Offical JVC DLA-RS4810 Owners Thread - Page 27

I did two calibrations yesterday with my RS45/Mini/ChromaPure, one for my 1.78 pic and one for 2.35 (I am doing the zoom method on a 2.35 screen). I calibrated my 2.35 using standard color and my 1.78 using Wide 1. Besides the bit of under saturation like you mention (which is on my RS45 as well), the color calibration came out better overall using standard color with no noticeable issues on paper or in viewing. Using Wide 1 for whatever reason, things looked good for the most part, but there were a few spots that were a bit out of whack (cant tell you what exactly now) on paper and in viewing last night things looked a bit off at times with a bit of red push it looked like. My 2.35 setting in comparison where I calibrated using Standard looked excellent. I am going to recal the 1.78 using Standard in the next few days or so since I dont know why Wide 1 gave me some minor issues in comparison. Neither setting gives "perfect" color calibration though for me, but Standard is closer overall.
Edited by Toe - 1/19/13 at 10:17am

Yes, mine is undersaturated as well. I don't believe the problem is fixable with Firmware and even if it was I doubt JVC would release such a fix. But hey, JVC word at CEDIA was they widened the colorspaces this year so we could fully calibrate it....go figure.

That said if your Green ∆E ≤ 3 you should be fine. If not you may want to try using the JVC CMS to see if you can minimize the problem (it didn't help much in my case but it did help some).

EDIT: Caution.....if you're using a Lumagen and also want to try tweaking the JVC CMS it is absolutely essential that you either A) Reset the Lumagen CMS or B) Use a Lumagen CMS that does not have any gamut corrections (ie, a blank, or unused, one). The reason for this is: Patterns generated by the Lumagen are altered by the Lumagen CMS (unless you use the Lumagen reference patterns which Chromapure does not use). When you choose a blank Lumagen CMS the patterns aren't being altered and if you're adjusting the JVC CMS you don't want the patterns to be altered by the Lumagen CMS.
Edited by Geof - 1/19/13 at 10:44am

If it's X35 or X75, they have wider gamut and lumagen would be useful to correct CMS but RS4810/X55 does not have wide gamut. I mean the color profiles are close to REC 709 but there's no oversaturated green or cyan in any color profiles to apply Lumagen.
For me, the best dE for Green with RS4810 was 3.5 which is not satisfactory at all compared to Lumagen's perfect CMS with prior JVC PJs .

Thanks. I already reset Lumagen's grayscale & color gamut to default before running Chromapure. Do you get ideal result or less than 1 dE when running JVC CMS and Lumagen together?
Edited by freebits - 1/19/13 at 10:59am

No, mine is around ∆E=3 but a ∆E of 3 is barely (if) noticeable.

You're preaching to the choir about this....I had a similar issue with my RS55. And to add insult to injury most of the other JVC models this year have a fully correctable colorspace (though it does vary a bit). Sadly, most of the comments I've read to date indicate the RS4810's have a green under-saturaiton problem.

It was my 2nd calibration after exchanging the 1st due to convergence issue. Both of them have green/cyan problem.
Lumagen is useless in this case
Edited by freebits - 1/19/13 at 11:28am

It's true that the Lumagen cannot expand a colorspace. The only chance of being able to expand the colorspace is with the internal JVC CMS. Unfortunately it came up short as well but maybe yours will be different (though I would not put good money on that). The Lumagen did correct some problems for me though (notably Yellow, Magenta, and a 75% green hue error was reduced). Cyan in undersaturated because Green is and neither the internal CMS or the Lumagen helped much with that. The beauty of the 125Pt cal is that it calibrates 25, 50, 75, & 100% saturations. So it can help reduce some errors that would have been uncorrectable otherwise. I wouldn't call it useless but it is not as effective as one would hope given our colorspace issues with the RS4810.

The problem is with the design of the JVCs. Nothing more, nothing less. The choice of the filters is the culprit and one can believe the choice was made to increase light output. i think it was a good design choice given that most would not likely be viewing any colors not possible with the saturation limisation of the green primary.

Hoping that the Lumagen CMS or the JVC CMS could fix this is not reasonable nor does it reflect an understanding. An external CMS or an internal one can not make a projector exceed its design limitations. In real life viewing there is no fix needed.

