NEW MITSUBISHI HC 9000 : FULL HD, 3D and..... Lcos panels - Page 10

I would think brighter, if its lumens you want and price

Hi Mark,

Thanx very much for your observations.

Can you share some thoughts regarding Contrast, Black level, Shadow detail and perhaps how well does this projector perform against the obvious competition i.e Sony and JVC.

Thanx

Ran

Hi Ran. Too long since we last spoke on the phone.

I need to check panel alignment with a white grid test pattern. The internal grid pattern is green only.

I don't remember if Tom measured on/off, I wasn't in the room all the time, instead was on the phone taking some projector orders. I am sure it will be in his review though. I will measure it tomorrow. Measuring ANSI CR is not something I choose to do. I will look at the other things you want Ran but it will take some time viewing content.

Unfortunately, I have not had a VPL-vw90ES in my HT and really can't compare one to the Mits 9000D. I have read the reviews though and I expect the machines are very similar with the Mits being brighter but the Sony having a much better DI. I suspect the Mits will be a better 3D performer having a brighter image with less cross talk because to obtain the same brightness the shutters wouldn't have to be open for so long.

Comparisons with the JVCs must for the most part remain off line, call me if you want. Briefly, the blacks and on off are better on the JVCs with the Mits being brighter and I suspect having a better lens. But remember the Mits only has 15 hours on the bulb.

I spent some time with the new Mits SXRD projector in 2D mode only.

It throws a little over 800 lumens in high lamp and about 650 in normal lamp (iris open). Either is much too bright for Mark's 110" StudioTek, but using the iris brings the brightness down to a more comfortable 14-15 fL.

Native on/off contrast in the 5000:1 range (15/0.003 fL). You can get it as high as 7000-8000:1, but almost all of this comes from higher peak output, rather than lower black level. The iris affects peak output dramatically, but lowers the black level only slightly. I didn't use the auto-iris.

With onboard controls you can get a very good grayscale and a reasonably good gamma, though it rises at the high-end. With the Lumagen in the signal path both were essentially perfect as was the color performance.

Convergence was very good. Blue was off by one pixel in the vertical plane, which I fixed with the onboard pixel adjust feature. White field uniformity looked very good as well, though I did not measure it. I didn't notice the light corners Mark saw, but I wasn't looking for it.

I also liked the motorized lens shift, focus, and zoom controls that work in very small increments making fine adjustment possible.

Problems

There were two issues with this projector.

• First, the CMS is broken. It doesn't work at all.
• Second, the human interface is occasionally awkward. One issue irritated the devil out of me. When using the calibration controls for gamma, white balance, and CMS, the adjustment menus remain planted right in the middle of the screen making measurements of the test patterns difficult and time-consuming. The Vango suffers from this design flaw as well. Very annoying.

Bottom line

I agree with Mark on three points:

• The CMS is useless. This is odd, because the CMS on the Mitsubishi rear projection DLP works very well. This is no doubt fixable with a firmware update if Mits wants to devote resources to doing so.
• The image is very sharp and crisp and with good depth, a little sharper I think than the JVCs, though this is a subjective assessment.
• Normal lamp is essentially silent and High lamp is very quiet. The noise of my laptop fan made it inaudible to me sitting just a few feet away. Mark was even closer and could just hear it.

I generally prefer DLPs. However, paired with an external processor with good calibration controls, such as the Lumagen Mini or the DVDO Duo, this is the first LCoS projector that got me nervously fingering my credit card. On the other hand, without a processor, I couldn't live with it because of the color issues, though I am particularly anal about such things. If you don't care about color accuracy, then it is a very, very good projector as it is.

Equipment Used
Software: ChromaPure Professional
Signal Generator: AccuPel DVG-5000
Color Analyzers: JETI Specbos 1211/Klein K-10
Processor: Lumagen Radiance

With Mitsubishi's controls











With Lumagen's controls

Hi Tom,

It has been a while since we last talked.

Just to get this straight. It seems that your on/off measurements for the 9000 are considerably lower than other reviews I read which stated more in the 18,000 range.
They also seem far from the measured on/off for the RS40.
Am I missing something?

I have seen the RS40 on my 130" High Power and was totally blown away from the cinematic picture and the inky blacks. I can understand people complaining about brightness though with the HP it seems to be fine.
However, I'm holding my decision as I wish to have more brightness for 3D and be covered once the lamp ages, so the 9000 seemed to be the obvious choice, now after reading your numbers I'm not sure.

