Questions about screen calibration from a newbie to all this

[Logue]

Hi everyone! I’m new to this whole calibration and color management thing. I bought a Calibrite Colorchecker Display (same as the X-Rite i1Display Studio, the cheapest one from their lineup since I don't NEED it) and am in need of some help. I’d like to start by saying I’m not a professional and do not NEED all of this for a living… I’m mainly doing it because I’m a tech enthusiast and would like to have a (decently) calibrated display. My monitor is an LG 34GL750 (via DisplayPort, don’t know if that matters in some way). It’s my main and only monitor to my desktop, so I use it for basically everything (from gaming – a bit less these days – to movie/show/video watching in general, browsing the web, etc.). Now, I have already done my first calibration with DisplayCAL which I’m using right now and I think it’s pretty good (the actual first one I’ve made with the sRGB preset turned out crap – I think because I set the Whitepoint to “As Measured” or something like that). I’d like to maybe make a few more calibrations now with some adjustments to get more accurate results after I gather some answers to some questions I have, which are as follows:

1. This monitor has 4 gamma settings. I’ve watched the Hardware Unboxed tutorial video on calibrating displays and it’s said there to basically reset the monitor settings and disable any kind of enhancement, which I’ve done: in my case, black level stabilizer is set to 0, no energy saving mode is enabled, no motion blur reduction/backlight strobing, ‘tho Freesync is enabled (also have an AMD GPU). The gamma settings on the OSD are just labeled “Mode (1, 2, 3 or 4)”.  Modes 1, 2 and 3 are progressively darker (1 being the brightest and 2 being the default which is selected after reseting the settings). However, in the display's manual, it’s said that “If it’s not necessary to adjust the gamma settings, choose Mode 4”. So, which should I choose? The “no adjusment” (mode 4) or the one it defaults to after reseting the settings (mode 2)?
2. Should I measure the gamma using the "uncalibrated monitor report" tool in DisplayCAL and then try to reach the value there? Say, if it says 2.43, I should input 2.4? Yes, no?
3. Which whitepoint should I choose? Is D65 the same as 6500k? And is that the same as the sRGB whitepoint? Or should I use As measured? Or am I writing nonsense here? Is this just preference, in my case? Is this just a matter of choice in my case?
4. What about white level, should I stick to the 80-120cd I see recommended everywhere? Before buying this, I always used my monitor at 50 brightness (also the default after settings “factory reset”) and that is measured to be around 180cd. Is that “too bright”? Should I calibrate it the way it is or is that going to “hurt” the calibration process? Is it better to calibrate at 120cd and then change the brightness after… I assume not, correct? Is it possible for my monitor to be more accurate being less bright (i.e., with the brightness level at around 120 or 100?)?
5. Should I use Gamma 2.2, 2.4 or sRGB for the Tone Curve? The displayCAL ReadMe talks about maybe using gamma 2.4 along with using ambient light measurements, that kinda confused me. Are they different (2.2, 2.4 and sRGB)?
6. Another ReadMe doubt: “A white point temperature different to that native to the display may limit the maximum brightness possible.” so…? Should I just use “As Measured”?
7. How big should the calibration screen be? I've seen some people use it at the default size/position (small square in the center) and others who set it to be a maximized window. Does it matter?
8. What are all the files that DisplayCAL generates? ti1, ti3, bpcti3, b2a1 post clut and postclutsmooth (PNGs?), wrz. What are those for?

[Stahlmann]

1 hour ago, Lacanian Wizard said:

1. This monitor has 4 gamma settings. I’ve watched the Hardware Unboxed tutorial video on calibrating displays and it’s said there to basically reset the monitor settings and disable any kind of enhancement, which I’ve done: in my case, black level stabilizer is set to 0, no energy saving mode is enabled, no motion blur reduction/backlight strobing, ‘tho Freesync is enabled (also have an AMD GPU). The gamma settings on the OSD are just labeled “Mode (1, 2, 3 or 4)”.  Modes 1, 2 and 3 are progressively darker (1 being the brightest and 2 being the default which is selected after reseting the settings). However, in the display's manual, it’s said that “If it’s not necessary to adjust the gamma settings, choose Mode 4”. So, which should I choose? The “no adjusment” (mode 4) or the one it defaults to after reseting the settings (mode 2)?

I'd guess they're 2.0 (1), 2.2 (2), 2.4 (3) and 2.6 (4) gamma. But it could be anything in between.

 

You can easily test it using DisplayCal and your colorimeter:

Select gamma 1 and then in DisplayCal click on "uncalibrated monitor report". After it's done measuring you get a small report giving you some basic information about the measured contrast, whitepoint, gamma, etc. Do this for all 4 modes and pick the one that ends up closest to 2.2 going forward.

