Anyone used ColorEyes/BasICColor/Spectraview with NEC *90UXi?

<p>Having just received a NEC LCD2090UXi, I am wondering what the consensus is on the best way to

calibrate the thing. I run an (Intel) Mac Pro with latest Mac OS X Tiger 10.4.8, and use an EyeOne

Display 2 puck for monitor calibration. The video card is NVIDIA GeForce 7300 GT (I guess you can tell I

don't play games). Anyone who has knowledge on any of the following matters, please post!</p>

 

<ol>

 

<li>Contrast setting must be left at default 50%, and only brightness should be adjusted for calibration

to target luminance, right?</li>

 

<li>The monitor is able to automatically adjust brightness to ambient level, which NEC judiciously

chosen to call AmbiBright. Should this be set to "OFF", "1", or "2" for work with color? Does brightness

adaptation mess up the profile? If no, is it better to calibrate via DDC programming or via video card

LUT? For now, I set it to OFF (default setting).</li>

 

<li>NEC *90UXi series monitors support DDC, right? If so, why am I not seeing "One-push button

calibration" option in EyeOne Match 3.6.1 (latest version)? Am I looking in the right place? Does EyeOne

Match support DDC on Intel Macs as of yet?</li>

 

<li>In order to test whether DDC is working, I downloaded BasICColor Display 4.1.0 (14-day trial

version). BasICColor was able to adjust monitor brightness without my intervention, and apparently it

programmed the RGB controls automatically (the RGB OSD setting switched from "Native" to

"Programmed"). So DDC works, and my video card is out of blame. <i>But</i> why does BasICColor

make a mess out of my vcgt curve? I thought vcgt curves were a thing of the past, superseded by

monitor-internal programmable LUTs! To illustrate, here is what BasICColor

produced:<br /><br /><center><img

src="http://eugenescherba.com/files/vcgt_basiccolor.png" /><p>BasICColor vcgt curve (at native

color temperature)</p></center> And here is the EyeOne Match result <i>at exactly the same

settings</i>:<br /><br /><center><img src="http://eugenescherba.com/files/vcgt_gretag.png" /

><p>EyeOne

Match vcgt curve (at native color temperature)</p></center> The funny thing about these curves is

that if EyeOne Match does not support DDC and BasICColor does, an exactly opposite result would be

expected!</li>

 

<li>Adding to my disappointment with BasICColor and NEC 2090UXi combination, the monitor shifts

its

appearance to horrible green-magenta cast when after I switch inputs back and forward! The cast goes

away after I switch to "Native" RGB from "Programmed" RGB, but that, of course, means my BasICColor-

produced profile is invalid! This is simply unacceptable for me, as one of the reason I purchased a

higher-end monitor like this was to be able to switch inputs from the Mac to my Linux server. I should

not have to re-calibrate the monitor every time after switching inputs!</li>

 

<li>There are two alternative software packages available, ColorEyes Display Pro and NEC Spectraview.

Both of them support DDC. ColorEyes Display Pro apparently has just been released in a Mac-only

version to <a href="http://www.luminous-landscape.com/reviews/software/coloreyes-

pro.shtml">rave reviews</a>. The NEC software, on the other hand, was specifically designed for the

*90UXi monitors. But here is the catch: both of these programs are based on BasICColor code! So where

does that leave me in regards to worry-free DDC support? Is Spectraview going to give me the same

headache as BasICColor if I buy it for whatever it costs?</li>

I have the NEC2190UXi and have been using the NEC Spectraview software to calibrate. I find it does an excellent job. It calibrates the monitor directly, makes no changes to the video card, and creates the profile for the systems software. I am not a color expert and am sure that others may be able to provide you more detail. I have no regrets with my choice of monitor or calibration software.

