Calibrated and profiled Dell u2312hm (spyder express3+dispcalgui)

Started Feb 25, 2012 | Discussions
standish76 New Member • Posts: 12
Calibrated and profiled Dell u2312hm (spyder express3+dispcalgui)

Recently purchased a new monitor, received a spyder epxress3 and have been experimenting with all of the above.

My monitor settings:
Custom R85 G82 B97
20 brightness 75 contrast

The tutorials here have been a fruitful read and helpful, but am wondering why my contrast result is so low--487:1? I know it's not the monitors strong suit, but it's quite a bit off from reported numbers at the TFT website in the 700's.
http://www.tftcentral.co.uk/reviews/dell_u2312hm.htm

Also how does one interpret the Curves graph that Argyll makes?
Report on calibrated display device
Command line:
dispcal.exe
-v2
-d1
-c1
-yl
"-P0.498522895126,0.496108949416,1.99295774648"
-r
Setting up the instrument
Instrument Type: Datacolor Spyder3
Hardware version: 0x0437
Current calibration response:
Black level = 0.25 cd/m^2
White level = 119.57 cd/m^2
Aprox. gamma = 2.18
Contrast ratio = 487:1
White chromaticity coordinates 0.3134, 0.3308
White Correlated Color Temperature = 6455K, DE 2K to locus = 5.3
White Correlated Daylight Temperature = 6455K, DE 2K to locus = 0.9
White Visual Color Temperature = 6268K, DE 2K to locus = 5.1
White Visual Daylight Temperature = 6426K, DE 2K to locus = 0.8

Ricoh Caplio R8
If you believe there are incorrect tags, please send us this post using our feedback form.
Ethan Hansen Senior Member • Posts: 1,186
Re: Calibrated and profiled Dell u2312hm (spyder express3+dispcalgui)

Dialing the monitor's white level down to 120cd/m2 forces it to run at a less than optimal value. These panels want 140 or more to show off full gamut and shadow detail.

Don't get overly concerned about your contrast ratio being below what TFT Central reports. The reason is that TFT Central reported a calibrated black level of 0.16 cd/m2 while you saw 0.25. Contrast ratio is simply the ratio of white level (120 cd/m2) to black luminance.

The calibration software was different - Argyll in your case, LaCie BlueEye in theirs - as were the measurement pucks. Each calibration package differs in what assumptions it makes as to a minimum usable luminance is. Your Spyder3 is more accurate in the shadows than the i1Pro TFT Central used. Their reported 0.16 black level could easily be off by more than the 0.09 cd/m2 difference in your measurement.

-- hide signature --
OP standish76 New Member • Posts: 12
Re: Calibrated and profiled Dell u2312hm (spyder express3+dispcalgui)

Ah. That makes sense. In the review at TFT the authors indicated how different devices and software returned different results in the discussion. Thanks for pointing that out.

Should I ramp up the white value starting with 140 cd/m2 to see if it's comfortable on the eyes and recalibrate to that?

Which testchart should I use on the recalibration? The default for curves and matrix or can I use the existing one for the most recent calibration?

OP standish76 New Member • Posts: 12
Re: Calibrated and profiled Dell u2312hm (spyder express3+dispcalgui)

standish76 wrote:

Should I ramp up the white value starting with 140 cd/m2 to see if it's comfortable on the eyes and recalibrate to that?

I think the below thread you wrote earlier answers my question on the white value. I'll give it a shot above 140 since it seems ok.
http://forums.dpreview.com/forums/read.asp?forum=1004&message=40301298

Chris Noble
Chris Noble Veteran Member • Posts: 3,219
Monitor brightness level

Ethan said:

Dialing the monitor's white level down to 120cd/m2 forces it to run at a less than optimal value. These panels want 140 or more to show off full gamut and shadow detail.

When you say "these panels", which panels do you mean? And where did you read about a min brightness level of 140 cd/m2? Just interested, as this is the first time I have seen this, and my measurements on my panel (not the OP's) don't support your statement. Just asking, not trying to be confrontational.

