Samsung D7000 & D8000 Settings/Calibration Thread - Page 91

Don't kill me... I am simply looking for the post you are referring to for guidance...Not having much success

Here:

http://www.avsforum.com/avs-vb/showt...5#post21768285

I recommend the AVS HD 709 test disc because it is free and complete (but you will need a DVD burner and bluray player). If you want the brightest possible output in a roughly "calibrated" picture, select movie mode (and disable all processing), set your cell to 20 and then set contrast as high as possible with the white clipping pattern (if you start to see pink tones in the bars, then reduce contrast a click or two). Then use the APL clipping pattern to set brightness, followed by blue only mode for color and tint. You can set sharpness to your preference, but 20 or less is recommended. The PDF directions will tell you exactly how to use each pattern.

zoyd,

Okay, I have selected my films and others for evaluation. They are "Birdman of Alcatraz" and "To Kill a Mockingbird" for black and white sources. Birdman has many very dark scenes that are a killers for any TV that has crushed blacks while Mockingbird has many bright scenes in addition to some fairly dark ones.

"Fried Green Tomatoes" and "Out of Africa" were chosen for their color fidelity and straight forward, no-finagling with color and lighting method of filming. (Most of Sydney Pollock's movies are good reference sources. I miss him.)

Then just for variation, I have added the HBO BD version of "Rome" and the UK region-free version of "Downton Abbey" both of which should be in anyone's list of reference disks.


I have attached both your new gamma curve and the 'custom' gamma that I use. They are presented with the display gamma with black compensation format. Also. I will be comparing these to a flat 2.28 gamma curve.


Larry

Nice selection. Which patterns did you use to calibrate windows or APL? The above was done with windows and APL patterns give me a little flatter curve but the 10% control is the same for both. I also had to bump brightness down such that levels 17,18, and 19 are clipped otherwise it rises too fast.

I used plain windows. I had to drop the Brightness by four ticks to adjust the clipping point about the same -- 17 and 18 are clipped for the same reason that you describe.

Larry

zoyd,

It has turned out to take a bit more effort then I had anticipated to compare the different gamma curves. I had to set up the Movie, Cal-day, and Cal-night and adjust each with a different curve while keeping the peak white level the same for each. Now I simply pause the source on a particular frame and can switch between gammas while watching for changes in telling portions of the frame.

So far I've looked only at some of Birdman. Near the beginning of the movie, there is a significant difference in the detail of the very dark grays in the scene where Lancaster is in the unlighted solitary cell -- particularly when the cell door is opened to let him out. The right side of the screen is in total blackness and the left side of the screen shows the exterior wall of the cell lit with the equivalent of about 40% luminance. Only a little light illuminates the interior of the cell.

For all three gamma curves, the black on the right side and the illuminated wall on the left side appears the same. It is a different story for the image of Lancaster crouching on the cell floor. With the constant 2.28 gamma, there is a loss of detail in much of body. In particular, facial detail and the texture of his prison uniform are lost. With your camera gamma curve and my custom one, there is little if any difference -- the detail is restored and the picture is much more realistic.

More to come...

Larry

Thanks. Yes, I have read that instruction, thanks. You are one of the guys I was hoping would respond.

Before I describe what I'm seeing, let me say that the only difference I can see on my 51D8000 between 4:2:2 and 4:4:4 outputs from my Panasonic DMP-BDT310 Bluray player is with the Chroma Zone Plate. All those other patterns look the same, and the only difference between 422 and 444 is that with 422 the moire of the little circles (which are located between the big circles in the middle and the circles cut in half at the edge of the screen) are not symmetrical. In one the center is dark and in the other it is light. With 444 they are symmetrical.