<p>Thanks Morrie, this confirms my suspicion that NEC didn't

disclose some of its specifications to BasICColor in favor of its

own SpectraView. Which is a pity because BasICColor supports many

other monitors e.g. Eizo ColorEdge quite well.</p>

 

<p>My question to you: If you ever used two computers with the

monitor (for example, a laptop plus a PC), were you able to switch

inputs after calibration with SpectraView? If you used just one

computer, don't worry about this.</p>

i have just used basiccolor on a 1990sxi, and it was a massive improvment to the spyder

software that i was using. It didnt give me the option to fully hardware calibrate, but it

looks like it has done something to the monitor.

 

At 100% brightness the monitor used to read roughly 270lux, so to get it to 100lux i had

to take it down to around 40%. After using basiccolor at 100% brightness it is reading

100lux.

 

Am i right in thinking that this is what hardware calibration would do?

 

And will it have done this to the color temp etc as well?

<p>An update:</p>

<p>I have since purchased SpectraView II. After a little trouble of installing it (they

shipped me an older release), I realized there is an updated version online which works

with Intel Macs. The software works very well: all you have to do is to specify the desired

white point (I chose D65), gamma (I chose 2.2), and brightness (I chose 120 cd/m2), and

the rest is fully automatic. I did the standard "grey gradient" test in Photoshop, and found

3-5 times less visible color banding than after a similar calibration with EyeOne Match. I

also tested the new ColorEyes Display Pro and BasICColor, which gave similar results to

that of SpectraView. ColorEyes Display Pro fared about the same as SpectraView, while

BasICColor was a bit worse. None of them were as bad as EyeOne Match, at least in this

gradient test. Only the SpectraView fully utilized the 12-bit LUT, and it was also easiest to

calibrate with, so I'm keeping it.</p>

I hate to belabor the discussion but.....

 

I just purchased a 2690 with SpectraView II and the GM hockey puck 2. I don't want to

become a color scientist. All I want is as good a calibration/profile as I can get so I can get

match prints out of my printer.

 

Everything I've read (elsewhere) says that setting the display to sRGB (6500), gamma 2.2,

black point of .50-.60 cd/m2 and white point of 115-120 cd/m2 should yield good, close

match prints.

 

I've figured out how to get SpectraView to set to the white point (Intensity) but there seems

to be no way to set the desired black point to get the software to backsolve for the black

point also. I keep getting black points down in the .24 range when I set the white point

(Intensity) around 115-120?

 

I've tried going with Native gamma instead of 2.2 (per drycreekphoto's recommendations)

and maximum intensity (wow-that's bright!). I don't know if I should go into the hardware

adjustments with the OSM functions and adjust things or what. Why the heck is this such

an unbelievable mystery? I could do better than this in a darkroom with my eyes closed.

This is ridiculous.

 

The color of this monitor and it's flexibility is amazing. Now, if it just had some simple,

easy to understand, written for photography instructions, it would be perfect. Like I said,

all I want is a match print. Nothing more.

 

Thanks for your help in advance. If you can post screenshots with captions that say "look

at this, click here, set that," that would be lovely. Again, thanks.

<p>Don't listen to Dry Creek -- listen to me instead. As I said

above:

choose D65, gamma 2.2, and 120 cd/m2 for brightness -- the rest is

automatic.</p>

 

<p>No, you cannot set blackpoint, because I would have mentioned it

otherwise. On all LCDs blackpoint is an almost linear function of

brightness (only contrast ratio is fixed). The higher the

brightness, the higher the blackpoint.</p>

 

<p>No, you should not go with Native gamma, because (1) that defeats

the whole point of buying an expensive LCD with an internal 12-bit

LUT, (2) native gamma sucks for non-colormanaged web (even in Safari

CSS elements are not colormanaged).</p>

 

<p>No, you should not go with maximum brightness, because that's not

what I said. Go with 120 cd/m2, unless your office or your workflow

is really crazy.</p>

Bless you, Eugene. Thank you for quick, simple, easy, English, non-technical, non-scientific

instructions.

Oh, I left the contrast ratio to monitor default in SpectraView. Is that what you did?

Yes, contrast ratio should be left to default. Cheers.