 

1 hour ago, Lacanian Wizard said:

1. Should I measure the gamma using the "uncalibrated monitor report" tool in DisplayCAL and then try to reach the value there? Say, if it says 2.43, I should input 2.4? Yes, no?

So you already found it  You don't need to input your gamma anywhere. After trying all modes out select the one closest to 2.2 and pick sRGB as your target profile in DisplayCal. This will calibrate your monitor to follow the standard 2.2 sRGB gamma curve.

 

1 hour ago, Lacanian Wizard said:

1. Which whitepoint should I choose? Is D65 the same as 6500k? And is that the same as the sRGB whitepoint? Or should I use As measured? Or am I writing nonsense here? Is this just preference, in my case? Is this just a matter of choice in my case?

D65 and 6500K is the same. Don't use "as measured", because it will use your wrong out-of-the-box whitepoint as a baseline for it's calibration.

 

1 hour ago, Lacanian Wizard said:

1. What about white level, should I stick to the 80-120cd I see recommended everywhere? Before buying this, I always used my monitor at 50 brightness (also the default after settings “factory reset”) and that is measured to be around 180cd. Is that “too bright”? Should I calibrate it the way it is or is that going to “hurt” the calibration process? Is it better to calibrate at 120cd and then change the brightness after… I assume not, correct? Is it possible for my monitor to be more accurate being less bright (i.e., with the brightness level at around 120 or 100?)?

Brightness is entirely dependant on your viewing environment and personal preference. For sRGB content i like to calibrate to 100 nits. On some monitors that means 80% brightness, on some 10. The value in the OSD is unimportant. It is best to calibrate to the brightness you will actually use, but changing it afterwards will not introduce huge inaccuracities.

 

I found that 100 nits is the sweet spot. Not too bright at night time, not too dim at daytime. Even if it feels too dim at first, give your eyes 10 minutes to adjust and it will look just fine. Conveniently it's also the mastering brightness for sRGB content, so you're as close to "the creators intend" as you'll get.

 

1 hour ago, Lacanian Wizard said:

1. Should I use Gamma 2.2, 2.4 or sRGB for the Tone Curve? The displayCAL ReadMe talks about maybe using gamma 2.4 along with using ambient light measurements, that kinda confused me. Are they different (2.2, 2.4 and sRGB)?

2.2 is the target for sRGB calibration.

 

1 hour ago, Lacanian Wizard said:

1. Another ReadMe doubt: “A white point temperature different to that native to the display may limit the maximum brightness possible.” so…? Should I just use “As Measured”?

That is correct. If you set your monitor colors to 100% on red, green blue and the backlight to 100% this will be the brightest your monitor can get. As soon as you take out specific colors to match the 6500K whitepoint you take out some maximum brightness. Again - this only matters when you use your monitor at 100% brightness, and you already said you don't.

 

1 hour ago, Lacanian Wizard said:

1. How big should the calibration screen be? I've seen some people use it at the default size/position (small square in the center) and others who set it to be a maximized window. Does it matter?

Just leave it at default. If your monitor has any type of local dimming, check the "black background" box so bright objects behind the measuring window don't impact the dimming zones during the calibration.

 

1 hour ago, Lacanian Wizard said:

1. What are all the files that DisplayCAL generates? ti1, ti3, bpcti3, b2a1 post clut and postclutsmooth (PNGs?), wrz. What are those for?

Never looked into that tbh. I can't say.

[Stahlmann]

Here is a short walkthrough of my typical sRGB calibration process:

1. Find the best gamma mode with the "uncalibrated monitor report" feature i mentioned above.

2. On the home screen of DisplayCal select "sRGB" in the settings bar,

3. Select the correction (depending on your exact monitor technology used) down below.

4. Go to the calibration tab and tick the "interactive display adjustment" box

5. 6500K white should be automatically selected because you chose "sRGB" in the preset before

6. Set white level to whatever your target brightness is

7. Ignore the rest of the settings and click on the "calibrate & profile" button

 

During the interactive display adjustment:

1. On the first tab adjust your RGB values in the OSD to get as close as possible to the 6500K whitepoint

(Tip: If your default RGB values are 50 for example - not 100 - only adjust the ones you need to adjust to lower values, don't go over the default with any color. If all are at 50 and the weakest color is green, then adjust red and blue down, not green up.)

2. On the next tab adjust your brightness using the monitor OSD until you're as close to your target as possible

3. Ignore the other 2 tabs and click continue

 

After the profiling is done, install the resulting ICC profile as your standard and make sure you have the DisplayCal Profile Loader running at all times in the background. Then run a color checker from the "verification" to verify your results and check if the calibration did it's job.

 

With the steps above you can calibrate pretty much any monitor to perfectly reproduce sRGB colors (as long as it covers the color space to a good extend). If that is your target there is no need to dig much deeper. 

[Logue]

Thanks for all the answers so fast!

