Reseller BBB

Store

DLA-RS3000/DLA-NX9

Model:
Availability: Reveiw
Shipping: $0.00

Quanity:



Product Description

Review: RS3000

Cleveland AV owner Chris Majoros was kind enough to send me a new native 4K JVC RS3000, pulled from regular stock, to test and evaluate over a two week period. The RS3000 temporarily replaced my JVC RS67, and was shelf mounted right behind and just slightly above seated viewer’s heads. It projected from a distance of 144” onto a 119” diagonal, 16x9 Da-Lite High Power 2.8 gain screen. 

Usability was excellent, and several things caught my attention as being especially noteworthy in comparison to other JVC models. Audible fan noise, even in high lamp power mode, seemed very quiet. This was especially welcome since my head was normally about two feet from the lens, and I did not use high lamp mode very often on the older RS67 due to excessive fan noise. Also, the resync or black out time when changing resolutions or signal type to the projector is much shorter than on any other recent JVC model, with worst case scenarios (SDR to HDR, lamp power change, aperture going from fully closed to fully open, color filter change) taking less than 11 seconds, with more typical times less than 7 seconds. The redesigned, backlit remote looks sleeker and feels more solid, but unlike the older design, it was very difficult and frustrating to try to operate by feel. Finally, there is a provision to instruct the projector which picture mode to switch to when HDR is input, which can be surprisingly handy. The RS3000 supports the popular HDR10 format and also HLG HDR, which is starting to emerge on a few sources including YouTube. Active HDR tone mapping, an important new feature, is designed to maintain the best brightness balance without crushing highlights. It also supports 3D which was not tested, but judging by other recent JVC models the RS3000’s 3D mode will be world class in every respect.

Displaying a black screen showed slight light corners, something JVC has struggled with for some time. It was not quite as bad as on my RS67, and while it did not bother me, I do know some people who would find it distracting. This characteristic seemed somewhat middle of the road in my opinion, having personally seen both better and worse examples in late model JVCs. Gray and white field uniformity was excellent with the lens aperture in the upper two thirds of it’s range, with a just barely perceptible green shift to the right. However, the green shift became much more visible at or below an aperture setting of -11, so even with small or high gain screens staying at or above -10 may be advisable.

Focus was superb over most of the screen, reminding me strongly of JVC’s excellent RS4500, though the bottom right was slightly softer.

Before calibration: 

Familiar scenes in Lucy looked great in the HDR10 picture mode, with excellent highights and definition. The picture was extremely punchy and dynamic. The Active Tone Mapping appeared to be working very well, although my unusually bright setup does not provide much of a challenge for it. Shadow detail was surprisingly strong, perhaps a bit too much for this movie which normally crushes near black detail. Blacks within the movie frame were slightly elevated, causing the 2.40 movie frame to glow. This is not a byproduct of the RS3000’s contrast; rather, it is caused by the factory calibration jumping out of black too fast. Otherwise, the blacks looked excellent. Colors were very pleasing on Lucy, but checking my reference images showed the out of the box HDR color to be somewhat on the pale side. Watching First Man on 4K HDR Blu Ray for the first time was not nearly as satisfying, lacking in definition and appearing excessively grainy, though it appears the projector was just displaying flaws in the movie. 

Watching a bit of normal 1080P TV shows before calibration was very enjoyable; the image was powerful and very much like a large, high quality flat panel. The only distraction were some skin tones and brown shades that seemed a bit violet or lavender.

Calibration:

The Spyder autocal software available on the JVC Kenwood website has been a useful tool for making gamma and colors more linear on recent models, and a new version of the software must be downloaded for each model year. The autocal should always be verified and fine tuned by third party software and better quality meters, or all that is likely to happen is the inaccuracies will just be shifted around with little overall improvement. In the case of the RS3000, running autocal degraded overall more than it improved. Despite running a color autocal run at each 5 step interval of the lens aperture, after autocal the white balance started off relatively accurate at an aperture setting of 0 and gradually became very cool as -15 was approached. I reran the various autocal procedures with no improvement, and then even repositioned and reaimed the Spyder5 before running the procedures a final time with no better results. Having preformed autocal on all the recent JVC models, I knew better than to expect a completely bug free experience, but this along with other shortcomings in the autocal results was especially discouraging.

The RS3000 was very quirky when it came to precise calibration. Say, for example, the user menu Color Temp. and Gamma menus were precisely calibrated after running autocal. If I then went in and changed the Color Temp. and Gamma numbers, measured the results, and then put the original calibrated numbers back in and measured again, the final results were often different than the original results using the same numbers. The shift may not have been enough to be visually significant to most viewers (change in dE2K was generally less than 2 or 3), but it is enough to cause a great deal of frustration when fine tuning the calibration. I have seen some other recent JVCs with similar behavior, though plenty do not show it.

