problems with my new NEC P221W

[taschin]

<p>Hi,<br />I bought, recently, a NEC P221W LCD monitor. I chose this model for the 10 BIT LUT <br />programmable, unique for a LCD monitor at this price. I bought, then, a spyder 3 pro colorimeter and I tested the hardware calibration with the SpectraviewII software.<br />My OS is windows XP media center edition, and the graphic card is a NVIDIA GeForce 7300 LE. <br />The monitor is driven with the DVI output.<br>

I thought that in the native configuration the monitor had values of white point and gamma close to D65 and 2.2 respectively. I am very surprised to measure a white point of 8070 K in native configuration and yet a contrast ratio never above 950:1 (the black point is never below 0.17 cd/m2, is it normal?). Contrary, the gamma value seems to be around 2.2.<br>

If I observe a grey gradient in native configuration this is perfect (even if I can distingue very well the 256 step of grays: A first question: is it normal that I can see the RGB steps? are 8 bits not enough?). The gradients that I observe are those shown for examples in the eizo's test software, or those shown in the same spectraviewII software --> tools --> test pattern.<br />After the calibration D65, 120 cd/m2, black level: default, and g=2.2 the gradients are worse and if I calibrate with a TRC of L* the gradients are even worse, especially in the black tones where clear bandings are visible and some color dominant is present.<br />I expected that by the hardware calibration the monitor response were improved. Does It <br />depend on the only 10 bit internal LUT? I'm making a mistake in something? Could it <br />depend on my graphic card?<br>

I need help to know if the behaviour of my monitor is ok or not.<br />Thank you very much<br>

Andrea Taschin</p>

[Ed_Ingold]

<p>It's possible you are pushing your graphics card to extremes to achieve your target values. It's better to get "in the ball park" using controls on the face of the monitor. I'm not familiar with the Spyder 3 Pro, but I can relate the general steps I follow.</p>

<p>I'm using X-Rite Photo to calibration my P221W, which gives me the option of rough-tuning the monitor using its hardware/firmware controls, while reading the values through the spectrophotomer. I had to play a bit with this to get the white balance and gamma close to the desired results. Beyond this point, the calibration uses the graphic card to make fine adjustments, which are very close to the desired targets. The corrections are charted at the end of the calibration cycle.</p>

<p>If I make the hardware corrections up front, the RGB curves are practically flat. The NEC P221W is an extraordinary monitor (compared to my Viewsonic VP2250wb). Both are touted as intended for professional graphics and photography. The NEC has no observable color changes due to the angle of view when I move side to side, and is much smoother in its presentation of gradients. I see no banding in either monitor, operating at 32 bit color depth.</p>

[digitaldog]

<blockquote>

<p>I thought that in the native configuration the monitor had values of white point and gamma close to D65 and 2.2 respectively. I am very surprised to measure a white point of 8070 K in native configuration and yet a contrast ratio never above 950:1 (the black point is never below 0.17 cd/m2, is it normal?). Contrary, the gamma value seems to be around 2.2.</p>

</blockquote>

<p>I’m not sure how you know this is indeed the native behavior of the display but the values are not surprising. What is key here is what you want the values to be, not what the “native” values out of the box or with some factory default says. Pick a Luminance (cd/m2) based on the print viewing conditions by the display, as well as a white point for a visual match and ask the software to provide that. 2.2 TRC gamma is fine. Adjust to taste (the right values are those that produce a match). </p>

[chas_frady]

<p>Here's a good illustrated tutorial:<br>

http://www.imagescience.com.au/kb/questions/142/How+To+Calibrate+An+NEC+Monitor+With+SpectraView+II</p>

[taschin]

<p>Thank you, Edward, Andrew and Chas. <br>

<em>.."Beyond this point, the calibration uses the graphic card to make fine adjustments, which are very close to the desired targets."</em><br>

But the hardware calibration should be important just to not affect the graphic card LUT so avoiding banding. Is not it so?<br>

<em>"I’m not sure how you know this is indeed the native behavior of the display but the values are not surprising. What is key here is what you want the values to be, not what the “native” values out of the box or with some factory default says. Pick a Luminance (cd/m2) based on the print viewing conditions by the display, as well as a white point for a visual match and ask the software to provide that. 2.2 TRC gamma is fine. Adjust to taste (the right values are those that produce a match)."</em><br>

Andrew, I inferred the native behaviuor just measuring the white point and the black level in N modality of RGB tag of OSD, and for the gamma value just seeing the corrections after the calibration. Is it correct?<br>

I'm agree with you about the procedure, but what I observe after the calibration is a worsening of the gradients.<br>

And what about the 256 step of grays? Is it normal that can I see them?<br>

Some questions about preferences in spectraviewII: first, in the calibration tag, calibration priority, what is better "Maximize Contrast Ratio" or "Best Grayscale color tracking"? second, in the ICC profile tag, source of primary color... section, what do I choose, "Autom..", "calibration sensor measurement" or "Factory measurements"?<br>

Generaly I choose the option "calibration sensor measurement", but the primary colors are very different, expecially the green, from those measured for example by X-rite colormunki in the link suggested by Chas.<br>

Thank again.<br>

<br /><br>

<br /><br>

<em><br /></em></p>

[digitaldog]

<blockquote>

<p>But the hardware calibration should be important just to not affect the graphic card LUT so avoiding banding. Is not it so?</p>

</blockquote>

<p>Its a high bit (internal) panel so no, it doesn’t. </p>

<blockquote>

<p>I'm agree with you about the procedure, but what I observe after the calibration is a worsening of the gradients.</p>

</blockquote>

<p>It shouldn’t affect gradients again, its a high bit panel. You are seeing this in ICC aware applications like Photoshop (at 100%)?<br>

Best Grayscale. Automatic for most devices. </p>

[taschin]

<p>Yes, I also made the gradeint by PS, but being careful to avoid a monitor compensation (that is, I assigned the same monitor profile to the gradient image). It is rigth to do this?<br>

What I observe in native configuration is a perfect gradient. After the calibration I see ligth reddish bandings in the dark tones. These are not many, but the gradients are not perfect. Do you have any suggestions? Could the monitor not work properly?<br>

Thanks again Andrew.</p>

[digitaldog]

<blockquote>

<p>I assigned the same monitor profile to the gradient image). It is rigth to do this?</p>

</blockquote>

<p>Yes, if you want to see the effects without DISPLAY USING MONITOR COMPENSATION.</p>

<blockquote>

<p>Do you have any suggestions? Could the monitor not work properly?</p>

</blockquote>

<p>Well the P221W doesn’t have all the bells and whistles of the higher end spectraview’s (colorComp, AutoLuminance) and is not an ISP panel like the higher end units. So its possible this is the best you’ll get off such a unit. The closer you can drive it native, the less of this banding you’ll see but in the end, you want to set the target aimpoints to match a print. That’s a bit more important. <br>

<br /></p>

[taschin]

<p>Ok, now I understood that this behaviour seems typical for this monitor. So, do you confirm that it is normal to have light banding in 221W when we move away from the native configuration? I'm still curious to see an other P221W at work! About the primary colours, is spyder 3 able to measure them correctly or should I choose the "Factory measurements".</p>