I'm gonna do these things later, I'm now calibrating the other computer which was in far worse shape (pretty old display, a LG W2353V). But thanks a lot, pretty much all I wanted to know. And I kinda did exactly what you described in your step-by-step little guide there, except for the sRGB portion. I didn't use the sRGB because the first calibration I did came out like crap because whitepoint was set to "as measured" (even tho' I had selected the sRGB preset... weird, not sure what happened there). But now I know what went wrong (my second calibration worked great bc I specified 6500k as the whitepoint, not "as measured". But then I used the tone response of Gamma 2.2, not sRGB.

 

Also, about the gamma modes in the display, I don't think Mode 4 is gamma 2.6, because it looks a lot like the Mode 2 (which supposedly = gamma 2.2). Like I said in the first post, in the manual Mode 4 is described as a mode to be used "when you don't need to adjust the gamma". IDK what that means exactly... Do they just use gamma 2.2? Then why have 4 modes and not just 3, with the Mode 2 being gamma=2.2? And although it is similar to Mode 2, it's not quite the same. I'll check both with the report tool later and I'll come back here with some logs/reports, maybe there's a difference.

[Stahlmann]

7 minutes ago, Lacanian Wizard said:

Also, about the gamma modes in the display, I don't think Mode 4 is gamma 2.6, because it looks a lot like the Mode 2 (which supposedly = gamma 2.2). Like I said in the first post, in the manual Mode 4 is described as a mode to be used "when you don't need to adjust the gamma". IDK what that means exactly... Do they just use gamma 2.2? Then why have 4 modes and not just 3, with the Mode 2 being gamma=2.2? And although it is similar to Mode 2, it's not quite the same. I'll check both with the report tool later and I'll come back here with some logs/reports, maybe there's a difference.

These gamma settings are different from brand to brand and even then sometimes different across other models of the same brand. Like i said, the only thing you can do is to test what mode does what. Some monitors i tested even had gamma presets that were straight up called "2.2" or "2.4", but when measured they weren't even close to that.

[Logue]

So, final question: is there any reason not to choose the Gamma 2.2 in the Tone Response Curve setting (in calibration window) and go for the sRGB instead? Like, why not the Gamma 2.2? Cuz I've seen people say sRGB is bad if you watch video (??) cuz dark portions get significantly brighter (allegedly, I haven't tested it yet - intend to create 2 profiles with everything the same apart from that setting).

[Stahlmann]

5 hours ago, Lacanian Wizard said:

So, final question: is there any reason not to choose the Gamma 2.2 in the Tone Response Curve setting (in calibration window) and go for the sRGB instead? Like, why not the Gamma 2.2? Cuz I've seen people say sRGB is bad if you watch video (??) cuz dark portions get significantly brighter (allegedly, I haven't tested it yet - intend to create 2 profiles with everything the same apart from that setting).

The sRGB gamma curve is not flat. It's mostly 2.2 but falls off on the low end. Selecting 2.2 will make your tone response curve flat 2.2, which is not how sRGB content is supposed to look. If some people think it looks bad that's just their opinion. But it's not accurate.

[Logue]

6 hours ago, Stahlmann said:

The sRGB gamma curve is not flat. It's mostly 2.2 but falls off on the low end. Selecting 2.2 will make your tone response curve flat 2.2, which is not how sRGB content is supposed to look. If some people think it looks bad that's just their opinion. But it's not accurate.

 

Right, I see. In my case, that may not be much of an issue (again, I'm not doing professional work with colors or photos etc.). I'll just make two profiles with everything else being the same and compare the two for my use case and see what looks better. And if I ever need an sRGB profile, I have it too.

 

One thing that's happening however is that it seems the profiles are removed after I close the main DisplayCAL program. The DisplayCAL Profile Loader is active in the system tray and it's colored (not grayed out as it gets, for example, when you open DisplayCAL), but it doesn't seem to be doing it's job. You know the "Download" button (looks like one) that you can click in DisplayCAL to install the profile you created after the calibration process is over? Yeah, so, that... When I click that (even if I'd already done it before), the colors change to the "correct" ones. Great, now they are correct. However, as soon as I close DisplayCAL (with the profile loader still active in the system tray), the colors revert back to what they were before calibration. Is that correct? I had that happen with both PCs I've calibrated - both use AMD GPUs, don't know if that may be a problem with software/driver interaction and Windows 10...

 

During installation, I DID select "Let DisplayCAL handle calibration" on both computers. If I uninstall DisplayCAL and just use Windows calibration (then selecting the "Use Windows calibrattion" tickbox in the Color Management settings in Control Panel), it seems to work just fine. Is DisplayCAL doing anything ELSE besides applying the color profile (that'd be the .ICM file, correct?)? There's that "1D LUT" it generates, that's applied directly to the GPU or something like that...? Does that remain even if I uninstall DisplayCAL? Is that applied together with the profile when you select it in Windows settings? Is it another file? Or is it a setting just accessible/enabled if DisplayCAL is installed?

