Official JVC RS20 / HD750 Calibration and CMS thread (NEW FIRMWARE V1.1) - Page 9

Originally Posted by JeffY

Filters on the own maybe, but using a combination of (imperfect) colour meter, test screens, filters and eyes you can do better than an inprecise colour meter on it's own.

Originally Posted by TomHuffman

All filters do is give you a visual cue about how to set the brightness of the primaries and the hues of the secondaries. They might have some marginal value with secondary hues on some displays, but even inexpensive colorimeters are quite accurate in their measurement of color brightness.

Originally Posted by Mark Petersen

That's what I use. It works well, even into the low IRE's.

Originally Posted by Scott_R_K

Do we have a working Link to Gregr's Spreadsheet ? Can't find it here and I thought he posted it on the Accupel Site but can't find it there either .

Thanks in advance,

Scott....................

Originally Posted by stereomandan

Yes, there is a 5% error from one dE calculation to the other, but the overall difference in dE is only 0.2 at worst case from one calculation to the next.

Originally Posted by Manni01

It wasn't everyone, and I wasn't upset, just curious to understand the differences.

Now I'm also ashamed you have spent so much time on this.

I guess next time I'll make it clearer it's about the pursuit of knowledge rather than criticism, which it wasn't.

Thanks, and sorry.

Originally Posted by TomHuffman

See Greg R's comments below about using filters with pseudo primaries.

http://www.avsforum.com/avs-vb/showp...&postcount=206

In such cases, filters can yield grossly inaccurate recommendations. Not a good idea.

Originally Posted by JeffY

KT, I think there is a fair bit of variance between projectors but nowhere near as much as you get with grey scale where the lamp has a huge influence. Since I know my colour meter isn't accurate I used Manni's CMS settings as a starting point. I believe I'm a pretty good judge of hue. Ramps and filters help dial in saturation and brightness. Manni, has offered to lend me his i1pro at some point so I'll be able to see exactly how far I'm off. In any case I only take rec 709 as a point of reference, I think it's possible to get a few percent extra performance from the display by not follwoing it ridgedly (all IMO of course).

Originally Posted by gregr

OK, for everyone that was bothered by the 0.1-0.2 dE differences between the various spreadsheets and calculators, I increased the internal precision of the calculations in the Display Calibration Calculator to 16 decimal places !!! That should be accurate enough for government work and keep everyone happy! .....

Originally Posted by Manni01

I just select the HCFR program window, do ALT-Prt sc to save the screenshot, launch MS Paint (or any graphic editor) and do CTRL-V to paste the screeshot. I then save it as .jpg using paint. You can use Prt Sc by itself, but it takes a shot of the whole screen instead of the currently active program window when in combination with the alt key.

Originally Posted by KTTV Images

Thanks Jeffy. Can I assume then that after using Manni's settings as a starting point (I assume attempt #3) that you used your own meter and then at the end used the filter approach to do the final tweeking. Or, did you just use the filters and your eye?

KT

Originally Posted by Manni01

First of all, thanks for all the help regarding my meters dilemma.

I have solved it by opening a second i1pro that has been sent to me by mistake and that hasn't been collected after 3 weeks and many emails.

It returns exactly the same values as the one I have.

As my eyes also tell me that the i1pro is right, and as the greyscale it returns is superior (ie more neutral) to the greyscale returned by the d2, which is slightly on the warm side, I have decided to put the matter to rest and declare the i1pro the reference and the d2 the off one, which is logical.

The consequence is that THX, at least on my PJ, is seriously undersaturated, but that also tallies as you all know, with my experience of THX (again, on my PJ).

I have put my new lamp back in to clarify that this was not an issue either, so this calibration is done in exactly the same condition as my first attempt (new lamp, i1pro). The main thing that's gone is the human error (too much brightess on green, which induced the slight lack of linearity in my first attempt).

So I have redone a greyscale (just color temp, no work on gamma yet, apart from selecting a 2.3 custom default) using the i1pro. I reckon the greyscale is now flat enough for not being an issue (I made sure I was as close as possible to D65 at 100 IRE).