Obtaining a de of 1 or less is also meaningless because of probe measurement errors. You can measure a de with your meter, but there is no basis to assume given the probes you guys are using that a de of 1 is any more accuate than a de of 2 and besides even assuming a de is accurate, one's eyes can not discern a de difference of 1.
Edited by mark haflich - 1/19/13 at 6:05pm

I'm surprised that the X35 has a wider gamut than the X55 since I assumed they were basically the same projector hardware-wise (except for the e-shift mechanism). Assuming the panels in the X35 and the X55 are the same, and the X35 has a wider gamut, it seems like JVC could issue a firmware update to widen the X55's gamut, right?

I also found green to be undersaturated like everyone else here. I only used the JVC's internal CMS to calibrate my projector and was able to get my dE under 3 but increasing the green saturation control in the CMS obviously did absolutely nothing. It took some fiddling to get the dE as low as it is since the two green controls in the JVC CMS that did work (hue and brightness) didn't behave as I'd expect.
Edited by Schwa - 1/19/13 at 2:10pm

Tthe panels have absolutely nothing to do with it. the panels are basically like 1920 x 1080 water values allowing as eacg crystal or valvue rotates a certain amount of light through. What is determinative is the primary colos, green hitting one panel, red another, and blue the third. The amount of light hitting each panel does not change nor does its colors. That is set by the filters between the light source and each panel. The more intense the filtering, the wider the gamut or the greater the saturation. Its easy to choose a stronger filter when designing the projector, but the greater the degree of the filter, the more light one loses. Filters filter out light, for example the red filter lets red pass, what exact red determined by the degree of the filter.

Makes sense, thanks. I guess what I meant to say is that I thought the optical engines were the same between the X35 and X55 except for the e-shift mechanism, but I guess the e-shift mechanism might absorb enough light to necessitate the use of less, uh, "aggressive" filters in the X55.

Yes. I remember Michael Chen from THX calibration course commented it.
I'm not a perfectionist.The point is green does not move while others do. That means I would have more color error than 3 dE with 4810 anyway.

Why? It's not that green doesn't move, it's just that you can't increase saturation past a certain point. I got my green dE less than 3; in fact, I think everyone who's commented on green's undersaturation has still been able to get their green dE below 3. Why would you be any different?

I am seriously considering upgrading to the 4810, currently I have the RS40, very interested in improved 3D and an end to bulb drama. However I'm having trouble choosing between it and the RS46, as the RS46 is about $800 cheaper. The RS46 lacks convergence, CMS and e-shift. CMS is nice, but the RS40's natural mode is very nearly at Rec 709, so I'm not sure how much difference a CMS would yeild, and I've always had good luck with pixel alignment with my JVCs.

As for the e-shift, I read conficting opinions on it. By and large how do you owners like it?

Oh... below 3 in CIE94? In my case, if I increase saturation, hue is lowered and vice versa with green. Brightness works okay.
Which color mode & profile did you choose?

Actually mine is above 3 (its actually closer to 4 as I just looked at my cal report) and it's not just a saturation error - there is also a hue error. The internal JVC CMS is virtually useless for green as all I can do is move the hue and saturation around a bit . There is some unit to unit variation though and some achieve better results than others. Kinda like convergence and bright corners....just a crapshoot (gee, I lost on all three counts this year, again). I think the maddening part is a CMS implies one should be able to calibrate out error. In the case of Green and Cyan there are limitations that don't allow much, if any, correction. Shouldn't be that way really.....
Edited by Geof - 1/19/13 at 8:44pm

The short answer is how far back do you sit. If you sit 1 screen width or less then yes, eshift helps. If you sit further back than not so much. Personally I think this years eshift is a step backwards from last years version (see Zombies photo's a few pages back).

I struggled with a similar choice X35 verses X55. However whereas for me I already have a Lumagen (for CMS), Darbee and an Isco lens it made it more a question of whether it was worth paying extra for the Eshift. Personally I found it seemed quite similar in effect to the Darbee (which I already had) and pixel density is already really good in my set up with the A lens. Like the Darbee it seems possible to overdo the Eshift settings and cause side effects on the two demo units I've seen myself. It is more adjustable than the Darbee and can even be combined with the Darbee as well, so it's an improvement in some ways so long as the settings aren't overcooked.