[Logue]

On 4/7/2022 at 11:14 AM, Stahlmann said:

Here is a short walkthrough of my typical sRGB calibration process:

1. Find the best gamma mode with the "uncalibrated monitor report" feature i mentioned above.

 

6. Set white level to whatever your target brightness is

About 1) - the "best" gamma mode would be the one, between the four available, that is closer to Gamma 2.2 on the Uncalibrated report, correct? Strangely enough, that's mode 1. Mode 2 = gamma 2.4 and mode 3 = gamma 2.6 (pretty dark). Mode 4 is gamma 2.29~2.3, kinda weird.

However, the "default" (the one selected after you reset the settings to factory defaults) is mode 2 (gamma 2.4). Should I choose mode 1 then? I did that and calibrated with it selected - I can't get used to how "bright"/washed out things are. Probably my eyes are used to more contrast/saturation.

 

About 6) - should I set it to what my current monitor brightness is in cd/m² (by measuring it)? Or should I set 120 even if my display's brightness is measured to be 160cd/m²? Maybe that's why it's "washed out"/bright? With this latest calibration I did I left it on "As measured":

 

 

 

[Stahlmann]

9 hours ago, Lacanian Wizard said:

About 1) - the "best" gamma mode would be the one, between the four available, that is closer to Gamma 2.2 on the Uncalibrated report, correct? Strangely enough, that's mode 1. Mode 2 = gamma 2.4 and mode 3 = gamma 2.6 (pretty dark). Mode 4 is gamma 2.29~2.3, kinda weird.

However, the "default" (the one selected after you reset the settings to factory defaults) is mode 2 (gamma 2.4). Should I choose mode 1 then? I did that and calibrated with it selected - I can't get used to how "bright"/washed out things are. Probably my eyes are used to more contrast/saturation.

Gamma doesn't affect saturation. Just the brightness levels. If you prefer 2.4 then go for 2.4. Since you're no professional you're free to choose whatever you prefer. You just have to accept that you will watch most content darker than what it's supposed to be.

 

9 hours ago, Lacanian Wizard said:

About 6) - should I set it to what my current monitor brightness is in cd/m² (by measuring it)? Or should I set 120 even if my display's brightness is measured to be 160cd/m²? Maybe that's why it's "washed out"/bright? With this latest calibration I did I left it on "As measured":

 

 

Set your white level to whatever TARGET brightness you want to achieve. It doesn't matter what your display is currently set to. Once you enter the interactive display adjustment before the real calibration starts you can then change your brightness to match the value you want.

[Logue]

1 hour ago, Stahlmann said:

Gamma doesn't affect saturation. Just the brightness levels. If you prefer 2.4 then go for 2.4. Since you're no professional you're free to choose whatever you prefer. You just have to accept that you will watch most content darker than what it's supposed to be.

Got it, thanks. I'll try gamma 2.2 for a few days to try and adapt to it.

 

1 hour ago, Stahlmann said:

Set your white level to whatever TARGET brightness you want to achieve. It doesn't matter what your display is currently set to. Once you enter the interactive display adjustment before the real calibration starts you can then change your brightness to match the value you want.

Right, it's what I thought then. But is there a reference? Like, what value should I aim for in the white level? 120cd/m²? IF I go for higher than that, what does it mean in terms of accuracy? Let's say sRGB aims for 120cd/m². If I'd like to use my display at 160cd/m² (which is what I set it to), am I giving something up? Like, my dark content will be brighter, is that it? Not that it matters that much for my use case, I'm just curious at this point.

 

Yeah, gamma 2.4 seems better to me because I've gotten used to it, but I'll try to adapt to sRGB and gamma 2.2 over the next few days. Basically, I feel like gamma 2.2 some dark parts are brighter than they used to be - which I think is what's actually happened, right? Going for gamma 2.2 after having used 2.4, things WILL look brighter (cuz they ARE). Especially the dark portions. But blacks (#000000) are still just as black. Just the darker greys that look brighter now, but I think that's okay. For example, the command prompt and the Windows File Explorer (using dark mode in Windows 10) are much brighter now than they used to be, even tho' they are almost black. It's like there's more gradation now between pure black and 70% black (or sth like that).

Also, if I want to check if the calibration works, I should just use the verification tab, right? 

[Stahlmann]

22 minutes ago, Lacanian Wizard said:

Right, it's what I thought then. But is there a reference? Like, what value should I aim for in the white level? 120cd/m²? IF I go for higher than that, what does it mean in terms of accuracy? Let's say sRGB aims for 120cd/m². If I'd like to use my display at 160cd/m² (which is what I set it to), am I giving something up? Like, my dark content will be brighter, is that it? Not that it matters that much for my use case, I'm just curious at this point.

sRGB's reference brightness is 100 nits. But the beauty of SDR is that brightness is relative. It means it will be as "accurate" on a 500 nit display as it is on a 100 nit display. You can set it to whatever is comfortable for you. I personally found out that the 100 nit reference value is comfortable for me, so i'm sticking to that. But i also know it's too dim for many people.

