Display Calibration [Archive] - Page 8

Siso

24th February 2020, 16:20

Next select Madvr 709, you stay in the same monitor mode.

Under calibration, you now select white point 6504 as before, but under tone curve, select default gamma 2.2, then change that to 2.4, This will carve out a 2.4 curve, then use that curve to drive the madvr lut creation. This won't affect your monitor's profile, this part is only the creation process for 3dlut madvr.

Under 3D lut, leave everything default, EXCEPT, tone curve, select 2.2 relative, then change to 2.4 relative. leave other settings default.

When I calibrate for madvr 3dlut, I set these settings - http://imgbox.com/0kFJzViu What about black output offset?

tp4tissue

24th February 2020, 16:31

When I calibrate for madvr 3dlut, I set these settings - http://imgbox.com/0kFJzViu What about black output offset?

You can't use the setting above, unless the default behavior of the monitor is nearly perfect. Which it almost certainly is not.

BUT, only after you've scanned through the hardware settings on your monitor to find the best mode for calibration. Otherwise, you won't have the optimal setting. Contrast ratio optimization is very important for IPS, because it has so little of it to begin with.

Use these settings when doing madvr. (not before looking through the modes for the best behaved one)
https://imgur.com/a/LRpNxb1
https://imgur.com/rC2jntH

Siso

24th February 2020, 16:44

You can't use the setting above, unless the default behavior of the monitor is nearly perfect. Which it almost certainly is not.

BUT, only after you've scanned through the hardware settings on your monitor to find the best mode for calibration. Otherwise, you won't have the optimal setting. Contrast ratio optimization is very important for IPS, because it has so little of it to begin with.

Use these settings when doing madvr. (not before looking through the modes for the best behaved one)
https://imgur.com/a/LRpNxb1
https://imgur.com/rC2jntH

Everywhere people say that in Custom mode of the monitor you can reach the rgb channels to manually set them to match the white point. I always use Custom mode.

tp4tissue

24th February 2020, 16:55

Everywhere people say that in Custom mode of the monitor you can reach the rgb channels to manually set them to match the white point. I always use Custom mode.

The custom mode MAY be good, you don't know that for sure, unless you check the other modes as well.

Again, don't go by what people say online <except Tp4issue>, they either don't have a colorimeter and generally don't understand their device/measurements.

Load up default 2.2 setting from the top menu, go to verification. Then run the simulation profile testing as below

In Most cases, to maximize the contrast ratio, you'll prolly need to go into the dell hidden menu and tug on the rgb gain settings. UNLESS those settings work out in the custom mode, or possibly standard mode is already peak.

Test the modes first: vv

https://imgur.com/a/hj5jHvS

These things come out very different from the factory, one mode might do something for someone else, but NOT YOU, because the gains are set wrong at the factory.

Just looked at the manual for ur monitor, u can tell this is not targeted @ videophiles, it spends most of the pages telling you how to use the backlit leds hahahaha.

OK, test the following modes, under standard, test Warm, Cool, and Custom(set to 6504k via displaycal's interactive whitepoint). Test the custom mode a second time, with all the rgb values pushed to the highest. it'll look wrong, but you're just trying to figure out which gain is the weakest, and you'll try to target that channel's peak brightness with the other channels so that you get peak contrast.

Make sure the dynamic contrast is disabled, and all other settings that affect contrast. such as that black stabilizer.

Then go into the factory hidden menu, and check what goodies are available in there.

Siso

24th February 2020, 17:03

In the other modes I won't be able to set the rgb channels.

tp4tissue

24th February 2020, 17:17

In the other modes I won't be able to set the rgb channels.

Right , you can't set it in the main menu, but you want to test them anyway, in case you get higher contrast out of it, or a better tone response, or something else.

Don't worry about the RGB

With dell monitors, the factory gain is always available in the hidden menu. So if it turns out the other modes are better, you can adjust those gains in the hidden menu.

But you should test all of them to figure out the best behavior conditions.

Siso

24th February 2020, 17:22

Right , you can't set it in the main menu, but you want to test them anyway, in case you get higher contrast out of it, or a better tone response, or something else.

Don't worry about the RGB

With dell monitors, the factory gain is always available in the hidden menu. So if it turns out the other modes are better, you can adjust those gains in the hidden menu.

But you should test all of them to figure out the best behavior conditions.

Will test them, the next time I'll start the calibrations. BTW what to do if this comes up when calibrating for madvr - http://imgbox.com/00pf6MEE I wonder if this will pop up with your settings, with mine it always pop up to ask me.

tp4tissue

24th February 2020, 17:46

Will test them, the next time I'll start the calibrations. BTW what to do if this comes up when calibrating for madvr - http://imgbox.com/00pf6MEE I wonder if this will pop up with your settings, with mine it always pop up to ask me.

mine doesn't ask me that, I'm assuming it's because i'm setting the tone curve in the setting for the calibration.

don't start the calibration, before you run the test, you don't need a calibration to test the behaviors of the modes.