Matt, I'd suggest you use a luminance that's around 150cd/m2 as was suggested to me by

NEC. Reason is, the lower limits of this display is around 120 cd/m2 and you need some

fudge factor here should the software want to adjust towards the lower end of its usable

range. The luminance is based on ambient light around your workstation and the print. For

ambient light, you can't go too low! Then, you should have no problem getting any decent

light box to produce around 150 cd/m2.

 

You should go with a 2.2 gamma on this unit not Native due to how it deals with setting

the display internally using Gamma Comp and Color Comp. With lesser LCD's, yes Native

white and gamma are going to produce less banding but this pup has no such limitations.

 

Note that you'll find the color temp is going to be a tad off (about 500K) because this is a

wide gamut unit and the off the shelf Colorimeters don't have filters designed for this type

of chromaticity. Not a big deal. A Spectrophotometer would nail this much closeer, but a

Spectrophotometer doesn't do anywhere as good a job in measuring the blacks and dark

tones. This is why NEC mates a special EyeOne with their real wide gamut LED display. It

has special filters mated to this backlight.

Eugene and Andrew:

 

Thanks again for your help. Andrew, I'm going out to buy your book. Any author that takes

the time to answer a question in a forum like this has earned my support.

 

Using the settings that Eugene provided me with, my test print is about 3-5% darker

(more saturated looking) than the screen. If I up the intensity (white point) to 150 cd/m2,

my print will be even darker than the screen. So, I'm going to try what William

Hollingsworth suggested below. I now know, however, between the help that you guys

have given me and William's answer, I'll get this nailed down pretty quickly.

 

For ambient lighting, I have a pretty dark environment. It's down around 32-36 lux. And,

for print viewing, I'm using a Solux 4700 degree, 50 watt, 36 degree-angle desk lamp that

I got from Harrington Lights in N.H. That seemed to be the broad consensus

recommended poor-man's print viewing light.

 

I finally received an answer on how to set a black point with the NEC 2690 from William

Hollingsworth, NEC's Professional Display Support Chief (basically, he's the top dog at NEC

concerning SpectraView and the *90 series of monitors/lcd's. Here's his answer:

 

You asked how to create a target with:

 

A black level of .50-.60 cd/m2

An Intensity 115-120 cd/m2

 

I'm going to choose 0.50 and 115. The contrast ratio with these values

is 115/0.50 = 230:1.

 

So in the SpectraView application:

 

1. Click the "Edit Target Settings" button.

2. Choose whatever gamma and white point you want. E.g. 2.2 and D65.

3. Select "Specific Level" in the Intensity group.

4. Enter 115.

5. Select 250:1 in the Contrast Ratio (closest to 230:1).

6. Click OK and save the Target under a new name.

7. Calibrate the monitor.

 

Thanks again, William, Eugene and Andrew!

>If I up the intensity (white point) to 150 cd/m2, my print will be even darker than the

screen.

 

Which means you have to adjust the light to the prints accordingly.

 

BTW, it was Will who suggested I use a 150cd/m2 target instead of 120 which I was

initially going to use.

I'm happy to report that after following William Hollingsworth's NEC 2690

calibration/profiling instructions that I posted above using a white level of 120 cd/m2, 2.2

gamma, .60 black level based on setting the contrast ratio at 200:1 (120/6=200), I was

able to produce a match print to a test photo that I had in the AdobeRGB working space

using Photoshop CS3 (with the application controlling the colors) on my HP B9180 that is

probably around only a 1% differential from the screen display. Wah-hoo!

 

I'm going to bet that the negligible difference was attributable to the fact I was using the

HP Advanced Photo Glossy canned profile with Epson Photo Glossy paper (that's all I had

on hand). I'm not complaining.

 

After all the frustration I had with Epson printers and my (former) HP L2335 display and

Monaco EZColor2 with the Monaco Optix XR DTP-94 puck and Color Eyes with the same

puck, I think I'm a permanent NEC/SpectraView customer.

 

Thanks again to everyone. What a great forum and web site!