However, the bigger point being that in the UK the price difference was about £2,000 which looking at exchange rates this morning is approx $3,200 not just $800. That difference really falls into the 'no brainer' area to me and now I see why you guys seem to question us X35/RS46 owners wanting to save a little money especially if you don't already have a Lumagen and a a Darbee. I can't understand how come the difference is so small for you guys compared to double the price for us. For me the X55/RS48 would have to have much higher contrast as well to justify the extra cost, but that model is of course the X75, which is another £2,000 ($3,200) more again.

Kelvin1965s,

We have the same price differential in Sweden. For me it also makes it a no brainer to not go for the X55/X75. However, there are new X70 units available for just a little more than the X35 (£500 or so) and there you get the CMS, extra contrast, lens cover and e-shift1. I am a bit torned with that choice. Though I don't really need the CMS as I have a lumagen mini and the 125p autocal, I would enjoy the extra contrast, would like to have a lens cover and I am intrigued by the e-shift. How much I would benefit I don't know as I'm sitting at 1.3 screen widths. My concerns are with the lamp problems of the older generation. I do like to watch some occational 3D and I understand that the X70 would turn into a ghosting festival as far as 3D is concerned after a few hundred hours? Also, to not have lamp problems would be a BIG plus. There are of course no guarantees that the X35 would far better in these regards, but the inital reports are promising. I have a gut feeling that the real world image quality differences would be very subtle and perhaps not noticable in my situation unless in a direct A/B comparison. In which case the better lamp and 3D would weigh much heavier. But I haven't compared the two so I don't know. What do you guys reckon?

Geoff, Schwa,

Have you seen the small crosshatch pattern after Eshift on? For me, it seems blur, out of focus and emphasizes mis-convergence more. It looks like dirty & blur white lines with smearing colors around it. And all the texts in menu looks distorted as Zombie described before.
And the image would not that sharp if the enhance in e-shift menu is decreased to 0 to get rid of such distortion in text.

I'd go with X70 if I don't own X55 now. But lamp can still be an issue after year or so.

I would go with the x55 because of the better lamp.

I had the green problem solved by using STAGE color profile. When I tried to calibrate my X55 using STANDARD color profile I could not get green right.
Edited by vesa - 1/20/13 at 11:30am

Yes, it does blur it quite a lot. More so than last years eshift1. I sit at 1.6 SW and I could see benefits to eshift1. This year not so much...I have to leave the MPC settings cranked up to see any benefit and then the image looks over processed and more blurry to my eyes. In fact I've been trying it with eshift2 turned off and I think the image looks more natural.

That would be my recommendation as well. I think the lamp issues have been overblown (don't get me wrong, they exist[ed?], but if one goes back and starts counting the number of posts complaining about lamp failure you're going to discover that of the seemingly many complaints a lot of them are by the same folks - folks who've posted numerous times about the same lamp). I also think the latest series 3 lamps are better than the first two versions as the lamp chatter has died down a lot (or did all those RS40/45/50/55/60/65 owners dump their units and buy something different - I think not). In theory this years lamps are supposed to last longer and dim slower. That appears to be true thus far but I think we need to wait until more folks get more hours on those lamps to really know. It may well be that the new lamps will result in the need for fewer calibrations if they do dim more slowly. In the end I agree that the new lamps are better but I'm not convinced there's a huge reliability problem either way.

E-shift implicity eliminates seeing any pixel grid. E-shift overlaps two sets of 1920 x 1080 pixels the pixels are biger than the space between them on each frame, by sequentially flashing each frame one overlaping the other, 3840 x 2160 pixel visable to your eyes. these pixels are about 1/4 smaller than the original pixels which makes sense because you are cramming 4 times the pixels on the same screen size. a tiny spot between 4 created pixels is not illuminate and I suppose you could call this a grid but essentially the overlap eliminates the grid.

One other consideration that doesn't get mentioned much is that, in my observation, motion blur seems to have been greatly improved. Maybe I had a horribly jerky jerky Rs40, I don't know.

Ah, great for you!

I measured most all color profiles and here are the initial out of box Green errors for my unit:

Note: The film color profile is only available in the film Picture Mode and I did not measure this - yet.

For me, the best color profile was animation mode but still dE of green and cyan is more than 3 after calibration.
Edited by freebits - 1/20/13 at 5:25am

Sounds more and more like I might be better off with the RS46 and investing in a Lumagen eventually. What is this Darbee thing? And why does everyone seem to have one?