 

Only for HDR the brightness is absolute and should not be changed by the end user for an accurate image.

 

26 minutes ago, Lacanian Wizard said:

Yeah, gamma 2.4 seems better to me because I've gotten used to it, but I'll try to adapt to sRGB and gamma 2.2 over the next few days. Basically, I feel like gamma 2.2 some dark parts are brighter than they used to be - which I think is what's actually happened, right? Going for gamma 2.2 after having used 2.4, things WILL look brighter (cuz they ARE). Especially the dark portions. But blacks (#000000) are still just as black. Just the darker greys that look brighter now, but I think that's okay. For example, the command prompt and the Windows File Explorer (using dark mode in Windows 10) are much brighter now than they used to be, even tho' they are almost black. It's like there's more gradation now between pure black and 70% black (or sth like that).

That's pretty much how it works. 2.2 will look brighter than 2.4. Still, like you said: Black is black.

 

26 minutes ago, Lacanian Wizard said:

Also, if I want to check if the calibration works, I should just use the verification tab, right? 

Yes. Just go to the verification tab and press the button at the bootom to start. After it's finished you can let it save an HTML. Open this file and a browser windows with all of your results will open.

[Logue]

Ok, thanks! Another curiousity moment:
Is there a difference between calibrating with different gamma modes? Like, if I choose Mode 1 for gamma (2.2) and calibrate it to sRGB. Is that gonna look exactly the same as if I'd selected Mode 2 for gamma (2.4) and then also calibrated for sRGB? Like, the same settings both times on DisplayCAL, the only change being the gamma mode in my display. There is, right? If I understand it all correctly. Color will still be accurate, but everything will be "darker" in Mode 2 (2.4)? That's the only change?

[Stahlmann]

There is one more thing you should know about display calibration: The more changes that have to be made with software (ICC profiles), the worse. That is because not every app is compatible with ICC profiles. So every app that doesn't support them is solely dependant on the changes you've made in the OSD. If your monitor is set to gamma 2 (2.4) and the ICC profile is made to match sRGB (2.2), then all ICC-supporting apps will be running at 2.2 while everything else runs at 2.4.

 

7 hours ago, Logue said:

Ok, thanks! Another curiousity moment:
Is there a difference between calibrating with different gamma modes? Like, if I choose Mode 1 for gamma (2.2) and calibrate it to sRGB. Is that gonna look exactly the same as if I'd selected Mode 2 for gamma (2.4) and then also calibrated for sRGB? Like, the same settings both times on DisplayCAL, the only change being the gamma mode in my display. There is, right? If I understand it all correctly. Color will still be accurate, but everything will be "darker" in Mode 2 (2.4)? That's the only change?

You will have to manually select gamma 2.4 in DisplayCal. If you yeave it at sRGB and your monitor to 2.4, then the ICC profile will transform your gamma to 2.2. And like i said above: The more the ICC profile has to do, the worse the end result. But if you select the sRGB settings preset and then just change your target gamma to 2.4 then it will be accurate.

 

Yes, the only difference between gamma 2.4 and 2.2 is the overall scene brightness. You can achieve the same levels of accuracity with either of them.

[Logue]

Thanks a lot for your answers and patience! If anyone is curious and has the same monitor model (LG 34GL750) as I do (I know there are variations from unit to unit, but hey... it may help someone), I've settled on these settings:

 

Display settings:

Contrast 70 (around 1100:1 measured, it varies)

Brightness 24 (it translates to 120cd/m² measured via DisplayCAL) - I found that after calibration, everything gets brighter anyway, so I might as well lower the brightness and the "washed out" effect I was noticing before is now gone and things are starting to """feel""" right (I got used to it, basically, lol)

Gamma Mode 1 (which is the mode that is measured to be 2.2 gamma via the uncalibrated report tool in DisplayCAL)

Color temperature - User mode -> RGB Values: 40 Red, 50 Green, 45 Blue

 

DisplayCAL settings:

Preset sRGB

Interactive display adjustment checkbox enabled but with low calibration speed and lots of patches (around 5000) for higher accuracy since this display isn't the most accurate (even after calibration, it says it only covers about 92% of the sRGB gamut with 104% volume)

White level: as measured (like I said above, around 120cd/m²)

Correction: LCD White LED family (AC, LG, Samsung)

[Stahlmann]

25 minutes ago, Logue said:

Thanks a lot for your answers and patience! If anyone is curious and has the same monitor model (LG 34GL750) as I do (I know there are variations from unit to unit, but hey... it may help someone), I've settled on these settings:

 

Display settings:

Contrast 70 (around 1100:1 measured, it varies)

Brightness 24 (it translates to 120cd/m² measured via DisplayCAL) - I found that after calibration, everything gets brighter anyway, so I might as well lower the brightness and the "washed out" effect I was noticing before is now gone and things are starting to """feel""" right (I got used to it, basically, lol)

Gamma Mode 1 (which is the mode that is measured to be 2.2 gamma via the uncalibrated report tool in DisplayCAL)

Color temperature - User mode -> RGB Values: 40 Red, 50 Green, 45 Blue

 

DisplayCAL settings:

Preset sRGB

Interactive display adjustment checkbox enabled but with low calibration speed and lots of patches (around 5000) for higher accuracy since this display isn't the most accurate (even after calibration, it says it only covers about 92% of the sRGB gamut with 104% volume)

White level: as measured (like I said above, around 120cd/m²)

Correction: LCD White LED family (AC, LG, Samsung)

A final tip: Leave the measuring speed and number at the default. You'll get pinpoint accurate either way. The extra measurements really don't improve your accuracity to any noticeable degree. You're basically just wasting hours of your time.

 

Happy to help a fellow enthusiast!

[Logue]

15 minutes ago, Stahlmann said:

A final tip: Leave the measuring speed and number at the default. You'll get pinpoint accurate either way. The extra measurements really don't improve your accuracity to any noticeable degree. You're basically just wasting hours of your time.

 

Happy to help a fellow enthusiast!

WHAT?!  How can that be?! Like, why?! Shouldn't 5000 patches render better accuracy than 175 (default)? Same for calibration, I noticed that with slow speed calibration it takes many more measurements before proceeding to profiling. I mean, if I can ACTUALLY get the same accuracy, then great, but, no offense intended here, how can I trust that answer? Is there evidence to back it up? Like, how do you know that? Cuz I've wasted some 4 hours in some measurements believing it'd be more accurate. LOL.  If I can get same accuracy with less time measuring, great, I'm all for it.

 

Maybe there are diminishing returns? Like, sure, I believe that 5000 patches might be too much (with more than that, DisplayCAL just stopped working - process froze - for some reason, both times I tried with that many patches). I just set to that cuz I left calibration running overnight, so it didn't really matter how long it'd take as long as it was done in the morning which it was (with 5000). Sure, ok, maybe 5000 is too much. But maybe 175 is too litle? Like, around 1000 seems to be a decent compromise, but maybe even that is worthless? Anywhere I could read up on that, like, maybe an article comparing the accuracy of X, X*2 and X*4 patches?

[Stahlmann]

4 hours ago, Logue said:

WHAT?!  How can that be?! Like, why?! Shouldn't 5000 patches render better accuracy than 175 (default)? Same for calibration, I noticed that with slow speed calibration it takes many more measurements before proceeding to profiling. I mean, if I can ACTUALLY get the same accuracy, then great, but, no offense intended here, how can I trust that answer? Is there evidence to back it up? Like, how do you know that? Cuz I've wasted some 4 hours in some measurements believing it'd be more accurate. LOL.  If I can get same accuracy with less time measuring, great, I'm all for it.

Evidence? Sure. Every reviewer i've seen using DisplayCal does it that way and the result is already basically perfect accuracity.

 

4 hours ago, Logue said:

Maybe there are diminishing returns? Like, sure, I believe that 5000 patches might be too much (with more than that, DisplayCAL just stopped working - process froze - for some reason, both times I tried with that many patches). I just set to that cuz I left calibration running overnight, so it didn't really matter how long it'd take as long as it was done in the morning which it was (with 5000). Sure, ok, maybe 5000 is too much. But maybe 175 is too litle? Like, around 1000 seems to be a decent compromise, but maybe even that is worthless? Anywhere I could read up on that, like, maybe an article comparing the accuracy of X, X*2 and X*4 patches?

Yeah, maybe i worded that wrong. But the diminishing returns are so small that it simply isn't worth the added time. You get 99% of the same accuracity out of the standard 175 measurements. Even with 175 patches you easily get maximum Delta-E's sub 2.0, which is the perceptable threshhold. Even if you get your maximum Delta-E from 1.9 to 1.0 with the added patches, you won't be able to perceive the color difference.

 

If you need more evidence, run another calibration with the standard 175 patches and then use the verification tab to check the accuracity of both modes you calibrated. That's the beauty of owning calibration hardware. You're not fully dependant on 3rd party reviews. Just try it out yourself.

[Logue]

Yeah, I'll try that! Thanks a lot! And, indeed, in the famous Hardware Unboxed video they made about color calibrating your display, the guidance is to just select sRGB and leave everything else at default (which includes the 175 patch default setting). That's great to know, I'll just compare my current measurement (I think I used 3400 patches, around there) with a new one using defaults and maybe something in the middle (i.e., 1500 patches). Compare the 3, see if there's any difference in deltaE between them and, if there is, how much so. But, yeah, like you said, probably not perceivable by humans.