Siso

24th February 2020, 18:19

mine doesn't ask me that, I'm assuming it's because i'm setting the tone curve in the setting for the calibration.

don't start the calibration, before you run the test, you don't need a calibration to test the behaviors of the modes.

Understood. :thanks:

Siso

29th February 2020, 10:08

OK what about correction? Should I use "Spectral: LCD White LED family (AC,LG Samsung)"? Or choose one from the online database for example this one - i1 DisplayPro, ColorMunki Display & Dell, Dell AW3420DW by 4KM (i1 Pro).ccmx http://imgbox.com/E3r2SatS Here it says W-Led https://www.displaydb.com/compare/dell-alienware-aw3420dw/

huhn

29th February 2020, 18:01

if you find a correction for the same screen, similar display or with the same panel use that else the w rgb is most likely the closest.

Siso

29th February 2020, 18:04

if you find a correction for the same screen, similar display or with the same panel use that else the w rgb is most likely the closest.

You mean - i1 DisplayPro, ColorMunki Display & Dell, Dell AW3420DW by 4KM (i1 Pro).ccmx?

huhn

29th February 2020, 18:06

i guess so.

Siso

29th February 2020, 19:36

i guess so.

I usually use Spectral: LCD White LED family (AC,LG Samsung), but I will try the one from the database.

Thanks, huhn.

tp4tissue

29th February 2020, 22:03

It's not white led. well it is white led, but it's got nano-ips, which means quantum dot coating, you need to use one of the quantum dot corrections.

Siso

29th February 2020, 22:17

It's not white led. well it is white led, but it's got nano-ips, which means quantum dot coating, you need to use one of the quantum dot corrections.

There is only one - Spectral: LCD Quantum dot LED (Samsung QLED Q9).

tp4tissue

29th February 2020, 22:50

There is only one - Spectral: LCD Quantum dot LED (Samsung QLED Q9).

yea i'm pretty sure that one will work, cuz the q9 has ~98% p3 like the quote on yoru monitor.

they all buy quantum dot from the same place. produced by genetically engineered bacteria.

Siso

29th February 2020, 23:18

yea i'm pretty sure that one will work, cuz the q9 has ~98% p3 like the quote on yoru monitor.

they all buy quantum dot from the same place. produced by genetically engineered bacteria.

These are the available corrections from the online database http://imgbox.com/Ddloufch perhaps one of them will work? I imported them in displaycal, when I choose "auto" it selects one of the spectral corrections.

tp4tissue

1st March 2020, 01:41

These are the available corrections from the online database http://imgbox.com/Ddloufch perhaps one of them will work? I imported them in displaycal, when I choose "auto" it selects one of the spectral corrections.

I'm pretty sure the q9 one is lab grade, I mean, it won't make a huge difference visually.

I would use the q9, if you want to use some dude's spectro, it's probably fine.

Did some googling, turns out, yea, you should use those corrections instead of the q9, apparently nanocell is not quantum. it's something else.

Siso

1st March 2020, 15:34

I'm pretty sure the q9 one is lab grade, I mean, it won't make a huge difference visually.

I would use the q9, if you want to use some dude's spectro, it's probably fine.

Did some googling, turns out, yea, you should use those corrections instead of the q9, apparently nanocell is not quantum. it's something else.

I see, on Auto it selects this spectral correction - http://imgbox.com/wlFXMtdZ it is one of the four available for my monitor in the database (2 matrix and 2 spectral corrections).

tp4tissue

1st March 2020, 15:55

I see, on Auto it selects this spectral correction - http://imgbox.com/wlFXMtdZ it is one of the four available for my monitor in the database (2 matrix and 2 spectral corrections).

I'm not sure if the i1pro is better than the colormunki. double check.

Siso

1st March 2020, 16:04

I'm not sure if the i1pro is better than the colormunki. double check.

This is the other spectral correction - http://imgbox.com/EWS3Rtle

The other 2 are matrix corrections. http://imgbox.com/tm162g4c
http://imgbox.com/Xh4USxfy

tp4tissue

1st March 2020, 22:16

This is the other spectral correction - http://imgbox.com/EWS3Rtle

The other 2 are matrix corrections. http://imgbox.com/tm162g4c
http://imgbox.com/Xh4USxfy

run the simulation test on all of them if you want to, but I don't think there's gonna be a huge difference.

