output to offer which ends up pushing the color temperature towards cyan. This creates a fairly common marginal error for such products and correcting it had marginal impact on light output.
For post-calibration the W10000 provides an excellent response for a consumer display pushing a near reference response except for the bottom end of black exceeding a .5 Delta C error. The RGB chart mirrors this excellent response and subjective viewing showed no obvious errors in the blacks. Errors below one can be difficult to perceive for the layman but is likely to be seen by a colorist professional. If you are doing professional work, this could cause an ever so slight color error that one might be tempted to compensate for when compared to the reference.
In the service menu for calibration, the BenQ does have blue channel isolation for setting color and tint but does not provide any means to isolate the red and green channels to professionally check color decoding.
This area of response is directly related to the Color Enhancement and, oddly enough, the 3D Color Management sections of the customer menu. Color Enhancement is one of those goofy color processes similar to the Mitsubishi Perfect Color that consumers perceive as useful and calibrators find totally annoying because neither one correctly address color decoding alignments, and using them can easily cause a lot more harm than good. The short version of the harm for the W10000 is that using these adjustments only affects unique phase angles for the color. Where you should have the same color and intensity of yellow within different sections of a color decoding pattern you would see a different response solely due to what other colors are next to it, the complex phase angle of color decoding. For the W10000, I calibrated two different ways; first using Color Enhancement and 3D Color Management to improve the color decoder; and second, ignoring it completely using only 3D color management. Both calibrations produced comparable results. This test became necessary because when calibrating in the service menu, the Color Enhancement controls do not appear and all you have to work with is 3D Color Management. While 3D Color management should only address color space it includes a saturation adjustment that only affects color decoding.
In the end I simply felt better about color decoding when the Color Enhancement controls are available. With HDMI and 1080p capability being the cat's meow, I calibrated HDMI using the customer menu and the analog video component input using the service menu.
HDTV BT709 Color Space Pre-calibration
HDTV BT709 Color Space Post-calibration
The W10000 response was a bit surprising for this alignment. Ultimately the color segments on the color wheel are outside of all industry standard color spaces and it is color management that brings them back in, on target. The color wheel creates a green primary way above the green target and some manufacturers leave it that way along with red or blue to differentiate their final image from competitors in the market place. Out of the box, the green veers towards yellow as well as pushing towards the edge of the chart. Using 3D Color Management I was unable to get the green on target. Looking at blue you can see it is slightly off and the controls allowed a precision alignment all around the target but not on target. Because the green can't reach its target, that has an effect on the secondary of cyan, between green and blue, as that is a byproduct of color decoding based on the primaries. I contacted BenQ about this mystery with green and was told that any change to this response will have to wait for the next generation product. Considering the level of performance provided so far, this was very unfortunate as the product is capable and would have allowed the end user to calibrate for any of the industry standard color spaces used around the world.
As with the wacky color processing used by Color Enhancement, the 3D Color Management provided a similar trait by providing color decoder saturation controls for the secondaries which plays absolutely no role in color decoding since their saturation point is a byproduct of proper color decoding for the primaries.
While the above image appears perfect, note that magenta in the 1st, 3rd and 4th blocks has a one-pixel error when it transitions with green and cyan, creating a 1 pixel darker line very evident in the 1st block. While not reference, this near perfect response was excellent compared to other consumer displays.
With the manual iris wide open, and using a calibrated D65 light output at 168 lamp hours, I obtained 501fl at 100IRE and .447fl for 0IRE yielding a contrast ratio of 1120:1.
The new firmware version supports the ISF CCC interface for calibration and alignment representing the ISF Day and Night preset modes. In the past this feature required an interface for a PC to access it, but the W10000 breaks new ground by allowing access via the service menu which is a huge plus since most calibrators have not had enough demand to justify the additional expense for the interface. Many calibrators have also recommended against using the feature because in past implementations all customer controls were blocked. The W10000 is ground breaking yet again for the ISF CCC system by allowing the user to make adjustments to these presets. Bear in mind doing so does not allow you to change the memory settings and your changes are only active during that particular viewing session which are reset to the reference values input by your calibrator once you cycle the power. This is a huge plus since any change you would want would also be related to the content you are viewing at that time.
