Author Topic: Greyscale/Gamma Calibration Tip  (Read 5674 times)

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #30 on: October 19, 2015, 09:36:13 AM »
What I'm trying to achieve here and I am achieving is getting accurate readings for d65, overcoming the inaccuracies of the meter at d65. If you only have one warm setting try standard, if it's 7500k then it will work as well.

I don't feel this is a good way to go about it personally. If you think it gives you good results then that's great.

You cant get an accurate bt1886 with a 2point, even a 10 point white balance control, anything up from 21 point will get you more or less in line with it.

But with your ee color box you should be able to manage it. Just keep the patterns small enough not to trigger a black level jump at the lower end.

My VT has a 10 point luminance adjustment alongside the white point settings (along with a few gamma presets). I have my panel tracking 1886 up until ABL becomes an issue (which is where good judgement takes presidence over numbers).
« Last Edit: October 19, 2015, 09:43:59 AM by KenMasters »

Offline idm

  • Trade Count: (-1)
  • AVForums Veteran
  • ***
  • Posts: 470
Re: Greyscale/Gamma Calibration Tip
« Reply #31 on: October 19, 2015, 09:50:54 AM »
As mentioned, using Ted's lightspace cms bluray disk, patterns are 11.11% the test disk is hightly acclaimed. When calibrating with ee color, use calman pattern generator 11.11%, or madvr 11.11%. I am sure ABL is not being triggered as the results achieved are actually awesome. Have you compared your meter to perhaps a klein colorimeter, or even a jeti specbos, how do you know that your meter is reading accurately?

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #32 on: October 19, 2015, 10:40:55 AM »
As mentioned, using Ted's lightspace cms bluray disk, patterns are 11.11% the test disk is hightly acclaimed. When calibrating with ee color, use calman pattern generator 11.11%, or madvr 11.11%. I am sure ABL is not being triggered...

The issue with your display is not ABL, the black level drifts depending on the average brightness of the panel. You need to watch out for that as you go through the levels.

Have you compared your meter to perhaps a klein colorimeter, or even a jeti specbos, how do you know that your meter is reading accurately?

There is quite a bit of leeway, accurate does not mean it has to be bang on, just within threshold - our eyes are really not that sensitive to colour. More important is that things track linearly - it's the shades we're sensitive to. The lighting in my living room is precisely D65, it's a good critical viewing environment and as an art director I'm pretty good at spotting when things are off.

How is your viewing enviroment? Do you watch TV with the lights on - what's their colour temp? Are your walls neutral? If your environment is not neutral, what you see on your display even though it measures correctly will perceptually not be accurate due to the bias these elements introduce. Of course I don't expect people to compensate for these things, it's just an illustration of the relativity of things.

There is so much more to a good picture than a pretty graph, the numbers should just inform your choices, the fact that you don't seem aware of the black level drift makes me wonder what attention you are paying to the actual display. In fact I find it quite surprising that you would invest so much in your equipment but don't seem concerned about the ability of the display itself.

Offline idm

  • Trade Count: (-1)
  • AVForums Veteran
  • ***
  • Posts: 470
Re: Greyscale/Gamma Calibration Tip
« Reply #33 on: October 19, 2015, 01:27:23 PM »
My black levels do not drift. Perhaps what you are basing on is the American versions of my tv model, my black levels remain constant, no brightness pops noticeable, do you have any experience with the same model number as mine? My lighting is switched off when watching tv, walls are a light brown so more to a d3500, this would of course help the d65 to pop.

I agree, there is so much more to a better picture than pretty graphs, but accurate measurement is even more essential, and given that you have not employed even a spectrophotiometer or spectroradiometer to confirm your measurements and are ignorant to any good suggestions made to improve that accuracy, I find that very hypocritical coming from you. You did not even mention what pattern sources you use, I would imagine that you are still using the avshd disk that has rounding errors.

What stimulus do you use when setting your colour gamut, 100/100, 75/100, 100/75 or 75/75, do you even have those patterns? How accurate is your colour checker sg? You want to talk about linearity, let's see your 10 point saturation and 10 point luminance measurements. These are imporatnt measuremsts to have accurate, not just pretty graphs and still will not give a good enough representation on how linear your display is. To even start measuring the condition of the display's colour rendition is to have accurate measurements.

Here educate yourself a bit more:

http://www.displaycalibrations.com/disk_features.html

http://www.lightillusion.com/tech_info.html

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #34 on: October 19, 2015, 03:03:38 PM »
My black levels do not drift. Perhaps what you are basing on is the American versions of my tv model, my black levels remain constant, no brightness pops noticeable, do you have any experience with the same model number as mine? My lighting is switched off when watching tv, walls are a light brown so more to a d3500, this would of course help the d65 to pop.

Like the E8000, the biggest flaw was with ‘brightness pops’, where image luminance will raise and lower at fairly frequent points, which can be distracting. Floating blacks are again present but the steps Samsung have taken to control luminance stability, down low, have certainly moved things on from last year and only the most sensitive of people to this phenomenon are likely to notice.

https://www.avforums.com/review/samsung-e6500-ps60e6500-3d-plasma-tv-review.350

As for floating blacks, those are here too, but they’re well disguised to the extent that they shouldn’t prove distracting. During a letterboxed film, if we paid attention to the black bars, we could see that they would sometimes rise to a brighter level during overall brighter scenes. However, surprisingly for us, they also sometimes lowered, too: during some scenes in Se7en they could reach 0.015 cd/m2, before returning to the baseline level of 0.02 cd/m2.

http://www.hdtvtest.co.uk/news/samsung-ps60e6500-ps51e6500-201207262022.htm?page=Performance

Also make sure when adjusting gamma not to start with a 0% pattern as the panel switches off.

I agree, there is so much more to a better picture than pretty graphs, but accurate measurement is even more essential, and given that you have not employed even a spectrophotiometer or spectroradiometer to confirm your measurements and are ignorant to any good suggestions made to improve that accuracy, I find that very hypocritical coming from you. You did not even mention what pattern sources you use, I would imagine that you are still using the avshd disk that has rounding errors.

I have a number of references, my own patterns, CalMAN, AVS HD, Disney WOW and DVE HD Basics.

Like I said, I don't think a spectro is a particularly smart investment for a hobbyist calibrator. Even a stock Display Pro's errors are shown to be below xy0.006 for both color and white with plasmas. To me the money is better spent on the panel itself, marginal differences in accuracy are nothing compared to the difference that cash would make on the quality of panel you could buy.

What stimulus do you use when setting your colour gamut, 100/100, 75/100, 100/75 or 75/75, do you even have those patterns? How accurate is your colour checker sg? You want to talk about linearity, let's see your 10 point saturation and 10 point luminance measurements. These are imporatnt measuremsts to have accurate, not just pretty graphs and still will not give a good enough representation on how linear your display is. To even start measuring the condition of the display's colour rendition is to have accurate measurements.

I'm not interested in making this a competition but I'll take some measurements and post them if you'd like. You have some interesting notions, I'd be keen to know what your basic picture settings are.

Offline idm

  • Trade Count: (-1)
  • AVForums Veteran
  • ***
  • Posts: 470
Re: Greyscale/Gamma Calibration Tip
« Reply #35 on: October 19, 2015, 03:08:59 PM »
I'm not interested in making this a competition but I'll take some measurements and post them if you'd like. You have some interesting notions, I'd be keen to know what your basic picture settings are.

Basic picture seetings? Why?

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #36 on: October 19, 2015, 05:23:15 PM »
Basic picture seetings? Why?

I'm just curious to see what choices you make for your base settings. I took some readings now and compiled them into a single image.

« Last Edit: October 19, 2015, 05:25:24 PM by KenMasters »

Offline idm

  • Trade Count: (-1)
  • AVForums Veteran
  • ***
  • Posts: 470
Re: Greyscale/Gamma Calibration Tip
« Reply #37 on: October 19, 2015, 05:39:03 PM »
Nice black level, what stimulus ar you using for the gamut? I will measure later on this evening, after everyone has gone to sleep and post, at the same time share my base settings.

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #38 on: October 20, 2015, 09:24:42 AM »
Nice black level, what stimulus ar you using for the gamut?

