- This topic has 4 replies, 2 voices, and was last updated 7 years ago by
.
-
Posts
-
Hello! I arrive here after buying an Acer XB271HU to sit between my two existing Dell U2515H. I noticed the white was noticeably gray on the Acer and started reading all about calibration and hardware choices. I messed around with importing ICC profiles from TFTCentral but was not very happy with the result and decided that after spending ~1200 on monitors warranted another 150 or so on calibration.
For starters, here is two pictures of my monitors. these were taken in as low light as possible, with all monitors at default/standard settings and brightness set at 35/100. The monitors appear very close already, with the Acer seeming to be just a bit duller.
So my question are the following:
- I am stuck between the Spyder 5 and the Colormunki Display. I don’t know / cant find enough info to determine what is “better” hardware or why it’s better. I can get a Spyder express for 120$, or a Colormunki Display for 130$ used 160$ new. I am leaning towards the Spyder because I am doing this for personal enjoyment and it is not a business critical use.
- I understand ICC profiles as something that nudges the color in the video signal to compensate for mistakes/inaccuracies of the display, similar to paint tinting at home depot to match an existing paint? However I do not know about 3D LUT files except I thought it was only for hardware that would store it on the display itself. Do I need to be concerned with 3D LUT for simple monitor calibrating?
- How exactly does DisplayCAL and other software work out matching different monitors together? Do I just calibrate each monitor and hope for the results to line up or does it compare data between calibrations?
Thanks for reading/replying and also for the open source software!
Calibrite Display SL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Hi,
I am stuck between the Spyder 5 and the Colormunki Display. I don’t know / cant find enough info to determine what is “better” hardware or why it’s better.
the ColorMunki Display has a light gathering lens (good for measurement speed and precision) and its filters are a good match for the CIE 1931 2° standard observer, making it suitable for a range of different display technologies. The hardware is the same as the i1 DisplayPro, but the latter measures a bit faster because the ColorMunki Display has an artificially imposed minimum delay of one second per read via its firmware (not overridable). Unless you want to do refresh (e.g. Plasma) measurements, these two devices are virtually identical and highly recommended, there is no other colorimeter on the market with equal performance at a sub 4000 $ price point.
The Spyder can’t read as low, is less precise and considerably(!) slower.
I understand ICC profiles as something that nudges the color in the video signal to compensate for mistakes/inaccuracies of the display, similar to paint tinting at home depot to match an existing paint?
No. ICC profiles just describe a response characteristic, i.e. in the case of display profiles, they record the display response in the state it was during the measurements. This allows ICC color managed programs to accurately display imagery in other color spaces (provided they are described by an ICC profile as well), by on-the-fly linking the image profile with the display profile (this is called a color transform). ICC profiles by themselves don’t adjust anything, but display profiles may be accompanied by adjustment data for the video card (also known as video card gamma tables or calibration) that must be loaded into the video card by either the OS itself (e.g. Mac OS X) or a small helper program (e.g. the DisplayCAL profile loader under Windows). This adjustment data then has a global effect on everything displayed, but does only affect white point and grayscale (three one-dimensional lookup tables). True color management operates on three (or more) dimensional color spaces. The adjustment data complements display profiling though, as it can make the display response more linear and allows even non-color managed applications to gain some benefits from the adjustments.
However I do not know about 3D LUT files except I thought it was only for hardware that would store it on the display itself.
3D LUTs are a “baked” form of a color transform. Instead of linking two color spaces on-the-fly, the transform is only generated once. As 3D LUTs are typically used in video workflows, where there is not a huge number of possible color space combinations, this lack of flexibility may not be a big limitation. The benefit of such a baked transform is that you are less prone to to be affected by bugs or shortcomings in the consuming software (or hardware) solution – in contrast, with ICC color managed software, different applications, given the same image data and profiles, may produce different results due to bugs, shortcomings or design choices in their color management implementation. This problem can be alleviated by avoiding buggy or incomplete implementations.
Do I need to be concerned with 3D LUT for simple monitor calibrating?
In terms of computer monitor calibration and profiling, 3D LUTs are a special case that you probably don’t need to concern yourself with.
How exactly does DisplayCAL and other software work out matching different monitors together? Do I just calibrate each monitor and hope for the results to line up or does it compare data between calibrations?
A profiler only needs to concern itself with the individual display.
Florian, thank you for your time writing me a thorough response!
I will definitely go for the Colomunki Display, ten bucks is worth it for faster. speed makes a difference when I have 7 computer screens to calibrate.
Regarding matching screens, let me ask in another way:
If I calibrate two different monitors and the software says each one is as accurate as they can get, but sitting side by side they are still visually off from each other, how does that work? If one screen has a noticable blue tint to everything, or faded looking greens?
I guess my end goal is not perfect accuracy but to have all three of my monitors as close to matching as possible.
If I calibrate two different monitors and the software says each one is as accurate as they can get, but sitting side by side they are still visually off from each other, how does that work?
This usually comes down to a visual whitepoint mismatch, i.e. while the instrument measurements may indicate that the white points agree numerically, the visual appearance may not – while the measured CIE color coordinates may match, the spectral distribution may not, and instruments “see” color differently than an actual human being. Instruments use filters (in case of colorimeters) or a diffraction grating coupled with an array of sensors (in case of spectrometers) and math to compute CIE color coordinates from the measured light intensities. The CIE 1931 2° standard observer color matching functions are generally used as a basis for this, which are modeled as an average of human vision. Individual humans may have a different spectral sensitivity than a standard observer model (which is to be expected, as the model can always only be an average). This is less of a problem when the emitted light spectra that are measured are relatively uniform, but if they are peaky, which is often the case with the various display technologies, then what the instrument sees as the same color will more likely differ to what the eye perceives. The work-around to this situation is to match the white points of the displays visually, and then use this pre-adjusted white as calibration target (“As measured”).
For a good color match it is also important that both screens have a gamut that is large enough to encompass the gamut of imagery to be viewed in color managed applications.
Florian, thank you again for the detailed response, this information makes sense and helped me understand what I am getting myself into!
I did some researching about my display specs and they are almost identical:
Acer XB271HU (Center Display)
- AU Optronics AHVA-IPS Panel
- W-LED Backlight with DC control, no PWM
- 10bit + FRC pixel depth
- ~100% sRGB, ~72% NTSC color spaces
- 350cd/m2 brightness
- 1000:1 contrast
2x Dell U2515H (Left/Right Displays)
- LG Display AH-IPS Paenl
- W-LED Backlight with DC control, no PWM
- 8bit pixel depth
- ~99% sRGB, ~72% NTSC color spaces
- ~350cd/m2 brightness
- 1000:1 contrast
I also found this article from another product site about perceptual color matching and white point matching: http://www.lightillusion.com/perceptual_colour_match.html
I believe this is what you meant, so when my Colormunki comes in this weekend my plan is to set the Acer center display as accurate as possible, and then adjust the two side display to be as close visually as i can before calibrating them.
Thank you again for the great help, I will be doing tons of tinkering but its nice to know where to start!