Just look out for if any of them have the primaries in a very different location, if they're all in similar agreement, you should be fine with any of them.

Siso

2nd March 2020, 12:34

run the simulation test on all of them if you want to, but I don't think there's gonna be a huge difference.

Just look out for if any of them have the primaries in a very different location, if they're all in similar agreement, you should be fine with any of them.

:thanks:

Siso

2nd March 2020, 21:19

These are the primaries of the two spectral corrections http://imgbox.com/t9BJcM5f
http://imgbox.com/RbsMzEW5

tp4tissue

3rd March 2020, 00:16

These are the primaries of the two spectral corrections http://imgbox.com/t9BJcM5f
http://imgbox.com/RbsMzEW5

They're reasonably close. the white point looks like it's moved a bit, but that shouldn't be a problem. After calibration, if the visual white looks really off, try the other one. I doubt this would happen though.

Siso

3rd March 2020, 10:21

They're reasonably close. the white point looks like it's moved a bit, but that shouldn't be a problem. After calibration, if the visual white looks really off, try the other one. I doubt this would happen though.

Will test both of them, but this one http://imgbox.com/t9BJcM5f seems better than the second, right?

tp4tissue

3rd March 2020, 23:22

Will test both of them, but this one http://imgbox.com/t9BJcM5f seems better than the second, right?

We can't know which one is better, unless you also buy a spectro. just because it's closer to some value right now, doesn't mean one is better, there's no point of reference.

Every measurement can also have some level of drift. If you checked it again today, it's not gonna be the same coordinate, because the temperature changed.

Prior to measurements, warm up the device running an rgb pixel fixer video for ~ 1.5 hours. The LED takes approximately 40 minutes to stabilize, some takes a little more than an hour. You want to exercise the LCD pixels as well as the LED backlight using the video, because you get more contrast out of it post exercise. If you leave a monitor in a box for a long time, the pixel gets lazy , and won't flip on all the way, exercise and warmth brings them back. Take care not to let the video get paused by youtube, or the computer/monitor going to sleep.

Leave the meter plugged in during this phase, but do not put it on the monitor.

The only thing you can go by is if there are visual anamalies. such as color casts. Then you know something may be up.

Damien147

17th May 2020, 15:44

huhn

17th May 2020, 18:20

just verify it every 3 month.

foxyshadis

19th May 2020, 09:14

For digital displays it is a good idea to re-calibrate your TV once a year. (https://support.datacolor.com/index.php?/Knowledgebase/Article/View/1340/84/how-often-should-i-calibrate-my-tv)
....but here I've read some people mentioning calibration every 300 hours:confused:
Is it truly beneficial to the eye or we are talking about numbers only?I mean for movies.
Once a year sounds good because I am lazy but am I missing something?:D

If you only ever calibrate once and never again, your TV will still be better-calibrated forever than 90% of everyone's. It's like when to wash your car; some people are obsessive about weekly washes, some people never bother. Aside from an initial burn-in period and the end-of-life rapid failure, the color won't change much if at all through the lifespan, but if it is new, it's worth calibrating every few weeks until you get roughly the same results twice in a row. If it's on the way out, of course, you can't calibrate your way out of that for long.

Damien147

19th May 2020, 15:25

Your first sentence sounds good to me.:)
I calibrated first time a year ago and I don't see anything wrong but you can't measure with eyes.The verification(never done it) that huhn mentioned before calibrating now can show how off are things and start from there I guess.
I have in mind to calibrate every October for example and I am not a perfectionist but I also don't want my calibration to become pointless now that I have a meter,something like that.
DeltaE >3 will be perceptible to the viewer I've read.I'll keep this as an indication I guess.
I have to move my TV first to calibrate every time and I don't see how others in same occasion have the appetite to do it every 300 hours:scared:
They have better TV I guess.I don't think it's worth it for the cheap 50 inch samsung I have.
Thanks guys.:)

huhn

19th May 2020, 17:16

it depends on the screen not every screen drifts a lot some drift like they have nothing better to do.

PJ are known to drift more in the first 100-1000 H there are always exception and so on.

Damien147

19th May 2020, 19:37

I meant regular movie watching on a TV.As you say results vary and it makes it hard to answer.
I'll try the less than DeltaE >3 thing when it's time to see how it goes.Thanks again.

RXP

3rd June 2020, 14:28

Hey guys, hoping someone can help here. I've searched everywhere for an answer but can't find my exact use case.