[Logue]

8 hours ago, Stahlmann said:

Evidence? Sure. Every reviewer i've seen using DisplayCal does it that way and the result is already basically perfect accuracity.

 

Yeah, maybe i worded that wrong. But the diminishing returns are so small that it simply isn't worth the added time. You get 99% of the same accuracity out of the standard 175 measurements. Even with 175 patches you easily get maximum Delta-E's sub 2.0, which is the perceptable threshhold. Even if you get your maximum Delta-E from 1.9 to 1.0 with the added patches, you won't be able to perceive the color difference.

 

If you need more evidence, run another calibration with the standard 175 patches and then use the verification tab to check the accuracity of both modes you calibrated. That's the beauty of owning calibration hardware. You're not fully dependant on 3rd party reviews. Just try it out yourself.

Maybe while I have your attention... 

 

I'm now looking into creating 3D LUTs to use along MPC-BE and madVR, to watch videos (movies, TV shows, etc.). I've already found plenty of links and tutorials for that. One question I still have tho': do I need to add MPC-BE and madVR's executables to the Exceptions list using the DisplayCAL Profile Loader (right click in the system tray)? Should I do that? And, if so, should I disable both the profile loader AND reset video card gamma table? Because the 3DLUT generated will already contain all the information thatt madVR needs to render the video with the correct colors, right?

 

Also, I'm using the 3DLUT Maker program, NOT the main DisplayCAL program to create the 3DLUTs because I wanted to use the measurements/profile I already have. I'd just like to make sure that I'm supposed to use the default settings for that... Here's what it looks like right now:

 

 

Are those correct? After creating the 3D LUT there I can use it in madVR just fine? Or do I HAVE to measure it all over again and do it via the main DisplayCAL program? I see that gamma is set to 2.4, but that is default for tthe Rec. 1886 tone curve, right? I don't need to change anything, even tho' my display's gamma is now 2.2?

 

EDIT:

After trying it out and creating the profile with the settings above, things look pretty good, better than before the 3D LUT was applied. So I'll assume I did everything right until your next answer. What I found weird was that this process produced four files, not just one. One is a .3dlut which was what madVR was asking for. However, there's also a PNG and 2 ICM profiles that get generated. I'm assuming I just ignore the other 3 files, what matters to me is the .3dlut file, correct? Here's a screenshot:

 

Also, in the madVR calibration window, there are fields to insert different 3D LUTs depending on the color space. I just created one for the BT.709 because it's what I mostly watch/use.

 

However, could I possibly create 3D LUTs for BT.2020 and DCI-P3 too, even tho' those are wide gamut and my display is not (it's SDR)? If so, which 3D LUTs should I choose for each? As you can see below, there are multiple DCI-P3 and BT.2020 options...

 

 

Which DCI-P3 and which BT.2020 do I choose?

[Stahlmann]

3 hours ago, Logue said:

Maybe while I have your attention... 

 

I'm now looking into creating 3D LUTs to use along MPC-BE and madVR, to watch videos (movies, TV shows, etc.). I've already found plenty of links and tutorials for that. One question I still have tho': do I need to add MPC-BE and madVR's executables to the Exceptions list using the DisplayCAL Profile Loader (right click in the system tray)? Should I do that? And, if so, should I disable both the profile loader AND reset video card gamma table? Because the 3DLUT generated will already contain all the information thatt madVR needs to render the video with the correct colors, right?

I'd suggest to make an exception. Otherwise you'll have 2 software calibrations working against each other. But then remember to have the ICC profile already disabled when creating the 3D LUT.

 

3 hours ago, Logue said:

 

Also, I'm using the 3DLUT Maker program, NOT the main DisplayCAL program to create the 3DLUTs because I wanted to use the measurements/profile I already have. I'd just like to make sure that I'm supposed to use the default settings for that... Here's what it looks like right now:

 

 

Are those correct? After creating the 3D LUT there I can use it in madVR just fine? Or do I HAVE to measure it all over again and do it via the main DisplayCAL program? I see that gamma is set to 2.4, but that is default for tthe Rec. 1886 tone curve, right? I don't need to change anything, even tho' my display's gamma is now 2.2?

You should select 2.2 Gamma here aswell. And change the input encoding to full range 0-255.

 

3 hours ago, Logue said:

EDIT:

After trying it out and creating the profile with the settings above, things look pretty good, better than before the 3D LUT was applied. So I'll assume I did everything right until your next answer. What I found weird was that this process produced four files, not just one. One is a .3dlut which was what madVR was asking for. However, there's also a PNG and 2 ICM profiles that get generated. I'm assuming I just ignore the other 3 files, what matters to me is the .3dlut file, correct? Here's a screenshot:

Sorry idk about that. All i can do is make an educated guess. It looks like the 3D LUT is also dependant on an ICC profile that is created at the same time, thus probably incompatible with your other calibration.