What I'm seeing and posted about is with those tall skinny "diamonds" in the Chroma Alignment screen. In particular the ones in the lower left, which is the quarter of the screen with a gray background. The red diamonds are not symmetrical. I see two pixels of red "bleed" into the gray on the RIGHT side and one pixel bleed on the left side. So, I'm wondering if this is a setting in my TV, possibly someplace in the SM, if the disc has a flaw here, or if it's just my BR player. As I already said the output (444 vs 422) from the BR player shows this to be the same. Same for all those patterns that S&M describes as well as turning Sharpness to 0, making sure picture is Screen Fit, and all "enhancements" are OFF. If yours looks like it's supposed to, maybe it is my player, and I was specifically wondering if the Oppo players would get this right.

So, if you have a chance to look I'd be appreciative if you would tell me if your particular combination of TV and player show the same thing with the red diamonds in the lower left against the gray background. (What TV do you have?)

Also, I'd be interested to hear what you see with the Chroma Zone Plate with 444 vs 422.

Probably one of many, but a lot of heavy hitters weighing in ...

http://www.avsforum.com/avs-vb/showthread.php?t=1369670

Sure, I will have a look when I get a chance. Too bad Secrets hasn't yet expanded their benchmarking project to BD players, it certainly made shopping for a DVD player simple.

thanks, I had read that a couple of months ago. Ultimately no consensus but an interesting read.

@Larry - thanks for taking up "Studio grading" with the two gammas. They do produce very different results and my gut feeling is that it will help in some cases and hurt in others, so the question will be if it is net plus for certain types of material or a crapshoot.

I took a look at my S&M disk ps3->D8000 (no sampling options available) and the alignment patterns were all perfect for blue, there looked to be a very small misalignment with red on the right side, maybe 1/2 pixel.

chroma zone plate showed some moire and darkening in the corners but I don't know how much is typical.

I'd trust the PS3 as well, it's a top notch bluray player. What color space are you using, zoyd? I'll have a look at the Oppo on my 59D7000 tonight.

You mean output format? YCbCr and RGB give the same results.

zoyd,

Your gut feeling is absolutely correct.

Late last night, I popped in the 2010 BD version of "Out of Africa" and right away during the opening scene with the train winding through the beautiful terrain, it was obvious that the picture was not right. It was too bright in the mid-range luminances with your camera gamma. The flat 2.28 looked best and very similar to my custom gamma. "Fried Green Tomatoes" which is a 'bright film' looked washed out with the camera gamma but looked fine with the other two curves. I went back to Birdman and it still looked good with the camera gamma. Mockingbird looked good with all three gamma curves.

I viewed portions of episodes of Rome and Downton. They also looked wrong with the camera gamma but quite satisfactory with my custom gamma They looked even a bit better with the flat 2.28 gamma however with a loss of detail in some of the very dark areas of the picture.

This has convinced me that, for these Samsungs, a flat gamma from 90% down to 30% and a slight fall off below that is ideal. The lack of detail in the dark areas is a negative property of these TVs mentioned in some of the reviews that I have read. So after doing all this comparing various source material, I have modified my gamma curve slightly.


It's been fun. And thankfully I could sleep in today.

Larry

Thanks Larry, nice work. I'll reproduce your curve and give it a try.

zoyd,

Here is the chc file for my modified custom gamma.

Larry

 

gamma modified #2.zip 1.283203125k . file

I am using CNETS settings for my 59d7000. Have tried others frome here but really like the picture I am getting. I just want to increase brightness some. Is it best to use brightness, cell light, contrast or a combination?

Brightness is for setting the black level -- confusing to novices, I know. Increase Cell Light for a brighter picture. Don't set Contrast higher than about 93 since it will cause a pink overtone.

Don't use any D8000 settings for the D7000. They will cause a deterioration in the picture. For the D7000, it's really best to use the factory Movie mode if you don't have calibration equipment.


Larry

Good information. Thanks Larry

Yeah, Oppo and Spears & Munsil use the term "color space". So, I think you're saying 4:22 and 4:4:4 look the same via PS3, right?

On my system, zone plate darkens some in the corners (and L/R edges), moire looks more symmetric in 4:4:4 but the light to dark gradient at the corners/edges is a bit smoother in 4:2:2.

