Projector Review: JVC DLA-RS640
LCoS – JVC DLA-RS640 Projector
After reading every review I could get my hands on, for both native 4K and pusedo 4K projectors, and researching the current abilities of projector technologies to support the new 4K UHD standard I decided that the one for me was the JVC RS-640. Why? Read on.
Considering the price tag that this projector carries, and the fact that it’s JVC’s top of the line Reference Series projector, excluding their new native 4K 4500, I was amazed at the less than stellar packaging. Just a single layer of molded polystyrene, that in my case, had broken in several places, which I am sure was due to mishandling by the shippers. The good news was that the projector didn’t seem to have sustained any damage.
So you may be asking yourself why choose this particular model? Here are the reasons in order:
- Best native contrast ratio of any projector. This is an essential for outstanding 4K UHD support, as is total room light control.
- The latest e-shift 5 technology provides an excellent 4K pseudo performance.
- Very bright at 2000 lumens. Again a requirement for good 4K UHD support.
- Supports REC 709 out of the box and is pre-calibrated to support both ST2084 and DCI-P3 color space for 4K.
- Supports HDMI/HDCP 2.2 for 4K HDMI.
- Supports 18Gb/s, 4K60p 4:4:4, 4K60p 4:2:2/36bit and 4K24p 4:4:4/36bit.
- Ease of calibration for both REC709 and DCI-P3.
- Ease of calibration for both SDR gamma and HDR OETF functions. The latter is very important in order to correctly handle HDR.
- Outstanding iris performance, if required, to give that extra dynamic range.
- Outstanding range of menu functions to configure the projector for virtually any environment
- Stellar reviews by well known and established individuals.
- Good AVS Forum support group for all those questions.
Installation
By the time the projector arrived I had already installed my new Oppo 203 and two new fiber optic cables to support 4K, 18Gb/s and 1080P, 10.2 Gb/s. It turned out that my current ceiling mounting plate wasn’t large enough for me to drill to match the mounting holes of the JVC. So, after several hours of work and two visits to Home Depot, I managed to attach heavy duty rigid extensions to my current plate that would now carry both my Panamorph lens on its motorized sled, and the projector. Boy! Was this assembly heavy. So, with the help of my daughter, I managed to slip the new assembly into its matching quick release ceiling bracket and square it all up. I quickly attached all the cables and power and I was ready to go. HDMI 1 was connected directly to the Oppo 203 and supported both 1080P and 4K. HDMI 2 was connected to my AVP Pre amp to support 1080P and all my other players and devices.
Top View – Quick Release Plate and Three Mounting Arm Extensions
Front View – Panamorph Sled Attached With Lens Removed.
Projector In Place Ready To Be Aligned.
Initial Geometry Setup
Having spent several hours reading the JVC manual, in great detail, I felt I was now ready to set up the projectors geometry. Unfortunately I had not already inverted the projectors image so it took more than just a few minutes to find the image and get it to point at my screen, as most of the image was actually on the ceiling! Selecting position, zoom and focus I quickly had the display all squared up and the geometry matching my 16:9 screen with the masks down. I moved the Panamorph lens into place, retracted the screen masks, and adjusted the Panamorph lens such that the light passed through its center and the image was square and parallel to my 2.35:1 screen. Well that all went well.
At this point I checked the focus and panel convergence using the projectors internally generated signals and my DVDO. Both were very good on the right hand half of the screen but as you moved to the left hand side of the screen both the red and blue convergence moved out by about half a pixel both horizontally and vertically, and the resulting focus became less than perfect. I spent about an hour converging the left hand side of the screen only to find out later that while the projectors test signal now looked good, as did the focus, the process added a slight green hue to the corrected zones low level gray scale and various test signals; more on this later.
Left Hand Side Panel Miss-Allignment
Right Hand Side – Perfect Panel Alignment
According to JVC engineering my throw distance of 14′ 6″ causes my lens to be close to one end of its limits as my 16:9 screen only has a 90″ diagonal. They believed that the convergence/focus problem was a result of this, and were not anxious for me to return it for service or generate an RMA to send me another unit. Especially as I had no faulty pixels, had really good uniformity and no bright corners or other issues. According to the JVC lens calculator my screen distance is right in the center of the lenses capability for both of my image sizes!!
To be honest, not only could I not see the convergence problem from my seat, but I couldn’t tell any difference in focus across the screen either. Both issues only being observable when I was about 6 feet from the screen. I decided after much wailing and gnashing of teeth, enough was enough, I turned off the convergence correction and moved on to looking at REAL pictures.
UPPDATE: 5/25/18. After sending a couple of dozen photographs of the less than stellar convergence (and resulting impact on focus) to JVC engineering and customer relations, JVC concluded that this $8000 projector was within manufacturing specifications!
