Samsung D7000 & D8000 Settings/Calibration Thread - Page 82

Hey, that's good service. I would expect that for a faulty set, but I didn't know they would swap them just because you didn't like the picture. I still think you're fine with the D7000 once you get the picture properly adjusted. The preset modes also vary from set to set (most are still way off from a calibrated picture and too dark or bright for my taste), so it's always better to calibrate yourself whenever possible.

See my post #2416 on the previous page. You can download a free calibration disc and instructions right here on AVS. The basic calibration takes a couple minutes once you get the hang of it.

Thanks ill do that.

I tried the 10 point white balance but it didn't really get me anywhere. My gamma for 80 was around 2.0 and 90 was around 2.5. Turning the contrast down dropped the gamma for 90 to around the same level as 80 for some odd reason. When using the 10 point adjustment I just didn't have enough wiggle room to get it dialed in. Any tips? Should I try and do all the 10 point adjustments before 80 and 90 IRE, and will that even things out? I've read that you can get the gamma on these sets ruler flat.

What are your contrast, gamma and cell settings and what patterns are you using?

My contrast setting was around 85, and read 33 ftl. Raising it did no good for the gamma but lowering it a couple of notches caused 90 IRE to drop below 2.2, but further dropped 80 IRE. Gamma was set to -1, and the cell setting was at 16. Raising the cell to 20 didn't do much of anything except add contrast to the picture. Oh and the patterns i'm using are the avs709 hd calman windows

set your contrast at 95 to align the 10pt controls better (9 should be closer to 90% etc.) and then turn down cell to 12-13 to get your 33 ftL. the controls should now give you the adjustment space you need.

I basically reset everything and will be trying again later tonight. I don't know if its the case when I calibrated the grey scale on the set, but when I turn the contrast up to 95 all bars in the contrast setup flash pink. Is this something I can calibrate out of the set?

So set contrast just below where you see clipping (pink). If that is too low you can fix it by doing an ADC calibration.

So is the clipping causing the gamma problem?

With my D7000, any Contrast setting at or above 85 allows the 10 point intervals to align perfectly with the luminance percent.

In Cal-day mode, Contrast at 85, Cell at 14, and Gamma at +1 yields 31 ftL with a near ruler flat gamma of 2.22 and deltaE less than 1.5. Color can be adjusted to be near perfect -- except for green and cyan which is common on the D7000 model.





In Movie mode, Contrast at 85, Cell at 10, and Gamma at 0 yields 26 ftL with nearly the the same results as above.

For either of the above settings, I can increase the Cell value to produce a peak output about 37 ftL with hardly any deterioration in the results

Larry

what would be the best settings to watch sports? March Madness!!

Joe, you may not get much feedback on that one. There are no aditional, "special" settings for sports. Just to be clear, in case this hasn't sunk in, this thread focuses on calibrating to a specific standard which, all else being equal, is independent of source. Personally, I think live sports feeds look absolutely perfect with a proper calibration (they look exactly how the broadcaster wants them to look). But I've never liked the moronically high color temperatures of most TV presets, unlike much of the general public it seems.

Like I said, any deviation from movie mode (warm 2) is moving away from an accurate picture. If that's your preference, then just adjust by eye however you like, it really doesn't matter at that point. Just be aware, you are a "victim" of how the manufacturers market their TVs - the one with the most obnoxiously bright picture typically sells more. Why should you use calibrated settings? This is a pretty good explanation with examples:

http://www.flatpanelshd.com/focus.ph...&id=1328263571

Anyway, if you want to simulate the artifically amped up settings many people seem to prefer, I suggest you calibrate movie mode like I suggested, then just change the color tone (e.g., try warm 1) to suit your preference.

Says who?
Last time I heard it was gamma 2.4 for the assumed dark room conditions (or 2.35 from EBU's assumption about a legacy CRT).
Others may think it should be a re-scaled inverse-encode curve (impossible to achieve with 10p controls, only external LUT processing plays here...).


By the way, the current calibration of my current TV is a mess if you check the gray-scale gamma only (Y row half-way between the colored bars, only filled for grays-scale) but it yields low overall dE errors on mixed colors: Pana G30 cal and this overall error is acceptable.

It's interesting how different results I got when I calibrated this TV using this kind of evaluation method. Even though I used the same patterns size (CalMan window size on the AVSHD disc) I ended up with very different CMS settings than I got with CalMan (mostly positive color luminance settings instead of negative values). The colors of random movie scenes look more like they look on my reference monitor (a CG275W calibrated with it's built-in color sensor), so I decided to go with this setup.

Here, every single color patch is compared to the color calculated from the Rec709 color space definition (+ I choose a target gamma of 2.4), not some kind of "temporal gamut" built over the actually measured gray-scale like CalMan does by measuring the actual 75% gray to judge the color luminance errors. (This calculation is relative to the 100% white point -note the zero error on 100% white- but I set that one to D65 as precisely as possible.)