 Chris Noble's gear list:Chris Noble's gear list
Panasonic Lumix DMC-GM5 Panasonic Lumix DC-GX9 Panasonic Lumix G Vario 7-14mm F4 ASPH Panasonic Lumix G 20mm F1.7 ASPH Panasonic Leica Summilux DG 25mm F1.4 +3 more
Ethan Hansen Senior Member • Posts: 1,186
Re: Monitor brightness level

Chris Noble wrote:

When you say "these panels", which panels do you mean? And where did you read about a min brightness level of 140 cd/m2? Just interested, as this is the first time I have seen this, and my measurements on my panel (not the OP's) don't support your statement. Just asking, not trying to be confrontational.

Chris,

Most LCD monitors that are not very high end (e.g. Eizo ColorEdge, Quato Intelli Proof, and, (barely) the NEC PA Series) do not perform at their best when white luminance is reduced below 130-140 cd/m2, depending on the panel. One of the features you get from paying 10x more for an Eizo than the Dell under discussion is relative freedom from gamut reduction, shadow banding and posterization, and other level compression artifacts at luminance values of 120 and below.

Depending on the screen, you may see visible artifacts when white luminance is dialed down below optimal levels. It can be difficult to determine whether a flaw is part of your image or simply a result of how your monitor displays it.

The danger of using high luminance is the dreaded "my prints are too dark" comment that one frequently sees. There are two solutions to this, both of which start with soft-proofing your images to the printer profile. If you use Photoshop, the Simulate Paper White option should be enabled for maximum accuracy.

The first option is to view your prints under a brighter light. If you are doing commercial work, you likely are doing this already. The ISO specs call for print illumination of 160 cd/m2 for "practical" evaluation. Below that, human color and shadow perception suffer. The soft proofing spec calls for a minimum display luminance of 100 cd/m2 with 160 recommended. If you take this approach, be sure to soft proof your images to the printer profile using Photoshop's simulation options. This avoids surprises due to the limited output contrast range of paper relative to your monitor. See this link for more: http://www.drycreekphoto.com/icc/using_printer_profiles.htm#Simulation

A second technique is to dial down your monitor white luminance until it roughly matches the white level of a blank sheet of paper in your viewing area. As I mentioned above, unless you have a top-quality monitor, this entails compromises in the fidelity of the images you see on-screen. It works, however.

-- hide signature --
Ethan Hansen Senior Member • Posts: 1,186
Re: Monitor brightness level

I should have added the option many of us used back in CRT days when a monitor luminance of 85 cd/m2 was pushing the envelope: Use a viewing booth with a dimmer or a viewing lamp (e.g. Solux or SolSource bulb) that can be moved closer or farther from the print. Again, the goal is to - roughly at least - match the print viewing and screen luminance.

-- hide signature --
(unknown member) Forum Pro • Posts: 14,338
Re: Monitor brightness level

Ethan Hansen wrote:

Most LCD monitors that are not very high end (e.g. Eizo ColorEdge, Quato Intelli Proof, and, (barely) the NEC PA Series) do not perform at their best when white luminance is reduced below 130-140 cd/m2, depending on the panel. One of the features you get from paying 10x more for an Eizo than the Dell under discussion is relative freedom from gamut reduction, shadow banding and posterization, and other level compression artifacts at luminance values of 120 and below.

Very interesting discussion. It was news to me but it does make sense.

I have just recalibrated my U2412M to 140 cd/m2. It was previously set to 100.

We'll give this a try for a few days. I'm concerned that 140 may prove a bit tiring given that I spend many hours/day in front of this monitor -- more text work than photo. Still, I'm sure I can stand a little more than 100. Maybe I'll compromise at 120 or 130 but I'll definitely give 140 a fair shot first.

Chris Noble
Chris Noble Veteran Member • Posts: 3,219
Conjecture...

Most LCD monitors that are not very high end (e.g. Eizo ColorEdge, Quato Intelli Proof, and, (barely) the NEC PA Series) do not perform at their best when white luminance is reduced below 130-140 cd/m2, depending on the panel.

I ask again, what evidence do you have to support this claim? Can you give an example of a specific panel that has poorer color fidelity at 100 cd/m2 than at 130? What are the measurements?

 Chris Noble's gear list:Chris Noble's gear list
Panasonic Lumix DMC-GM5 Panasonic Lumix DC-GX9 Panasonic Lumix G Vario 7-14mm F4 ASPH Panasonic Lumix G 20mm F1.7 ASPH Panasonic Leica Summilux DG 25mm F1.4 +3 more
Pictus
Pictus Veteran Member • Posts: 6,379
Re: Monitor brightness level

malch wrote:

We'll give this a try for a few days. I'm concerned that 140 may prove a bit tiring given that I spend many hours/day in front of this monitor -- more text work than photo. Still, I'm sure I can stand a little more than 100. Maybe I'll compromise at 120 or 130 but I'll definitely give 140 a fair shot first.