When you create a 3D LUT for MadVR using Display CAL you target a certain gamma. In MadVR's Gamma Processing tab could you then use that to change the Gamma of your LUT? e.g. say if I had a Gamma 2.4 LUT created, tracks perfectly 2.4 from verification. I now want a daytime viewing Gamma - could I just select 2.2 in Gamma processing? Will MadVR understand that the LUT is at Gamma 2.4? Or is it expecting a specific default Gamma for the LUT if you want to use Gamma Processing?

huhn

3rd June 2020, 21:42

madVR doesn't know the gamma of the 3D LUT. just create a new one takes like a minute.

if you create a 2.2 3D LUT it may work.

RXP

4th June 2020, 06:15

Hoping to get confirmation on that - because in that case the best thing to do with DisplayCAL is just to create a Gamma 2.2 LUT and let MadVR do any changes you need.

The reason I was asking was for 3D movies on my JVC x9000. Calibrating after the glasses is a bit of a faf compared to a regular flat panel. Some movies benefit from a lower Gamma due to a peak white of only 15 nits after the glasses. Doing it that way would save me time - but I guess I'll bust out the meter again or rely on JVC's own Gamma adjustments.

RXP

4th June 2020, 06:22

For digital displays it is a good idea to re-calibrate your TV once a year. (https://support.datacolor.com/index.php?/Knowledgebase/Article/View/1340/84/how-often-should-i-calibrate-my-tv)
....but here I've read some people mentioning calibration every 300 hours:confused:
Is it truly beneficial to the eye or we are talking about numbers only?I mean for movies.
Once a year sounds good because I am lazy but am I missing something?:D

If you're using a TV it's really not necessarily IMO.

Disclaimer: I use calibration to fix issues I can see. Otherwise it's a numbers obsession IMHO.

- My C6 had terrible low IRE greyscale that looked purple/blue and high gamma at low IRE which exacerbated compression noise. A MadVR 3D LUT took that out like magic. I'm not really interested in objective colour accuracy - it's a nice side benefit but if that drifts I'm fine with it.

- My JVC x9000 that I bought used had a calibration applied to it. I could visibly see the green looked luminous on Cover Art in Kodi/Plex. I used JVC'a Autocal and it corrected it nicely. With a PJ I do it after 800 hours or so. They dirft way more and historically have had Gamma droop issues.

- My C9 - ran a 3D LUT with MadVR and could see no visible difference between factory and the LUT after a minor whitepoint adjustment. In fact factory Gamma hid the compression noise way better, so I'm never calibrating that.

huhn

4th June 2020, 06:44

if you have proper measurements and you are using a 3d LUT already why would you need the meter to change the gamma?

RXP

5th June 2020, 07:35

I went ahead and created a 2.6 and 2.4 Gamma 3D LUT for my C6 OLED. Very happy with the results. I like to target higher gamma on the C6 to reduce compression noise, while having no shadow crush. After realising you don't need to measure Gamma on the calibration tab it's massively reduced the time it takes to run a 3D LUT so I don't mind doing it again.

huhn

5th June 2020, 18:24

you just need to create a new 3dLUT with a different target the source which is the time consuming part doesn't change.

Siso

21st July 2020, 20:55

How often is good to calibrate a monitor for madvr 3dlut? I'm making a 3dlut every 3-4 months, I'm asking cause it takes around an hour and a half for the 3dlut...1553 patches -_-

VBB

21st July 2020, 21:43

The usual recommendation is to re-calibrate about once a year, depending on your display type. Projectors tend to drift more than monitors/TVs. Every 3-4 months for a monitor is overkill, but then again why not, if you have your own equipment :)

Asmodian

21st July 2020, 21:48

I agree with once a year as good practice.

I might calibrate a new OLED or projector after the first 4-6 months of use and then once a year. At most every six months, every three months is way too often.

Siso

21st July 2020, 22:31

Thanks for the replies guys, I'm kinda fed up with calibrating every 4 months, I have I1 display pro, and even it is faster than the colormunki, it takes too much time for the 3dlut :(

VBB

22nd July 2020, 00:11

I mean, a 3D LUT is not all that involved compared to a manual calibration. It all comes down to OCD. When I first got my meter, I would bust it out almost every night, and this was before I ever messed with DisplayCAL, back in 2013. Then in 2015, when I got my fist 4K LED display, I started doing auto-cals using madTPG. Once I got the hang of it, I left everything alone until my next TV purchase. That was my current C7 OLED, and I don't recall exactly how many 3D LUTs I created, but I do know that the last one I made was in January of 2018. It still looks great, and I haven't noticed anything different. In fact, the TV has seemingly gotten better over the years, as can be the case with OLED. These days I'm simply too lazy to calibrate. My next TV will be ordered pre-calibrated from VE (by jrref, for those who know him). Calibration can be an awesome hobby, but the cost of keeping up with equipment and software can be very high, if you're intent on doing it right.