I was surprised at the RS3000’s scaling and pixel structure with 1080P focus patterns. Focus and other finely detailed patterns looked chunky and lumpy rather than crisp and sharp, and changing or disengaging the e-shift had no effect with 1080P. It was easily visible upon careful and up close examination, but could even be seen from my normal viewing position. Thankfully, it did not draw attention to itself on the normal 1080P images I viewed, and if it did an easy fix would be to upscale any 1080P sources to 4K before going to the projector. Owners of high quality 4K Blu Ray players should have the player upscale Blu Rays to 4K to avoid this. 4K (as opposed to 1080P) focus patterns looked great and did not suffer the same effect. The 8K E-Shift changed and rounded the fine structure of 4K just as the 4K E-Shift did with 1080P on JVC’s other recent models, and engaging it will be a matter of personal taste.

Input lag with a 1080P input, which is the only resolution the lag tester is capable of, measured 42 ms in any picture mode with any settings, provided CMD in the Motion Control submenu was turned off. With CMD on, it was 103 ms, again in any mode and with any other settings. Curiously, the low latency setting had no impact at all other than locking CMD in the off position.

CMD did increase motion resolution, but at the expense of introducing the much scorned Soap Opera Effect. The last item in the Motion Control submenu, Motion Enhance, did not increase real motion resolution but did slightly exaggerate sharpness of the motion resolution that was already resolved. There did not seem to be any significant difference in motion handling between the RS3000 and other recent JVCs with normal film and TV content, although admittedly serious sports fans sometimes notice differences that escape me in this regard.

The on/off contrast ratio after calibration, measured with a Klein K-10 facing the lens with a diffuser to maximize the meter’s dynamic range, was 20,868:1. Closing the lens aperture, which should be possible in many installations for normal HD, will increase this number, which should be viewed as more of a minimum. Just how much the aperture can and should be reduced is determined mainly by the screen size and gain, along with the throw distance and viewer preference. Also, using the Auto aperture settings will dramatically multiply these numbers. Although lower than JVC’s past lamp based flagships including my own RS67, I thought the RS3000’s contrast looked very rich and deep.

Full field on/off CR, manual aperture:
0: 20,868:1
-5: 31,442:1
-10: 42,144:1
-15: 59,920:1

10% size, 25% APL window CR, manual aperture: 
0: 1,788:1
-5: 1,953:1
-10: 2,329:1
-15: 2,312:1

Modified ANSI, measured only at one point of a 4x4 checkerboard, manual aperture:
0: 197:1
-15: 222:1

Switching from low lamp power to high gave a 27.5% increase in light output. Engaging the color filter cut light output by 15%. Since the HDR P3 color volume was near 94% even without the filter, I believe JVC’s choice of making the default HDR10 mode not use the color filter to be a wise one.

After calibration, the RS3000’s measured performance (color, grayscale tracking, and gamma) was virtually flawless in SDR mode. 

The following supplemental measurements are attached in zipped folders:

SDR gamut luminance, SDR skin tone color checker, SDR saturation sweep at 100%, SDR saturation sweep at 50%, SDR saturation sweep at 15%, SDR color volume analysis.

HDR NF color volume analysis, HDR NF gamut luminance, HDR Color Filter color volume analysis, HDR Color Filter gamut luminance. 

Running across some older but well preserved screen samples, it seemed interesting to do a comparison of different materials to get an idea of light output and performance compared to the long discontinued High Power screen. Even many of these materials have been discontinued, however. Nits and fL were measured directly off the screen with a Jeti 1211 reference spectroradiometer and are accurate real world numbers; lumens, on the other hand, are calculated and rely on accurate screen gain ratings. Since manufacturer’s gain ratings are notoriously inaccurate, the exact lumens numbers are not listed below. However, averaging the four most believable results gives an estimation of 1,593 lumens calibrated in high lamp without color filter. 

Light output at 100% stimulus, 119” diagonal screen, 144” throw, high lamp, no color filter, 97 hours on bulb:

Da-Lite High Power 2.8: 413.4 nits, 120.65 fL (used for calibration)
Da-Lite High Power 2.4: 334.7 nits, 97.7 fL 
Carada High Contrast Grey .8: 57.8 nits, 16.9 fL
Carada Brilliant White 1.4: 144.16 nits, 42.1 fL 
Elite Screens AcousticPro1080 woven (with added black backing) 1.0: 107.07 nits, 31.25 fL
Elite Screens CineWhite 1.1: 123.46 nits, 36.03 fL
Elite Screens MaxWhite FG 1.1: 133.26 nits, 38.89 fL

Same as above, except low lamp power:
Da-Lite High Power 2.8: 324.18 nits, 94.62 fL

Same as above, except with color filter:
Da-Lite 2.8 HP: 359.55 nits, 104.94 fL

After calibration:

Watching some Netflix SDR HD material provided an opportunity to see how some shows I had been watching on my LG 55C8 OLED fared on the big screen. One older drama took a little getting used to; compression artifacts were more visible, but colors, including fleshtones, were both vibrant and natural. The image looked a little more “enhanced” overall than on the smaller screen, but as the show progressed it made a valiant effort to draw me in. Overall, it was very good, but I felt the presentation was a bit too ruthlessly revealing of this somewhat mediocre quality content. Switching the source’s output from 1080P to 4K made no distinguishable difference, something that needed to be checked because of the RS3000’s odd look with 1080P focus patterns.