 

3 hours ago, Logue said:

Also, in the madVR calibration window, there are fields to insert different 3D LUTs depending on the color space. I just created one for the BT.709 because it's what I mostly watch/use.

 

However, could I possibly create 3D LUTs for BT.2020 and DCI-P3 too, even tho' those are wide gamut and my display is not (it's SDR)? If so, which 3D LUTs should I choose for each? As you can see below, there are multiple DCI-P3 and BT.2020 options...

No you don't need to do DCI-P3 or BT 2020 if you have an sRGB only monitor.

 

3 hours ago, Logue said:

 

Which DCI-P3 and which BT.2020 do I choose?

The "ST 2084 10000" and "HLG 1000" profiles are for HDR content, so for SDR pick the ones that just say "color profile". Again, you don't need them since your monitor can't do them either way.

[Logue]

Hey!

So, should I select gamma 2.2 or use Rec. 1886 in the dropdown menu and then input Gamma 2.2?

 

 

Like, the tone curve should be Rec. 1886 or Gamma 2.2?

 

Also, I can't change the input encoding to Full, it's not an option:

And output encoding is greyed out (see 1st image)

 

Another question: See that the calibration I performed wasn't intended for video/madVR (it was for sRGB). So my tone curve (in the calibration/profiling process, before the 3D LUT generation) was selected to be sRGB. Is that a problem?

[Stahlmann]

1 hour ago, Logue said:

Hey!

So, should I select gamma 2.2 or use Rec. 1886 in the dropdown menu and then input Gamma 2.2?

 

 

Like, the tone curve should be Rec. 1886 or Gamma 2.2?

I'd put it at Gamma 2.2 in the first box. But tbh there isn't a right answer for every situation. There is no way for us to know if a movie is mastered for 2.2, 2.4 or Rec. 1886. When calibrating my TV i calibrated ALL SDR content to sRGB Gamma 2.2. Wheter that is right or wrong is not really important to me, as all the content i watched over the last 2 years looks good to me.

 

1 hour ago, Logue said:

Also, I can't change the input encoding to Full, it's not an option:

And output encoding is greyed out (see 1st image)

Probably because this is meant for TV use, not really monitor use. If you cannot select something else then limited range is all you get. On a monitor this could (but doesn't have to) lead to raised blacks.

 

1 hour ago, Logue said:

Another question: See that the calibration I performed wasn't intended for video/madVR (it was for sRGB). So my tone curve (in the calibration/profiling process, before the 3D LUT generation) was selected to be sRGB. Is that a problem?

Tbh i think this whole 3D LUT thing is completely unneccesary. As your monitor is sRGB only and without HDR support just use your standard sRGB 2.2 calibration for all content.

 

I'm not saying this because i don't want to help you anymore, but i think you're already way past the point of diminishing return when it comes to calibration. Especially since you're an enthusiast, not someone sitting at New Line Cinema mastering the next Lord of the Rings movie.

 

Just look at your verification results: If they're good, then don't worry about it. Then re-check in a couple of months to see if you need to recalibrate.

[Logue]

Ok, you're probably right. I've done what you said and things look pretty good here. Thanks a lot for all your help, man! You were a godsend!

If anyone else is interested, I did find this forum post from way back in 2015, which has also helped a great deal understanding all this gamma talk and whatnot.

 

https://www.avsforum.com/threads/madvr-argyllcms.1471169/page-208#post-40108362

 

I ended up settling on these settings:

 

 

Gamma 2.2, Relative with 100% Black output offset - following the advice from the forum post I linked above. I also tried the other settings suggested in that post and I didn't find it to be much different, so I ended up with the "default" 2.2 gamma curve.

 

[Stahlmann]

1 hour ago, Logue said:

Ok, you're probably right. I've done what you said and things look pretty good here. Thanks a lot for all your help, man! You were a godsend!

If anyone else is interested, I did find this forum post from way back in 2015, which has also helped a great deal understanding all this gamma talk and whatnot.

 

https://www.avsforum.com/threads/madvr-argyllcms.1471169/page-208#post-40108362

 

I ended up settling on these settings:

 

 

Gamma 2.2, Relative with 100% Black output offset - following the advice from the forum post I linked above. I also tried the other settings suggested in that post and I didn't find it to be much different, so I ended up with the "default" 2.2 gamma curve.

 

If you're really so picky when it comes to color then i suggest looking out for a monitor / TV that can do hardware-level calibration when it comes to future monitors. (Basically creating a 3D LUT that is stored inside the display). This way you're indepentent of ICC profiles. Most high-end TV's support this and some high-end monitors do aswell. Mostly professional ones, but also a few gaming monitors like the LG 27GP950 / 27GN950.