For those struggling with brightness, contrast & backlight settings on LCD's - Page 4

I'll probably change the PDF to basically read if the Black Clipping test does not agree with with the APL Clipping test it's an acceptable compromise to use the lower setting from the two tests.

It currently says "Turn black-level no lower than the setting where it is possible to see 19-28 flash. Try for just barely being able to see 19 flash." for the APL clipping pattern. Do you plan on changing that to 20 or will you keep it at 19 (which is in agreement with other posters who have used both the DVE disc and the AVS disc)?

Probably a daft question, but there really is no scientific approach to setting brightness? It seems like the only thing that no one performs without a meter?

Video black is a standard and you set your display so you can see what you are supposed to see and don't see what you are not supposed to see with your eyeballs.

And then there is contrast.

And sharpness....

I haven't decided. Using 20 just means that redwolf4k doesn't have anything to complain about, because digital displays with questionable near-black performance in dim lighting are likely to end up exactly the same as DVE. I prefer 19 simply because it tends to agree more with the standard Avia patterns, and it means there's one more level above video black that cannot be cut off. Neither is absolutely 'correct', because of items like differing display performance and viewing conditions.

A meter can be useful, but it's not fool-proof. The main issue I see with measurements would be gamma, because there's little clearly-defined intent in this area in relation to how various displays actually perform. I've read about people on AVS taking an adjusing-iris display, measuring windows, and then using a computer or processor so window gamma measurements report as constant gamma. Wth the way adjusting backlight or iris features work, it's very unlikely such an adjustment would look at all similar to what the content producer intended. When people use items like 10 point gamma controls on LCD and Plasma displays there's likewise no guarantee that you're coming closer to the original intent, because unless you know the performance of the display used for production the actual intended gamma performance is simply not exactly-defined.

Here's one experts take:


In fact, the entire thread is fascinating. http://www.avsforum.com/avs-vb/showthread.php?t=1312439

Ok, I currently have left it at 19 just flashing on both clipping patterns (my CCFL-LCD doesn't vary light output based on APL).

If you look at a thread I created earlier this week (30 or 35 fL ideal for dark room), you'll see that a pro calibrator suggested that 2.35-2.4 seems to look better on LCDs and more closely matches 2.2 on plasmas. Something to keep in mind.

I posted in your thread.

There are lots of opinions but I believe Steve Shaw knows video better than most. Check out his company - LightIllusion.com

Where I was going on the gamma comment was basically to say that gamma is probably less well-defined than setting brightness. Most people tend to use windows for gamma measurements, and various technology displays do not all result in similar Y measurements using windows. Broad statements like "2.2 should be your target to get accurate representation of the images" can serve as a basis for clearly mis-guided 'calibration' attempts, such as using a video processor to make adjusing-iris or adjusting-backlight displays measure windows similar to other display types.



Like duke32 mentioned, a less subjective way of setting contrast is to use measurements.

You'll get no argument from me but I see no harm presenting the views of those who know a lot more than I.

One of the beauties of owning calibration equipment is the ability to endlessly play with your own displays so I try all sorts of combinations. The one thing I've learned is that you can't carry numbers from one display to the next and from one viewing environment to the next. Every TV seems to have its own demands....

This has been an interesting thread on black level and I've had a ton of fun making changes and trying to understand the results. Lord knows I'd give anything to have identical displays side by side as my visual memory when changing parameters is poor at best. Remembering what was seen relative to what is seen is something that evades the brain.

Here is something I use that is most interesting in a very dark environment. The patterns are from an Accupel DVG-5000.

From the manual:

Windows w/PLUGE - Calibrate color temperature, grayscale, and
gamma. PLUGE bars ensure black level doesn’t shift while adjusting grayscale
tracking. The background of the High-GS windows is 100% luma to
examine a display’s ability to differentiate levels near 100%.