I have then recalibrated the gamut at 75%, as I realised that it needed much less correction at this level of stimulus to get an almost perfect gamut, and the gamut at 100% is just as good.

I have also run the saturation measures for Stereodaman to dissect.

The conclusion is that the new firmware gets a clean bill of health.

It's as linear as one can wish (I don't think anyone has reported a consumer product with such a good linearity).

And it's as straightforward to use as one could wish, as it is now working as intended. There is very little interaction between the controls, which are now reasonably separate.

Based on my experience, I would just warn about pushing brightness too high, which I did by mistake in my first attempt and which generated the initial slight lack of linearity on green, and to calibrate at 75% rather than 100% as the results are at least as good, with less correction applied which I assume is better.

Here are my settings for those who are not entirely put off by the meter dilemna:

Custom color temp (new lamp)
Gain R = -30
Gain G = 0
Gain B=-58
Offset R=0
Offset G=-4
Offset B=0

CMS (HSB)
R= -4 -22 3
Y=6 -46 37
G=-11 -43 31
C=-3 -51 39
B=30 0 -7
M=1 0 -5

I wish I had posted this as my first attempt, and hope you'll forgive me for the confusion, but hopefully this sets the record straight.

Note: the greyscale and saturation data is in the 3rd attempt HCFR file, as this was my main calibration. The 100% just shows what you get at 100% stimulus, and only the gamut info is relevant in this file.

Originally Posted by frohlir944

can someone please tell me which cable from new egg to use, (is it usb to rs 232?) if yes which one? also they all seem to be wrong male/female wise. Please help with actual firmware steps. THX

Originally Posted by TomHuffman

I just recalibrated the unit I originally did in January.

First, the firmware upgrade is a pain in the neck. The first PC we tried wouldn't recognize the display. The second we tried recognized the display, but the upgrade wouldn't complete. We finally got it to work by downloading the .exe file, renaming it (because it didn't have an .exe extension), and then unpacking all of the files into the default folder.

Second, when finished I remeasured the THX mode and User 1. Then I used the CMS to correct User 1. The grayscale at 6500K was so accurate no adjustments were required. I put the gamma at 2.3 and made some minor adjustments at 3 or 4 points.

When I had completed this, the RS20 had the most accurate color I have ever measured from any display. I have a more accurate spectro now than I did in January, and even the THX mode measured incredibly well. For the great majority of people this mode would offer sufficient accuracy assuming my unit was typical. It may not be.

	User 1
	R	G	B	Y	C	M
	0.670	0.279	0.143	0.445	0.200	0.324
	0.330	0.711	0.047	0.549	0.324	0.143
	0.242	0.689	0.068	0.923	0.755	0.303
CIE94	8.3	10.4	3.6	11.2	4.5	3.1
	THX-Now
	0.642	0.307	0.148	0.421	0.230	0.323
	0.328	0.583	0.062	0.500	0.328	0.157
	0.211	0.734	0.075	0.902	0.802	0.296
CIE94	0.4	1.5	0.9	1.4	1.2	1.0
	THX-Before
	0.652	0.310	0.141	0.427	0.224	0.315
	0.322	0.592	0.059	0.502	0.324	0.149
	0.202	0.727	0.075	0.902	0.815	0.288
CIE94	1.5	1.7	1.6	2.0	1.8	1.0
	CMS
	0.642	0.299	0.151	0.418	0.225	0.322
	0.329	0.598	0.060	0.505	0.330	0.155
	0.207	0.724	0.075	0.923	0.784	0.285
CIE94	0.6	0.5	0.7	0.3	0.3	0.1

The default User 1 is even MORE oversaturated than the RS1s I have worked on.

I thought that a calibrated RS20 looked even better than the Planar I looked at a couple of weeks ago. It has better depth and a less noisy image. This only reinforced my belief that ANSI contrast plays a small role in perceived depth. The ANSI contrast on the RS20 is mediocre, but the illusion of 3D that it delivers is the best I have ever seen. I didn't notice the noise in the Planar at the time, but it became evident when comparing it against the RS20.

To my knowledge the RS20 is the best display of any type for calibration controls and if I have seen a projector that throws a better image I can't think of what it was.