Have u seen the JVC or the 90ES? Can u comment on the 9000 performance compared to either of them?

Thanx

Ran

Looks like the projector is pretty good. Havent had a chance to compare all three but it should compete well against the JVC and the Sony

Yes and no. Yes, in that the 18,000:1 figure is almost certainly obtained by using the dynamic iris, which I did not use. As I said, the numbers I quoted were native contrast. No, in that the RS40 has even higher native contrast, but to my eyes for the vast majority of material the difference between the two is visually negligible in this regard. Speaking only for myself and my own perceptual biases, I preferred the Mits image because of the DLP-like snap and clarity.

Also, the SMPTE standard for a reference projector is 4000:1. In my view, any projector that provides that much native contrast or better with adequate brightness offers excellent performance.

Thanx Tom,

Just for reference, did u have a chance to measure the on/off on the RS40 and if so do u remember the number?

Ran

I did but I don't recall the specific number. I do recall that it was similar to previous generation JVCs in the 20,000:1 range.

Mark and Tom,

Would be interested in hearing your comparisons of the pic from the Mits compared to that of the Vango (independent, if possible, from the higher brightness of the Mits).

PS Tom, the lumen #'s you report sound very similar to the JVC RS20 and 25, right?

Mark, If you get 14ftL from your 110" diag 1.3 screen, I calculate this to be 385 lumens that your Vango is putting out. Seems a bit less than the 400-500 lumens that you've often quoted. ?

Bill. You will get to see them all on Tuesday and Weds. We are going to dinner with Tom and my wife on Tuesday. Viewing before and after dinner. Tom keeps the measurements and I try to remember what they are for my breath taking narratives. I think that I may have stated 500 when it should have been 400 from the average Vango, with mine for some reasons being about 10% lower than the others we calibrated.

I can hardly wait! See you tomorrow afternoon (assuming I don't get lost!).

Hi Tom, the measurement of 17,000:1 was obtained _without_ the dynamic iris by using a lux meter which gave me 0.5 lux for 0 IRE and ~8500 lux for 100 IRE.
With the same measurement technique I got around 32,000:1 for my HD750 at the same mounting position.
And for comparison, projectors I owned or had for testing at my place:
- IIRC 2800:1 native for the Planar 8150 and around 12,000:1 with Dynamic Black engaged.
- Around 8000:1 for the Sharp Z21000 in high contrast mode.
- Around 14000:1 for my HD1/Dreambee

Due to the higher brightness this means the blacklevel of the HC9000 is roughly three times as high as that of the HD750 with the aged lamp of 650 hours. Sounds dramatic, but it is only visible in very very dark scenes.

Now if the HC9000 had only 5000:1 on/off, this would mean at least 10x times the blacklevel of my former HD750. Now this would be definitely annoying...

Kraine also measured ~ 19,000:1 native contrast http://www.audiovideohd.fr/tests/279...-HC9000-5.html

What settings did you use for this? As I said, I could get considerably higher contrast by simply pumping up the peak output. But I measured a fairly consistent black level of around 0.01 cd/m2 off the screen.

Settings were:
- high lamp, iris fully open, Zoom ~1.6
- Cinema preset, 6500K CT preset, cinema gamma, cinema filter engaged.
Measurement was taken ~1m from the lens with the lux meter.

Until now I had no time to a real calibration to match the colortemp. to my Cheaptrick. I plan to get this done once I have some days off.
After all I have ChromaPure and the Chroma5 at hand.
I hope the Chroma5 is doing fine with this projector? I read something about problems with the X3/X7 and some colorimeters due to the new lamps but I can't recall where

O.K., I remeasured the Mits in Mark's theater instead of in my living room, and I got much higher figures.

The only explanation I can offer is that I have a rack of equipment behind me, which was on while I took the measurements, and the LED lights from that equipment must have been enough to effect readings off the screen. These LED lights are so dim I didn't think much about it, but when you measure very high contrast projectors you must bee VERY careful about even the smallest light contamination.

Anyway, here's what I got. Numbers are in cd/m2.