Switching to a newer, more visually stimulating cooking show really brought out the RS3000’s strengths: popping color, wonderful depth, and photorealistic textures. The calibrated Dark Room mode, with it’s lower lens aperture setting, showed that in the right environment the RS3000’s contrast provides the image with a fantastic sense of dynamics. I tried the Auto aperture setting on both the Dark Room and Bright modes, and I could both see and hear it working more in the Bright mode. Fortunately, on these shows I saw no weaknesses or side effects of the Auto setting, and in the Bright room mode it kept fade to blacks nice and deep.

Switching to familiar reference 4K HDR Blu Rays, the image was, as expected, simply stunning in HDR Bright (a renamed User mode based off high lamp, no color filter, Auto 2 aperture, and custom gamma curve). The incredible light output, combined with the Auto aperture (AKA Dynamic Iris), led to a powerful image that had to be seen to be believed. Shadow detail was nicely balanced, and colors were rich and satisfying. Revisiting First Man, colors were noticeably better than before, but it still proved to be one of the less visually compelling 4K HDR releases.

I had until recently been lukewarm in my appreciation of JVC’s DI (Dynamic Iris). It has had problems with blooming of small and bright objects, maintaining accuracy in shadow detail of dark scenes, and general stability. I was pleasantly surprised when I tried two chapters in The Greatest Showman that have, on previous JVCs, highlighted these issues. First, starting around 5:53 of chapter 2, shadow detail on the RS3000 was not crushed even with the more aggressive Auto 1 DI setting. A ticking sound was audible as the DI worked, but it was clear that it made a substantial improvement in visible contrast of this dark scene without seriously compromising shadow detail.

Next up was chapter 11, where the DI in previous models has made a blooming mess of Barnum as a bright spotlight shines on him from a distant camera view. The RS3000 performed admirably on this test, with only the faintest hint of compression visible when compared to a Manual aperture setting. This was nothing like the waxy, glazed over look seen in older models, and good enough to firmly win over this DI skeptic.

A new feature of the RS3000 is the Auto Tone Mapping feature for HDR, which analyzes metadata in the program and gives a Mapping Level control with a range of + or -5 to adjust brightness. This Mapping Level control is almost always going to be used to brighten the image rather than to darken, so it can be thought of as a brightness boost with five levels of strength. Up to this point, the RS3000 had the advantage of an extremely bright setup. To give the RS3000 a nice torture test, I closed the aperture all the way and made a new 80 nit Custom gamma curve with Arve’s tool to compare with JVC’s Auto Tone Mapping. This will give a closer approximation to the light output of an RS3000 on a larger, lower gain screen. 

The RS3000 did not pick up any metadata from The Greatest Showman, so the Auto Tone Mapping could not be engaged. Planet Earth II, which would have been a good test for highlights with it’s bright clouds and snow, was the same. Switching to Lucy, it was not until the Mapping Level was turned all the way up to +5 when the overall brightness of the ATM approached that of the Custom gamma curve. At this setting, the JVC’s ATM looked polite and inoffensive, with very well controlled highlights, but somewhat anemic. Shadow detail was noticeably elevated, which when combined with the reined in highlights, robbed the image of it’s depth. The Custom curve was much punchier and looked more like what one would expect from HDR, with more balanced shadow detail and brighter, though occasionally more compressed, highlights. It is interesting to note that, while the ATM seemed very impressive in the before calibration round of viewing, the result in this round of testing was less positive. Before calibration, the image benefited from full brightness of several hundred nits and the Mapping Level was at it’s default of 0. On the other hand, in this torture test, the brightness was intentionally limited to better match what would be seen on a large, lower gain screen, and Mapping Level had to be increased all the way to give the image enough brightness. It appears that the RS3000’s ATM tends to loose punch and depth when attempting to brighten the image for low light output installations. Overall, I felt the ATM was a significant advancement over what has previously been available, but I believe users with low to moderate light output setups will find it lacking in impact and pop.

Conclusion:

The RS3000 is a significant advancement in many ways over JVC’s previous models, which is saying a lot. Native 4K resolution and excellent focus give fantastic detail. The improvements in the Dynamic Iris are significant, and make it a very useful enhancement that can be used without reservation. Even though the measured contrast numbers were not as strong as other recent lamp based JVCs, the DI improvement minimized or made up for any visible contrast difference in my situation. The RS3000’s image reminded me of JVC’s much more expensive RS4500, and it’s time in my home theater was amazingly satisfying.
 
Click Here for Charts and review on AVS Forum


Product Images