I didn't calibrate a Samsung PDP with this method yet but I plan to do so. (I should patch this software to do absolute colorimetry on the gray-scale instead of 100% WP relative. It would be "perfect" then. But it's usable right now and it's a pain for me to compile softwares.)

Interesting comments, Janos. I noticed that too and wondered where they got it from. Anyway, I figured this was still a good reference for the layman. People need to see what they are missing and hopefully they might begin to appreciate it.

@janos: nice work. I have some interesting measurements of real-time gamma vs. APL on my samsung that is probably applicable to the pannys and also may be related to how one calibrates for luminance relative to white.

One question, why do you calibrate color luminance using 75% luminance and not something lower? The histograms I looked at for typical movie scenes is heavily weighted to much lower luminance, closer to 30%.

Here is one example of the Gamma vs. APL measurements I'm currently working on. Note that this set was calibrated for gamma using AVHSHD windowed patterns. The results for fixed APL patterns look very similar on average but there are distinct differences in the details of how each level input level behaves.


Gamma calibration with target = 2.3, fairly flat using window patterns:



I then measured the gamma response for each of 7 input levels relative to peak white as a function of APL. I did this by using small (1% size) windows against a variable gray 100% size background. In a perfect world, each level would be a flat line at gamma = 2.3 as we vary the picture content level.



The average of all levels is not too far off from the desired behavior and shows the typical non-defeatable dynamic contrast effect at low APL (gamma is too low) and ABL effects at high APL (again gamma is too low). But the interesting behavior is the spread in the response of the various levels. This is not noise in the measurement, all of the structure seen here is repeatable to the 0.02 gamma level.

So to take an extreme example, a 5% input level has fairly flat response but it's far too bright relative to white (gamma=2.15) while the 15% level is generally too dark relative to white (gamma=2.4). The 10% level is the level which most closely approximates our target value of 2.3

someone just give me numbers to plug in... I used zoyds numbers from the first part of this thread and the monitor looks good, look very good... but i want more.

there seems to be a slight posterization/metalic to the look of the image.

Question, when you go to calibrate your grayscale you should start with the two point and adjust using the white balance gains and offsets. Now I know some people use 80% for the gains and others use 100%. Now if I use 100% and let's say my red is at 120% my blue is at 110% and my green is down at 85%, how should I go about adjusting this? Should I leave green alone and bring the red and blue down to balance all three or do I bring green up some and then bring red and blue down to balance them. I guess my real question is when do you touch green? I know that I can bring red and blue down and get them all to balance and this will give me a certain Y value. Or I can bring green up and then balance all three and this will give me a different Y value. Now please correct me if I am wrong here but isn't the Y value at 100% the base for all the rest, like at 50 or 60%? I do know that when you make changes to 100 and the Y value changes it changes all the other points. I guess this goes back to if 100 is the base for the rest and I can obtain different Y values depending on how I adjust red, green and blue what is the standard or the proper way to adjust these. I hope I am making sense here.
Thanks

In the above example I didn't calibrate with any fixed levels. That's the point.

I ran the full set of measurements (all the 58 patches you see on the linked image), modified some settings in the OSD and ran the whole set of measurements again.
I iterated the settings until the average error looked good enough.

This may sounds like a hell of a job but no, it was actually faster than using CalMan (media players can be so slow, and then CalMan itself is a pain to use...).
The software (dispread) generated the patterns automatically and the i1d3 is a quick, so it was about 1 minutes / set (no additional time to generate the report, only a second until I saw it).
And I needed about 10 sets to complete the calibration of a freshly reseted prof mode, so it took about 20 minutes. (I guess a Samsung >D550 or a Pana VT30 would require ~20 iteration steps to tweak everything. Still less than 60 minutes and I spent hours with a VT30 once...)

The software generates full range RGB values but the VGA card can convert this to either limited RGB or standard YCC before it outputs the colors on the HDMI port. You just need a laptop with a discrete AMD GPU (>HD4xxx).


So I plan to use this workflow instead of the shiny HDTV specific softwares. (I just need to modify the software to use absolute colorimetry but the source code is available for free...)


Ah, well. Of course, the size of the patches is still a question... and it's not possible to use fixed APL during this automated measurement (not technically impossible but I can't write the program code to generate APL patterns on the fly...).

You should mention you are looking for D8000 settings (if my detective work is correct)...


Gotta search the thread I'm afraid.

So I messed around with greyscale and gamma last night and still couldn't get gamma to flatten out. Everything looks good except at 70 and 90 the gamma raises to around 2.4. My settings are cell 15, contrast 80, brightness 55, and gamma negative 1. It looks like little mountain range at the high end. help

ok, now I get it. Yes, this is a good generalized approach that should capture and minimize quirks that standard pattern sequence techniques miss. This could be built into HCFR as well since it already has a pattern generator and the new version uses the argyll drivers. What is your figure of merit, a straight average of the 58 dEs? It still might be better to weight that to an assessment of an "average" movie scene.


It turns out that for gamma response you can't do anything about the structure inherent in the APL induced effects. One would need a 10pt correction for each APL level to do that. The constant APL patterns tighten up the variation a bit but tend to leave the low levels with too low gamma. I think that's why people tend to prefer windowed patterns for gamma calibration since it ends up with a more contrasty result.

An average movie scene...
Well, let's average the color values of every pixels of all the movies ever digitalized or digitally recorded. What do you get?
83,89,56? Is this vector the answer for THE BIG QUESTION of life itself? LoL. I don't think there is such thing that "average movie scene".


Of course, patches can be added or removed from this list according to the calibrators taste about the actual device.
If I will use this method for a Samsung PDP with 10p gray and 6p CMS then I will increase the number of the gray patches from 8 to 16 and also the C,M,Y patches from 1 to 8 (and make it ~100 at total with the automatically generated patches).

The "random" patches are actually not that random. Those were generated by ArgyllCMS/targen. You can control how it iterates the points which are probably the best to check.



Yes. I talked about a 4D lookup based (external) CMS system which tries to handle this problem but my conclusion was that this would require more work than the realizable benefit may justifies.

exactamundo... 84/8000 numbers.

I didn't mean it that way, although we already know the answer to THE BIG QUESTION (42). But there are typical distributions for daylight outdoor scenes for example. I don't think it makes sense to base the patterns themselves on these distribution but if one does an APL based background then it might make sense to base the background on something like that. Most of the time we are viewing scenes below an APL of 40%.

Regarding the multidimensional LUT for calibration, I agree that for white balance and color the benefits do not seem worth the effort but it might be for low APL gamma, it's a mess on these displays.

Using your expanded palette approach how large are the improvements in the dEs for the non-standard colors? If you add a column to your spreadsheet with dE from a normal calibration what would it look like?

Hey SeeJeffRun, I want to thank you for posting your pn51d7000 picture settings, Ali_Red is right! They are the best settings I´ve tried for my PN51D7000 (Build December 2011). By any chance did you complete your calibration for the Color Space values? Thanks Again!

hey guys,for whoever has been paying attention to my situation,i went ahead and made my own changes in the sm. it looks MUCH better and very close to how it looked out of the box. i know you guys advised against it but i just couldnt live with this thing having to be in "cool" mode anymore,also that green haze just killed me.

would this set still be difficult for an experienced isf or thx guy to "fix"? i know this tv can get the most accurate colors compared to pretty much any tv and i still have hope my set can be saved.

keep in mind,after my sm adjustments it looks pretty darn good and i'll never have anyone that isnt at least respected on these forums touch this tv. but i know it can get much better by an experienced calibrator.

that being said,would you guys if you were in my situation,keep pursuing an exchange from best buy or just hire a calibrator from these forums to touch up and clean this set as it is?

Here is the second draft of a methodology for using HCFR and the i1 LT (D2) or i1 Pro meters to calibrate the D7000 and D8000 models. It is still rough so please let me know about and additions, clarifications, or corrections. Send them by personal mail to keep down the traffic in this thread.

Draft #2a released on 03/15/2012. Added more detail to grayscale and gamma calibration. Added new section on calibrating the color space.

 

The latest version can be found here: http://www.avsforum.com/t/1333561/samsung-d7000-d8000-settings-calibration-thread/2880#post_22299923

Larry

Edited by LarryInRI - 8/11/12 at 8:35pm

Larry,I was interested in picking up one of those colorimeters you listed. Of those listed which would you recommend? I'm highly interested in learning how to calibrate and I was leaning towards the eye one d2.

man im exhausted,i have to be to work in 3 hours! lol.

anyway,ive come to an even better result doing this: in the service menu i entered the white balance settings menu. i set everything,low rgb's high rgb's sub brightness and sub contrast all to 128.

i left everything in the user menu alone except lowering sharpness to 0 and brightness to 50(and disabling all the useless junk like dynamic contrast black tone etc). im very pleased with this result and feel i have successfully restored my tv (or even improved from) to factory settings. everything looks great,primaries,secondaries and greyscales all seem accurate as far as my eyes are concerned.

looking forward to getting a real calibrator out here and improving it even more. not to mention i really want to see a pre cal comparison,lol.

anyway,im tired gonna go to bed for an hour and a half thank you everyone who offered advice it was much appreciated! good night guys!

So for each color patch you adjust CMS manually to hit a numerical target (no real-time target display) and it takes 20 iterations to arrive at a global minimum? You must be fast!