Chroma alignment diamonds seem to be nicely centered in 4:4:4 for both colors. However, in 4:2:2, blue remains centered while red is shifted 1/2 to 1 pixel to the right on both sides.

Think I'm feeling a bit more motivated to run through the checklist/worksheet now, but I still wonder if this amounts to anything visible in normal content. Thinking I may return to 4:4:4 regardless.

Edit: In checking multiburst for multiple conversions, 4:2:2 is much better (brighter). This may trump everything else since the purist in me won't let that one go.

So, your Oppo works best with Chroma Alignment using 4:4:4 setting, but in Chroma Multiburst you like 4:2:2 better. I'm going to have to look at Chroma Multiburst again, paying attention to the "brightness" comparison between 444 and 422. What did you mean by "multiple conversions"?

My Panny player shows no difference between 444 and 422 on Chroma Alignment.

What TV do you have?

I also have a PS3 but I dont find any setting to choose 4:4:4 vs 4:2:2

I ran the Spears & Munsils Chroma tests for my 59D7000 and Oppo 93 a while ago. I used the printable sheet where you can tick which option gives the best results.
With my setup it was actually the RGB Video Level that got most ticks in the boxes, or actually I think it was a draw for 4:2:2 and RGB Video, but RGB won because I gave more weight to some patterns than others.
Can't remember the exact results and I've lost the sheet. Not sure if I can tell the differences in real world material, but with test patterns they were very clear.

As S&M discuss in the "choosing a color space" tutorial, one of the goals is to minimize the number of color space conversions occurring in the display chain (the other is to figure out which device performs them more accurately). Assuming my Oppo 83 is about as good as it gets in this respect (the lazy approach, but I'm pretty confident in it), I want to identify the Oppo output format which results in the minimum amount of processing (conversions) going on in the TV.

If you read the S&M tutorial, they show you a specific region of the multiburst pattern (high freq, vert) which will be brightest when the number of conversions is minimized. 4:2:2 was the clear winner in this respect because there is no "muting" in any parts of the pattern. This seems to agree with Janos' findings, and I know he spent a lot of time analyzing his D550 (and other models which he has calibrated). Not sure if that would apply to the rest of the D series, but it seems to work for my D7000. The D8000 uses a different processor from the rest, so this may or may not hold true for your display. I didn't have time to check all the available Oppo output formats, but I'll get around to it eventually.

Seems it's usually a comprimise, so if it comes down to alignment vs. conversion accuracy, I'll probably go with the better conversion.

Anyone know what the latest firmware update did? Any effect on settings?

According to Smackrabbit, who seems to know about such things, the PS3 only outputs 4:4:4 YCbCr

I thought that under display-rgb mode-limited/full was (limited=4:4:4) and (full rgb=4:2:2)?

Usually it's limited 16-235, full 1-255. 422 and 444 are both YCbCr, so are 16-235 by default I think.

Zoyd,

Hello! Thank you for the constant updating of tv calibration settings! Question, your most recent settings seem to have a yellowish/greenish hue in certain places. I did some searching and noticed you addressed this before as one getting adjusted to the "warm2" settings. However, I've been using warm2 for years, and following your particular settings since the inception of the thread.

So, are there any reasons why I might noticing this now? Also, what would you say the biggest benefit of these new settings are over the prior ones? The scenes definitely look darker, but I'm not sure if I'm getting crushed blacks now.

Thanks!

Hi Nkon - I don't remember talking about yellow/green hue problems but have always recommended the warm2 preset. Those new settings are a bit experimental at the moment, thanks for the feedback. Based on Larry's comments and my own experience with them, I will be adjusting them a bit in the near future. Also, how they affect your display is very hard to diagnose because the settings will vary from display to display, especially the 10pt controls.

Yep, this should serve as a reminder to everyone. If you are consistently seeing unnatural hues in COPIED calibrations, then there is a high probability that the settings are simply not correct for your TV. Try disabling 10pt at least. Better yet, buy an i1D2 and do your own 2 pt white balance (super easy) - guaranteed better results than borrowing other people's real calibrations.