At this point I inserted my HD Spears & Munsil test disc into the 203 to see what test signals looked like of a real disc, as opposed to being electronically generated. Using the overscan test signal, and no e-shift, I managed to re-focus the image so that I could clearly see all the pixels, across the entire screen. A small amount of color fringing did exist, but it was extremely minor and invisible from my seated position. Switching in the convergence correction removed all color fringing and made the image almost pixel perfect. However, I originally choose not to use it due to the blotchy hue change that it caused. Apparently this is a ‘bi-product’ of using sub-pixel alignment. Later tests revealed that engaging the convergence correction did not impact the color of real pictures, and had the advantage of cleaning up most of the minor color fringing on text and some edges.
Green Hue After Sub-Pixel Panel Convergence for Both Red And Blue Panels
Brightness and Contrast Setup.
Feeding the projector with my HD Spears & Munsil and UHD Diversified HDR 10 test discs I adjusted the black and white levels for both HD SDR and HDR ST2084, using the recommended Auto HDMI input setting on the projector that selected Standard Level. This setting only passes levels 16 through 235, clipping everything above and below those levels. Enhanced supports 0 through 255 and allows above white and below black levels. The remaining Super White mode, levels 16 through 255, clips all levels below black but passes above peak white.
HDMI Input Level Selection
In the past I have always selected the HDMI input level that supports below black and above white, (Enhanced 0-255), and calibrated it accordingly. After extensive research I discovered that the projector should be set to STANDARD, levels 16-235, those levels actually encoded onto a Blu-ray. Even JVC advises the use of this input standard providing a notable improvement in black levels and not wasting bit levels. I also read that significant adjustment of the brightness control for the enhanced mode affects the gamma. So I left the input on Auto.
Setting the HD SDR black and white levels was straight forward using the JVC brightness and contrast controls. Only requiring minor changes to the controls default 0 positions. The SDR test signals were very bright, even in low lamp mode, looking excellent with outstanding black level and really saturated colors. All gray scale, resolution and color signals played back flawlessly and the short video clips looked magnificent. Turning now to calibrating 4K ST2084 OETF (4K gamma) was a new experience for me. Setting up the black and white levels for HDR is similar to SDR. However, I found that just using the brightness and contrast controls to optimize these levels didn’t yield a good HDR picture. Generally being too washed out, or too dark. So after extensive back and forth between these controls and the ST2084 OETF controls, I finally managed to get an image that looked very good. However, I have no idea what the resulting OETF curve looks like. Clearly this adjustment is not for the faint of heart and REALLY needs to be setup by an experienced ISF calibrator.
ST 2084 OETF Correction Controls
I stepped through most of the 4K test signals and all looked excellent. No color clipping below 1000 NITS, fabulous gray scale and color bars, resolution was correct and zone plates etc. looked nice and clean. I then turned to the included 4K images stepping through each. They all looked stunning and three dimensional. The only issue was that peak white levels occasionally looked to be slightly clipped. Attempts to fix this generally messed up the entire image so I left well enough alone. I have no doubt that the forthcoming ISF calibration will resolve this issue.
As I do not own a suitable colorimeter I relied upon the REC709, HDR, and BT2020 (DCI P3) factory color calibrations by JVC to be close to what they needed to be. Based upon reviews, the THX mode is considered to be just about spot on for REC709 with a gamma of 2.2. However BT2020 and its associated OETF needs to be calibrated once installed.
Both HD and UHD gray scales looked excellent with no visible tints. A quick use of color filters for HD, together with appropriate test signals, showed the color balance to be very close, as was the gamma to 2.2.
I was ready to rock and roll!
First Impressions
Using HDMI input 2, I selected THX picture mode as this is REC709 calibrated, 2K in order to disable the e-shift, low lamp mode and auto iris 2.
1080P THX Settings
I first slipped Galaxy Quest 2, a 1080P HD SDR Blu Ray, into the Oppo 103. WOW!! The colors, contrast ratio, blacks, peak whites and sharpness were awesome, edge to edge. What was really impressive was that even without my Darblet activated, and with the minor panel miss-convergence and focus issues the image was sharper, had more punch and overall depth than my single chip BenQ DLP with the Darblet. So far I was really happy.
I then turned on the Darblet at 35%. Much to my surprise, while it definitely improved the contrast and sharpened the image, the improvements were more subtle than with my BenQ. This is clearly a testament to the performance of the JVC. Already producing an image that was sharper and with more dynamic range than the BenQ plus Darblet. Nevertheless the image enhancement was well worth having, so I left it on for all HD SDR 1080P viewing. The one thing that I certainly will not miss were the ‘rainbows’ from the BenQ DLP.