Lamp	Iris	Black	White	CR
High	Open	0.007	104	14,857
High	1	0.005	81.7	16,340
High	2	0.005	55.6	11,120
High	3	0.002	24.9	12,450
Normal	Open	0.007	77.2	11,029
Normal	1	0.004	65.5	16,375
Normal	2	0.004	45.5	11,375
Normal	3	0.002	21.8	10,900

In addition to these presets, it also offers a slider where you can make fine adjustments. I did not test these. There were just too many options. Throw will also affect CR somewhat.

Iris 2 was the Goldilocks setting for Mark's screen. A setting of 1 gives even higher contrast, but was too bright for this StudioTek.

This thing looks real good. Sharp with good blacks for a digital. Tom's measurements in his lab just didn't reflect actuality.

I concur with Mark, who so graciously hosted a visit to his house Tuesday evening. I was very impressed by the Mits, especially its sharpness and motion handling capability, both of which struck me as significantly better than my (2 yr old) RS20. And though its o/f CR is still not quite as good as the RS20, it is nevertheless very good (see Tom's post above).

It was also great fun meeting Tom H. and getting to discuss ChromaPure with him!

so the tough question is, is this a better projector overall than the RS50? The lumen output sounds like no competition, especially in 3D mode.

The RS50 contrast advantage goes away since I can't close down the dual iris given the limited lumen output to begin with. The CMS is also a bit hobbled, so I would have to rely on a mini 3D anyway if we're comparing the cons of each projector.

The HC9000 lumen output is exactly what the RS50/X7 should have been based on their advertising of 1300 lumens which is a measurement with the color profile=off, nowhere near 8500k, it's something off the charts and no one would watch content in with this strong greenish/blue mode.

cheers to Mitsubishi for putting out respectable lumens much closer to their advertised spec, jeers to JVC this year for the overzealous advertising which has now been removed from their US website.

We don't have a RS50 in my HT to run a comparison. If someone in the area has one, I would be glad to run A vs B. Based on what I remember, I have my recommendation but you will have to call me for it.

Just curious, which light meter are you using? I didn't know any of the low cost meters were accurate enough to measure contrast on such high-contrast projectors.

Hi guys, I'm testing a unit for RapalloAV in NZ, so far I agree with Toms report and findings. Although I managed to get below de94 of 1 from 20~100%.

Two main findings though, the CMS, it can work, but only with the old fasioned mark one eyeball, filters and video generator decoder check pattern.
Pulling staturation down to -20 across the board allows for the gain to start working, which means the whole works is being over driven some what.
From the saturation and gain adjustment only, they just work enough to balance primarys and secondarys to be in balance to white and the filters. Not the perfect solution, however it can be about balance as much as absolute accuracy. Yellow was still out a little which could be seen in an image, but only if you are looking for it.The result does not show up on gamut charts though as overdriven thus Still broken but can be made to look reasonble without a lumagen.

The cinema filter, actually this I found to break the gamma correction tables, running brightness mode, open iris and controling light level with contrast and gamma controls i dialed in a gamma of 2.3 across the board(subjective setting).
The gamma control is interesting with minor settings low(great), but needs a few more points in the middle area. But like I mention to make it work you can't really use the cinema filter.

Many ways to the same sort of results here.

RapalloAV has a lumagen mini on the way to me, see what I can do with that combination.

Well Tom and I will have to check out the CMS a little more. We just moved the sliders and nothing changed returning them to where they were. We never thought one would have to put saturation way down to let the other controls do anything. Thanks for the help.

I have just posted some further discussion of my impressions of the Mits over in the Vango thread.

Voltcraft MS1500, around 70 from a german shop.
Of course I can't measure the on/off directly at the screen. Thus
I have to find a place where there is enough light from IRE0 so that the meter is still accurate enough. I prefer 0.25, 0.33 or 0.5 lux position to make calculations easier. You get a lux value for IRE100 that is still low enough so that it does not oversaturate the meter.

Not sure I understand what you mean to say here.

It sounds like you were successful using the CMS controls to correct any decoding errors, but were unable to correct the extreme oversaturation that the projector produces. I too was able to get nearly perfect luminance values and the hues looked reasonably good after a grayscale calibration. However, the really large color errors in the PJ are in saturation, and the CMS just doesn't help here. Even after calibration the dEs for color were still quite high (except blue and magenta, which were reasonably good).

Here's the post-calibration HSL results.

Just to clarify by after calibration you mean calibration using just the projector. After using the external CMS in the Lumagen, almost perfect results were obtained.