• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Dell U2412M gamma issue

L

lpcstr

n00b
Joined
Dec 29, 2013
Messages
12
I just bought a Dell U2412M and I would like to calibrate it but I have ran into an issue. The display only supports two gamma modes (2.2 and 1.8), the problem being that the 2.2 mode is actually close to 2.0 in reality. As far as I know the only other way to adjust the gamma is with the Windows or GPU (Nvidia GTX 760) color correction software. The problem with that is if I set the gamma adjustment to anything other than 1.0 I get obvious banding artifacts in gradients, which is worse than having the wrong gamma.

So, does this mean I am stuck with the wrong gamma if I decide to keep this monitor?
 
If you have a colorimeter, spectroradiometer, or even a luminance meter, you can do your own gamma correction, using software like argyll cms.
 
spacediver said:
If you have a colorimeter, spectroradiometer, or even a luminance meter, you can do your own gamma correction, using software like argyll cms.
Click to expand...

How is that going to solve the banding issue? What is argyll cms going to do that Windows or the graphics driver isn't?
 
You can set specific targets for your luminance function. Using the gamma slider in your video card driver is a very crude way of adjusting gamma. In your case, it seems to be amping up parts of the luminance function to the point where the difference in luminance between successive RGB levels is too big (resulting in posterization).

You want a nice perceptually uniform change in luminance as you go up from black to white.
 
It's not a question of how much I adjust the gamma. Adjusting gamma even the slightest through window monitor calibration or Nvidia control panel causes the banding artifacts. Both of these tools work the exact same way. They utilize the GPU's LUTs. So unless a program like argyll cms utilizes magic I don't see how it is going to make one bit of difference. If you take an 8-bit source and apply a gamma to it, you are going to get banding. The only way to avoid that is to 1. do the gamma correction at a higher precision using linear color space and then apply dithering when going back to 8-bits (not ideal and I see no way for a 3rd party program to be able to do this) or 2. have the gamma adjustment happen in the monitor, which isn't an option since this thing only has two fixed settings, neither of which is accurate.
 
Last edited:
Do you get banding artifacts across the entire range of input levels? And does this happen regardless of whether you decrease or increase the Nvidia gamma?

The reason that using a more finely tuned approach can make a difference, is that your display's luminance function isn't a perfect gamma curve. It may be lower or higher at various levels of input. If it turns out that, for example, your gamma is too low overall, but more so in the low ranges, then increasing your overall gamma uniformly (by moving the slider slightly downwards) is going to fix the low end (so stuff doesn't looked washed out near black), but will decrease the gamma below target at the high ends, resulting in too steep of a curve at the high end (which would result in posterization artifacts).

By using a more finely tuned approach, you can fix only the parts of the function that need fixing.

edit: you can get posterization artifacts in two ways: either by stretching out the difference between two successive tones, or by clipping too many tones together. Yours sounds like the former.
 
Last edited:
Nvidia control panel allows adjustment of gamma in 0.01 increments which is probably overkill to start with. Like I said, this display has a gamma of approximately 2.0, which means I need to set gamma correction to 0.9 in order to get an overall gamma of 2.2. Banding becomes apparent at 1.01 and 0.99. It has nothing to do with the tuning and everything to do with the method by which the gamma adjustment is being applied.
 
Is the banding visible in low tones, midtones, highlights, or across the entire range?

And if you load up a smooth grayscale ramp, does it become more posterized when you move the gamma slider up and down?

It sounds like something is off software wise.

Try xrite's calibration tester - it's a very simple exe file that will tell you if your video LUT has been tampered with.
 
They are visible more or less through the entire range but they are most obvious in the low end since the human eye is more sensitive to change there. It doesn't matter which direction you adjust the gamma, both look equally bad. The xrite tool isn't going to help me. I've already stated what the issue is. To make it perfectly clear, you cannot take an 8-bit image, apply a gamma (any gamma) to it and then quantize back to 8-bit without producing banding. You must perform a dither operation to hide the banding or better yet perform the gamma adjustment in the monitor using an analog method if possible. If I set the monitor to use MAC gamma (1.8) I get no banding, so obviously the monitor is handling the gamma adjustment correctly internally, but that doesn't help me since it doesn't support a custom setting. That only leaves me with one option, adjusting the gamma using the video cards LUT, and clearly Nvidias drivers/hardware are not performing any type of dither since banding become apparent immediately, regardless of how much gamma adjustment you apply.
 
Well that's another reason to use argyllCMS. It works in 16 bit precision, and can use MadVR as a rendering engine for the test patterns during profiling (which allows the dithering).

I've never experienced banding issues with Nvidia's gamma adjustment, but I've never owned a display that isn't a CRT.
 
spacediver said:
Well that's another reason to use argyllCMS. It works in 16 bit precision, and can use MadVR as a rendering engine for the test patterns during profiling (which allows the dithering).

I've never experienced banding issues with Nvidia's gamma adjustment, but I've never owned a display that isn't a CRT.
Click to expand...

I don't understand. I have never used argyllCMS, but unless it subverts the graphics driver or monitor firmware, then the only option it has to apply a gamma adjustment is though the same exact LUTs, and therefor will still produce banding for the same exact reason.
 
it's worth a shot, see this post here.

You can do a very quick experiment to see if your monitor can handle an argyllcms adjustment without causing the quantization issue.

1: download argyllcms from here. Just extract it to any folder.

2: download this zip, and extract the two files into the \bin folder of your argyll installation directory.

3: in a command window, navigate to the \bin folder, and type: dispwin BT1886.cal

If you want to unload the correction ramp, type dispwin -c (or reboot).
 
That curve actually lowered my gamma to about 1.9, and it displays banding as well.
 
Last edited:
yea that curve was generated to lower my own gamma, so it would lower yours as well. That's a shame that it still creates banding. Looks like with that particular model, you might not be able to effectively implement software gamma/color correction.

Just found this page - apparently your model has a gamma slider in the OSD?


edit: nm, I think they misused the term "slider".
 
It might have something to do with the fact that the display is 6 bit, and uses temporal dithering to achieve 8 bits. Perhaps it's only built to effectively dither with the 2 presets it has.
 
As near as I can tell the only way to apply a gamma correction to a monitor without it looking like crap is to either 1. have a graphics card that performs dithering when converting the 16-bit LUT values to 8-bit, or 2. have a monitor which has a programmable internal LUT with more than 8-bits of precision. It looks like with my U2412M and GTX 760 combo I struck out twice. Aint that swell.
 
If gamma correction beyond what's available on the OSD is important to you, return that puppy :)
 
Returning it would likely be easy. Finding a suitable replacement however wont be trivial. For starters, I was able to get this U2412M from a local retailer. It's much harder to find higher end monitors at these places which means I would have to order one online. If for any reason I had a problem with it, returning it would be a major hassle. That aside, I would still have to find something which offered all the features I want. Specifically, 24" 1920x1200, IPS 8-bits preferred, internal LUT and fine tuned gamma adjustments it would seem. Something like the U2413 looks somewhat appealing, except for all the talk of severe overdrive ghosting I have read about. I plan on using this monitor for both image editing and media/games, so ideally it would have good color reproduction and also relatively low latency.
 
I'm not really up to date on LCD technology, but the Eizo Foris FG2421 might be a good choice if you can afford it (it's a VA panel not IPS).
 
spacediver said:
I'm not really up to date on LCD technology, but the Eizo Foris FG2421 might be a good choice if you can afford it (it's a VA panel not IPS).
Click to expand...

That's a 16:9 display. I really need the extra vertical space. I'm not sure how VA compares to an 8-bit IPS but I know I certainly don't need 240Hz capability. I would probably prioritize photo editing before gaming. I don't game much, but when I do I certainly don't want to have an issue like that reported with the U2413 which is why I mentioned gaming.
 
my bad, i thought the FG2421 was 16:10

FWIW, half of the refreshes in 240hz are black frame insertions, so it's more like 120hz but with no eye tracking based motion blur (strobed backlight essentially). This is great for gaming where you want to preserve detail while panning the view.

As for picture quality, assuming you win the panel lottery, I don't think you'd be disappointed - it has excellent black levels and contrast, and apparently can be calibrated quite well.

But yea, as someone who works with FW900s I know what you mean about 16:10 being important.

I'd make a new thread, specifically asking for a recommendation based on your criteria.
 
You must log in or register to reply here.
Back
Top