4K HDR Input Settings
Selecting HDMI input 1, I then popped Star Wars -The Last Jedi, 4K HDR, into the Oppo 203, leaving the projector in all its auto modes. Upon detecting a 4K signal the projector switches to e-shift 4K, high lamp mode, HDR, BT2020 and ST2084 with dynamic lens mode 2. Just as with Galaxy Quest, the image was outstanding, from the deep deep blacks of space with its myriad of bright stars, to the bold saturated colors and notable increase in fine detail and image depth. Quite exceptional….but, and there is always a but! I could tell the the peak white and blacks were not quite right. Adjusting the avilable controls didn’t improve things. I had to assume at this point, that an ISF calibration is now required, together with some custom HDR ST2084 (OETF) curves. More on this after my ISF calibration.
Setup Changes
I did play with the manual iris to see just how much the picture black levels improved over the dynamic iris. Without proper calibration I could see no benefits of not using one of the iris modes. Happy to say that the iris is so quiet and well controlled that it provides exceptional black level control without becoming visually intrusive or mechanically noisy. I have yet to settle on Auto 1 or 2. However I feel that Auto 2, my default mode, provides less image processing than Auto 1. Better blacks always seem to be achieved by using Auto 1, but occasionally the iris can be seen to be pumping, usually only on credits, that to me doesn’t matter. Right now I am running a fixed aperture of -6 and Auto 1 for HD with low lamp, and an aperture of 0 and Auto 1 for 4K with low lamp or sometimes high lamp, depending upon the movie production.
Lamp And Lens Settings
After comparing the 21:9 stretch of the 203 and the anamorphic mode of the projector I could see absolutely no difference in the images. So I set my Oppo 203 to Direct so that an HD disc puts out 1080P and a UHD disc puts out 4KP and the player no longer did up conversion. This simplified the projectors setup procedures to create pixel mapping for HD and e-shift 4K for UHD.
The player or the projector can be setup to display 4K images from a 1080P and 4K signal. However, I felt that the up-conversion of HD to pseudo 4K didn’t really benefit the image, which to be honest, was already outstanding. Up conversion creating images that looked more ‘film like’ but not sharper or softer. Also allowing up-conversion and 4K playback for all discs runs the bulb in high power (you can manually set it to low power). Continuos use of high power will significantly reduce the lifespan of the bulb and causes the projector fans to be somewhat louder, but fortunately still not intrusive, being virtually inaudible in low lamp mode.
The menus have a huge range of presets for gamma, color, and adjustments of every kind you can think of. While I played a little with these just to see their effect. I decide that the default modes and settings provided such a terrific picture that I could now wait until it is ISF calibrated. After which I probably won’t want to touch it!
A Selection Of Color Temperature Adjustments.
The only other feature that I tried was Clear Motion Drive (CMD) as it improves motion. Set to low, it did reduce motion judder without visibly impacting the image. Motion judder was something that I rarely saw on my DLP projector. However, after reading several AVS posts it was discovered that enabling this control may introduce color fringing, something I never saw. It apparently will also randomly come on even without you setting it on. This meant that every time you turn the projector on, you needed to turn the CMS on then off to ensure that it is disabled. This is not a problem if you leave low latency enabled, which disables CMD. There is a JVC fix for this, but I understand it involves returning your projector to them.
Disappointments
I suppose other than the not quite perfect panel convergence and the resulting minor focus issue, my biggest disappointment was that the three controls that were available to taylor the ST2084 OETF curve, could not get it ‘right’. They certainly helped, but getting the image to match the required HDR 10, BT2020 test signals was never going to happen. Furthermore, I had no idea what the resulting OETF curve looked like. It is possible to have the Oppo 203 map HDR to SDR, and that certainly worked quite well, leaving the color space at BT2020, (Oppo set to HDR off BT2020) and adjusting the projector setup accordingly. However, I didn’t spend this much money to defeat what I was looking for, HDR. I therefore returned to HDR and ignored the occasional minor white and black clipping.
Conclusions
Do I have buyers remorse, well just a little. For a top of the line projector I expected near perfect panel alignment together with all but perfect focus. I got neither! Can I see it from my seated position? No. Does it impact my viewing pleasure? Not really. BUT, the problems are always there nagging at me! However, with an image this good, even without ISF calibration, I suppose eventually I will let them slide.
Next Steps
In my opinion it is foolish to spend significant monies on a 4K HDR capable system and not have it ISF calibrated in order to get what you actually payed for, HDR and BT2020, well at least DCI-P3. So Chad of HDTVbyChadB has worked his magic. See here for the ISF calibration results.
Click on the following links for more information:
