Samsung D7000 & D8000 Settings/Calibration Thread - Page 79

Im not sure what you mean by strange, 38-40 fL was my target point for the client's viewing conditions. This set looked frkn awesome once tweeked out.

Zoyd, you are not considering the client's viewing conditions or their field of view.

Hi Gregg,

Just responding to his description of a "dark" environment and night time settings. My experience with this display has peak whites >40 ftL with those settings so I was curious. I agree, if the viewing FOV is narrower than what I'm used to it would warrant higher values.

hey Zoyd.

He is moving back to DR in a couple of weeks. He will be sitting 12 feet from the display and will be using 2 sets of panel lights as back lighting.

Why are you only getting 28fL? Why are you using a gamma of -3? These settings seem kind of strange / dim to me.

BTW, this panel was not capable of hitting more than 38-40 fL ( I dont remember what my final reading was at).

Night time settings (no bias lighting, 8 ft from 51" panel), gamma 2.4, peak white 28 ftL

Day time settings (low ambient window lighting, north side of house, small windows) gamma 2.3, peak white 35 ftL

I'll also use the day time settings with a small lamp on at night while watching sports/news stuff.

Maximum white using a 1% size pattern was 55 ftL.

1% size ? why ?

Are you getting 28 / 35 with the 1% size also? If so your display is much too dim for any kind of reference viewing.

A standardized methodology is required when performing video calibration (which means setting the display to established standards).

I'm was just telling you what the peak white capability is relevant to low APL. I Use standard size patterns for the 28/35 and for those the maximum I've measured is 44 ftL.

For the DIY'ers that used the AVSHD disk to calibrate your D series plasma, I was wondering what the consensus is on which patterns you used to give the best results. I know they say that 10-12% sized patterns are what you want for plasmas, but the large APL's are too small and the window patterns are too big. So I was curious which ones you used.
Thanks

Use the patterns on DVE - BD Basics. The problem with this disc is that there are no RGBCMY patterns for setting the color gamut, but it works well for setting gamma, gray scale, and peak light output.

Another good disc is the THX Calibration BD. This is somewhat hard to come by and only officially given out if you attend a THX training (or sometimes a THX special event).

Gregg,
Doesn't the DVE also have the problem with pattern size. I also have this disk and thought that I had measured the pattern sizes and they were not within the 10-12% range.

correct, they are not 10-12%.

They are correct :-)

Hey Gregg
Thanks for the calibration now I see the full potential of the TV it looks incredible. The movers will come to pick the TV on Friday and it will arrive in DR in a month, I don't know if I can't wait that long without watching the TV you have spoiled me LOL. As I told Gregg when my wife saw the Tv after calibration she was looking at the minimum detail to see if she found anything wrong with the picture but she stood in total silence that to me is PRICELESS LOL. The day before the calibration she was complaining why was I paying someone to come and calibrate the TV, but I guess if it was money for a Shoe it would be ok LOL. Guys if you can pay for someone to calibrate your TV do so it does wonders to the picture quality. Highly recommend Gregg Loewen.

Hello Gregg, which pattern would I use to calibrate with if I only had the AVS HD 709 disk? I have a Samsung PND516500 plasma. I had DVE BD Basics but the disk would have play back issues with the blu ray spitting out the disk and not reading it, or if it did read it then it would just freeze or hick up. I returned the disk regardless. Probably just a compatibility issue with my player. I had no patience for it any more, my blu ray player is an older samsung bdp 2550 latest firmware which is also old now. Time to upgrade soon...

AVSHD 709 may not be optimal, but has all the patterns you need, including 75% RGBCMY windows. For grayscale, I think most people recommend the normal (not APL) windows.

Gregg, how often do you make it down to the DC area?

Gregg,

You misunderstood my comment about "strange." I did not mean it in any derogatory manner.

Chad B obtained a peak white of 54.6 ftL with a gamma of 2.2 as he reported in March 2011. He did not post his settings but I assume that cell was set at the maximum of 20 and contrast near 95. Even though he did not explicitly state that he used the movie mode for his calibration, he implied such. He did not state what size window was used for his measurements.

With the 17-18% area windows available to most of us, we have been able to achieve much higher levels than the approximately 38 ftL with the same settings that you used. That and the owner's stated viewing conditions of a dark room with the peak level near 40 ftL is the context for my use of the word "strange." There are some of us that cannot take that kind of visual abuse.

Larry

Funny you should ask that as I've been working on comparing how various patterns measure for gray scale, color, and gamma. This isn't a recommendation but some data to help compare the various options.

Here are my first results:



This is a table of comparisons using measured deltaE between a baseline calibration and various other test pattern sources. The baseline calibration was done using patterns embedded in a high resolution photograph with average RGB level of ~37%.



CIELUV dE>3 is highlighted yellow and >6 is highlighted red. Unusable readings are grayed out (i1 pro color for 17% stimulus and 10% gray level)

Things to note:

-Gray scale calibration is not as sensitive to pattern configuration as color calibration.

-AVSHD APL small patterns are an outlier and yield the largest errors relative to the other patterns.

-Baseline was very repeatable other than a 3 dE drift in red.

-Gamma was stable for all patterns except my high (55% avg. video level) pattern set.

-Comparison between my window patterns and AVSHD window patterns is in excellent agreement.

-Green had the largest variability and if you look at the details of the measurement the xy of green is what changes not luma error. So the green primary moves with APL.


I'm still digesting this data so please chime in with comments and take a look at the test patterns ( BD iso). The disks are rough as far as navigation goes.

zoyd,

A few questions.

Why are you using CIELUV rather than CIE94 or CIEDE2000?

Can you explain the relationship between Pattern Level and Peak White? IOW, how are the peaks all about the same with the different pattern level? It's probably semantics but it is confusing -- at least to me.



I'm downloading the disk right now.

Larry

Sounds like that guy tried to calibrate your TV somehow similar to how I do these Samsungs but he badly lost the track somewhere.
(I also alter the SM settings and I use the Cool color tone but the final characteristics are fine, not like yours...)
I wonder how he dared to ask money after he showed you those cal reports. Didn't he sweat a bit? LOL.

I'm using CIELUV just for convenience since that's what HCFR calculates and it's a sensitive algorithm similar to Lab.

Peak white is just for the 11 point gray scale ramp while pattern level is for the embedded color pattern.

My findings are that chromaticity of every colors can move between both APLs and IREs.

Colors on Samsungs have reasonably stable relative luminance levels but notably fluctuating chromaticity while on Panasonics, colors have stable chromaticity but incorrect relative luminance levels.
At the end, they both have a "red push" on skin tones (caused by either over-saturated or too bright reds).

Since the chromaticity diagram is a non-linear but monotonic curve which can cross the constant target at one level only and because I think over-saturation is still better than under-saturation, I set up the colors at 100%.
This way the picture has some red push (over-saturated red on skin tones) but otherwise bright colors are a little washed out while darker tones are still over-saturated which is even worse (I think).

All the color patterns are 100% saturated, the variable is luminance. Like you I see in general that higher luminance patterns drive the calibrated luminance values up so that colors are too bright (not saturated) if I use 75% or 100% luminance patterns. That's why I made the lower value patterns to do these tests with. I set the luminance of the pattern equal to the average luminance of the scene it's embedded in to simulate a real scene. I was also interested in the relationship between pattern luminance and average picture level because of the dynamic contrast in the Samsungs. Solid colors on black backgrounds are not a good representation of what the TV normally displays.

I just thought I'd post my recent experience as a relative newcomer to TV calibration with my PN64D8000.

I'd been contemplating purchasing something like CalMan + Colorimeter to do the job until I remembered that I had a "Lacie Blue Eye" that I'd bought for PC use but the software is totally unsuitable for TV calibration since it's very basic and is only useful for creating PC color profiles. After learning that these devices are just rebadged EyeOne Display2s, I started looking for supported software.

I then stumbled across the free HCFR program which everyone seemed to think was too complicated for novice use. After a day or two though of reading tutorials and messing around I've started to really get my head around how it works and have been able to produce stunning results that made me very happy that I'd made the effort.

If people want, I may make a YouTube video that shows the steps that I took and maybe a few pros can critique my work.

Here are some things I've learned about the Samsung plasma duing calibration.

- I generally prefer Contrast at ~70 in a light controlled room. 90+ might be OK for day use but it's just too retina-assaulting for a dark room.

- The 10p white balance adjustment is spread across the contrast range so it matches 10 IRE, 20 IRE... etc when using contrast near 100% but with lower contrast e.g. 70 only the first 7 will have an effect and they'll be spread across 10 IRE to 100 IRE making it a bit trickier to adjust.

- My particular display (1000 hours) requires a major cut in red gain when using Warm 2. If I use Warm 1 it requires a major cut to blue and red gain.

- This is the first display I've owned with colorspace settings and the accuracy of skin tones that can be obtained is simply AWESOME.

46"? So, it´s an lcd, not a plasma.

something that covers about 15-18% of the screen, it is okay to use a smaller screen window but not a larger one. Always use a window and never a field. If they only have a 12 % window then use that.

Hi, I regularly service New England and then the I95 corridor down to Virginia Beach. Every 3-6 weeks.

HI Larry, no problems.

Interesting data. Always remember that the an i1pro can be off by as much as 25% on the Y reading. It is more accurate at xy data.

Interesting, how do flesh tones look after you have done this ?

Remember not to spend too much time evaluating CIE graphs when the response of the display is non linear, it is a far better use of time and will also lead to a more accurate display if you actually evaluate reference material for flesh tones instead.

FWIW,

Gregg