Whether it is the lamp or filter that needs access, BenQ has given some thought into making that convenient for you. Both are accessed via the side panels allowing you to leave the projector on the mount, a huge plus. The manual covers this, yet I found myself struggling because neither of them swings out to release them as the manual infers. For the filter there are two plastic locking tabs that you push in then slide the panel towards the bottom maybe a quarter inch and it will release. For the lamp you loosen two screws (with retainers so they don't fall out) and slide it also about a quarter inch towards the bottom and it will release. To be clear, the bottom infers the end with the feet which when mounted on the ceiling would be the top. Now that I know how to do this, it is relatively easy! Mounts tend to be sensitive and enough force is required that you will likely have to center the image once you are done.
For the hands on videophile who has no problem driving their equipment using customer controls and menus, it is nothing more than a momentary irritation over a number of days requiring a few keystrokes on the remote. It did not happen everyday but came and went with a mind of its own presenting itself at power-on only; once remedied, the lamp remained stable. It does not appear to have any effect on lamp life. On the flip side if you have a videophile client who is technology challenged and catches this problem, that could be trouble and no amount of reassurance concerning its temporary effect may suffice. That said, the client with the ISF calibration never did notice or call about this problem and his was ceiling mounted so position appears to have no role.
As of the publishing date, I have received no further comment from BenQ on this matter. If they do respond, the article will be updated to reflect that response.
It doesn't work conveniently for my zoom 2.35 screen application due to the zoom limitations. I am forced to physically move it farther away or closer to the screen. Due to that same limitation I recommend you measure the throw distance if working with a small room or if you plan to install it on an existing ceiling mount; it may not work out or you will have to move your mount. Like most projectors you need to get it perfectly centered left to right since using any keystone correction destroys the all important 1:1 pixel mapping; this feature is disregarded altogether for that reason.
While 1:1 pixel mapping using the REAL aspect is a novel performance feature, most are going to want their images full screen. But if you want the projector to do this, you are going to be forced into some unnecessary overscan using the internal scaler. An external scaler solves that and also adds $1-2K minimum to your budget. Many performance systems already have an external scaler, so this may be a moot point for those upgrading. Another solution is to set your source for one scan rate and have it do the scaling to 1080i/p, but you could be faced with a similar overscan issue. The best solution would have been 0 or near 0 over scan for all content not in 1080. Considering the fact that most front projection systems will have 1-2% overscan applied by the installer to cover up imperfections in installation or the source, I fail to find any valid reason for excessively over scanning content that is not 1080.
I have yet to review reflective LCD technology, so I have no comment other than to say that Greg Rogers did a review on the Sony Pearl VPLVW50 for Widescreen Review and in the technical portion there were some response similarities to our currently reviewed Panasonic PTAE1000U. Both have a reputation based on the dynamic iris technique. At this level of performance, the JVC D-ILA series is another one that should be on your list for consideration.
Due to the focus uniformity, green primary errors and lamp problem, BenQ sent out a replacement. During the production run, BenQ rearranged a few things on this projector creating two different animals when it comes to calibration. The following additions reflect only those elements that are different between the two units.
For calibration, the ISF CCC system adds 7 gamma settings and separate gamma controls for red, green and blue. The gamma setting preset by BenQ is just fine. Otherwise the ISF CCC system is a duplicate of the user menu minus the Color Enhancement menu. As far as calibration goes there was not any real difference for the end result; both ended up with the same level of performance.
The new version has increased the on screen display graphics to double the size. For the end user this was a good move as the original menu pixel mapped to 1920x1080 providing very small fonts. This was most noted in trying to read the incoming scan rate display at the bottom right corner. The down-side is that the menu now is so large that it covers up areas in the test patterns. This can be worked around but the first version was more convenient for calibrating. This version also allows service menu access unlike the first, and in there I found factory resets for zoom, focus, iris and most importantly lens shift, so it can be returned to center.
The second product also went through a phase of the lamp pulsating light output.
Posted by Richard Fisher, July 5, 2007 8:28 AM
About Richard FisherRichard Fisher is the President of Mastertech Repair Corporation, serving north east Atlanta, Georgia, and has been servicing, calibrating and reviewing audio video products since 1981. Tech Services USA, a division of Mastertech, creates sites, communities and libraries for consumers and professionals to share their technology knowledge and learn from each other. These include The ISF Forum and HD Library. HDTV Magazine exclusively publishes HD Library and Forum for Tech Services USA.
Richard is ISF and HAA certified providing calibration and A/V reproduction engineering services. Richard is a technical consultant and also provides performance ISF and HAA home theater systems and calibration via Custom HT. Mastertech Repair Corporation is a factory authorized service center for Hitachi, Mitsubishi and Toshiba and a member of the National Electronics Servicing Dealers Association, NESDA, and the Georgia Electronics Servicing Dealers Association, GESDA.