Yea, the black level is quite something on the VT and ZT, also exceptionally clean in regards dithering noise. I intially set the colour at 100W/75% (the VT didn't really need touching out the box in Professional, I think it was only Magenta that crossed 3 dE), then run through 0, 25%, 50%, 75% and 100% to see if there are any issues (saturations I ran through for you yesterday, rest of the graphs come from an older calibration).
« Last Edit: October 20, 2015, 09:43:10 AM by KenMasters »

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #39 on: October 25, 2015, 12:05:14 PM »
Nice black level, what stimulus ar you using for the gamut? I will measure later on this evening, after everyone has gone to sleep and post, at the same time share my base settings.

Still planning to make a post?

By the way, any thought to using a backlight, makes for far more comfortable viewing.

Oh, and for interest sake, talking about how accurate some TVs are out the box, check out the new Panasonic:


Offline idm

  • Trade Count: (-1)
  • AVForums Veteran
  • ***
  • Posts: 470
Re: Greyscale/Gamma Calibration Tip
« Reply #40 on: October 25, 2015, 12:43:36 PM »
I will be, just swamped with some other stuff at the moment, what Panaonic is that? Out the box?

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #41 on: October 25, 2015, 03:27:51 PM »
I will be, just swamped with some other stuff at the moment, what Panaonic is that? Out the box?

It's their new OLED panel, but it's not unusual for Panasonic - these are the out the box measurements of their budget "4K" LCD:

« Last Edit: October 25, 2015, 03:31:14 PM by KenMasters »

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #42 on: May 18, 2016, 07:26:04 AM »
I've got a 55" Panasonic VT60 and use the plasma metering mode. Helpfully I also have a D65 Master TL-D 90 Graphica backlight and wall paint.

Random correction, just updated Calman for the first time in years yesterday and went though my settings to make sure there were no changes and it turns out I use the CRT metering mode for the Panasonic. The Plasma setting is skewed slightly towards blue, I would guess the 3000 focused field drive makes the panel behave more like a CRT than the standard "600Hz" subfield drive.

Offline idm

  • Trade Count: (-1)
  • AVForums Veteran
  • ***
  • Posts: 470
Re: Greyscale/Gamma Calibration Tip
« Reply #43 on: May 18, 2016, 07:41:29 AM »
Strange, I find that the plasma metering mode for the i1d3 is spot on compared to the i1pro2, measure's closest to the pro2 with plasma meter mode. But then again, if memeory serves me right you also have a C6?

Offline KenMasters

  • Trade Count: (0)
  • AVForums Veteran
  • *****
  • Posts: 4,409
Re: Greyscale/Gamma Calibration Tip
« Reply #44 on: May 18, 2016, 08:24:33 AM »
Strange, I find that the plasma metering mode for the i1d3 is spot on compared to the i1pro2, measure's closest to the pro2 with plasma meter mode. But then again, if memeory serves me right you also have a C6?

Well, most plasmas are like yours, driven by a "600Hz" sub field, the later Panasonics use what they call a focused field drive, so instead of the image being rendered across the duration of the frame in multiple passes (10x60Hz or 12x50Hz = 600Hz), each frame is rendered in a single 1/3000th of a second burst - probably closer to the single pass nature of a CRT. On a side note, I don't know if you're familiar with the Leo Bodnar lag tester, but this behaviour of plasmas is why it can't properly measure lag on a plasma - always look at the camera method measurement, not the lag tester.

I don't have the C6, just the standard Pro model, maybe you're thinking of the HDR version of the C6 I came across in the beginning of the thread? I have no true reference measurement device, I rely on my experience as an art director and the reference ambient conditions of my room to tell me when things are off. The THX preset on my TV is pretty accurate, but a little too much red on the high end (and a little green on the low), there's a slight blue/green cast across the range when using the Plasma mode (checked to see why I selected CRT mode yesterday), CRT gives me readings than concur with my observations (both the THX preset and the plasma metering mode) and results that look right to my eye.

Typing this the thought also occurs to me that the Panasonic used a more saturated red phosphor (in order to cover 98% of DCI [and appears more natural as previously red, even if measuring correctly, could appear undersaturated/luminescent]). That might be another reason for the inaccuracy as green and blue measure the same, just wild speculation.

EDIT: It might be interesting to compare the two modes across the THX preset and my own calibration - and to compare my THX results with those of reviewers who've measured the panel with their Klein 10As. Only if you're keen to see, wouldn't just do it for fun.
« Last Edit: May 18, 2016, 08:37:42 AM by KenMasters »