Compensate by increasing the luminosity of the environment.

Ethan Hansen Senior Member • Posts: 1,186
Re: Conjecture...

Chris Noble wrote:

Most LCD monitors that are not very high end (e.g. Eizo ColorEdge, Quato Intelli Proof, and, (barely) the NEC PA Series) do not perform at their best when white luminance is reduced below 130-140 cd/m2, depending on the panel.

I ask again, what evidence do you have to support this claim? Can you give an example of a specific panel that has poorer color fidelity at 100 cd/m2 than at 130? What are the measurements?

Let's see... We measured color gamut volume, grayscale neutrality, gamma curve accuracy and smoothness for neutrals and primaries, and tonal separation at minimum luminance levels.

Subjective evaluations used several pairs of highly critical eyeballs, the owners of which get top-end rates based in no small part on their ability to view and edit images accurately on screen. Objective measurements were made using a Photo Research PR-730 spectroradiometer. Details here: http://www.photoresearch.com/current/pr730.asp Our older measurements were made after calibration to a secondary NIST standard, while data from after 11/2011 were made with calibration to a primary standard.

Here are some monitors we have in-house or have otherwise characterized extensively and differences in output between a 140 cd/m2 white luminance and 100. Sorry, but we don't have as much data at 130. All screens are of at least minimal quality for graphics editing. When possible, all were calibrated using DDC and monitor LUTs. For panels not supporting DDC, all adjustments were carried out in the video card. Most were driven through the DVI port, while a couple required DisplayPort to access full functionality.

NEC:

  • PA241 and PA271: ~5% gamut volume reduction. Measurable banding/gamma inaccuracy in color channels, but at the threshold of visibility.

  • EA231WMi: > 10% Gamut volume reduction, banding in gradients, color crossovers in neutrals at lower luminance levels.

  • EA232WMi: Same problems as EA231, slightly worse banding and loss of shadow detail

Dell:

  • U2711: Increase in color crossovers, banding (gamma inaccuracy) in neutrals, blue channel.

  • U2410: Banding in neutrals, reduced shadow separation. Slight (5-7%) reduction in gamut volume.

  • U2312HM: Significant reduction in shadow resolution, banding in both neutrals and color channels.

HP:

  • LP2480zx: Mimimal (~1-2%) gamut volume reduction. Slightly lower shadow resolution - measurable, but only barely visible.

Eizo:

  • CG245W, CG275W, and CG221: Gamma inaccuracy in blue, red channels that could be measured but not seen by eye.

And I could go on, but actual work calls...

-- hide signature --
Chris Noble
Chris Noble Veteran Member • Posts: 3,219
Target brightness

We have characterized differences in output between a 140 cd/m2 white luminance and 100.

I understand your post, but I am not totally comfortable with increasing the brightness of the display as a way of obtaining a broader and flatter spectral content. Running a monitor at too high a brightness level is more often the cause of poor monitor/print matching than anything else.

The brightness of the screen is the product of two controls that can be set independently (as you know): the monitor's Brightness Control (i.e. CFL backlight power level) and the RGB gains. It is conventional to set the RGB gains as high as possible, and adjust the brightness using only the Brightness Control, but a White Point output of 120 cd/m2 (or lower) can also be set by reducing RGB gains and leaving the Brightness control at a higher level. The RGB resolution is reduced of course, but if your panel has a 10-bit internal LUT, you would not see the different with 8 bit video data until the RGB gains are 1/4 of full scale or less.

You need to try different combinations to find the optimum tradeoff, but I have found that I get just as good results with R/G/B all around mid-scale as I do at full scale, and with a resulting overall brightness that can be kept at 120 cd/m2 or even lower without running the CFL power at the bottom of its range. Back to work as well.

 Chris Noble's gear list:Chris Noble's gear list
Panasonic Lumix DMC-GM5 Panasonic Lumix DC-GX9 Panasonic Lumix G Vario 7-14mm F4 ASPH Panasonic Lumix G 20mm F1.7 ASPH Panasonic Leica Summilux DG 25mm F1.4 +3 more
Keyboard shortcuts:
FForum MMy threads