So, I've now had the chance to compare three different brightness settings: 41 (19 and above flashing), 40 (20 and above flashing), and 43 (17 and above flashing). I definitely get some black crush with 40 and 41 and the remaining shadow detail is too dark and hard to see even in a completely dark room. My display's gamma is around 2.4 and the black levels are very good at any of these settings. Therefore, all lowering brightness does is introduce black crush and reduce shadow detail substantially. I cannot recommend the approach suggested by redwolf to anyone with a display that has a very good black level and darker gamma (average of 2.35 to 2.4). The only thing it does on a display like mine is crush the 2-3 lowest levels just above black and reduce shadow detail considerably.

My display is a 55 inch Samsung LCD television. I tried both the AVSHD and DVE patterns. I find that when I get up closer to the televison, I can easily see the dithering on the DVE ramp pattern if I am looking slightly down on the pattern or at an angle from the left or the right by about 15%. If I am looking directly straight on, I have to turn the brightness control up in order to see it. The same applies with the AVSHD pattern. So it appears that I can easily see where the display is activating by looking closely at the display up close from different angles. Should I set my brightness in this fashion? I had previously sat back from the display by about 10 feet and adjusted brightness by watching the flashing pattern.


What method do you think would be best for my display type?

I prefer to use the black clipping pattern (not APL) and set it to the lowest brightness setting that still shows digital 17 flashing.

Razz, I think this post sums it up quite nicely.

Sorry for the ignorance. Could you please explain what MLL is and how it is measured?

I now feel a sense of frustration with my Samsung LCD displays. It seems that the brightness may be optimally adjusted for those sitting directly in front of the display while those sitting even 10 degrees off to the side will see a more washed out picture. Or it may be adjusted to those sitting slightly off to the side and then those sitting directly in front of the display will see crushed blacks. This is a bummer. I wonder if perhaps plasma displays have this issue. I suspect that I should not get up close to the display as suggested above and just sit back where the display will be normally viewed from and then adjust the setting from there. It is after all from this location that the picture will be viewed from.

This is an across the board issue with LCDs. Some LCD's are better, particularly IPS panels. But really if you want something that doesn't shift, you need to go to plasma or something else.

MLL is minimum luminance level or the black level of the display. Is is how much light the TV outputs on a completely black screen. I view my LCD from the center and set brightness by standing directly in front of the screen up close to set brightness.

Razz1 - where in MN are you located? This "might" be your lucky day....

You could set brightness for 2 seats at the same angle from center, but otherwise it's sort of a given that on some displays the image varies depending on viewing angle.

This is what I figured might be the case. I think perhaps a plasma might be something for me to consider in the future.


I am in Albertville, right next to the Outlet Mall.


Thank you for your response and explanation of MLL.

I see allot of different suggestions here. and then arguments. about 17 and 19 so what is the process to use ? Can someone summarized and most agreed upon method and where should the back light be set during this? Is the second post here still valid?

Thanks for clarifying.

If you're using the AVS HD 709 patterns, the APL Clipping pattern (Basic Settings - Chapter 2) is intended to be comparable to commercial patterns when you can barely see 19 flash from your typical viewing location with usual room lighting.

Set the backlight to whatever setting you want to use for typical viewing.

Both the backlight control and the black-level (brightness) control can affect where you will be able to see flashing on the APL Clipping pattern. There are different approaches for how to go about setting both controls, and it may also depend on the TV. Personally I tend to use the black-level control primarily to set where a digital display clips, and I tend to use the backlight control mainly to adjust the overall image for room lighting. Personally I start by setting the backlight very high to make it easier to see where the first level above the display's black actually begins flashing. On better TVs you may need to make the room entirely dark, use the black clipping pattern, get right up next to the TV, and even then it might be difficult to spot where exactly the flashing starts, but on other TVs it's very easy to tell exactly which levels are being cut off from the black-level (brightness) setting. After I have an idea where the black-level control causes clipping, I turn down the backlight for room lighting. This process of setting black-level first and then adjusting the backlight will not work for all displays, but it's how I would try to demonstrate that the black-level (brightness) control generally has a different function than the backlight control.

There are some gaps between the general theory of calibration and how displays actually perform. As a consumer you can buy a number of different types of displays, and generally the displays have some slight differences in how they perform. Basically with calibration the intent is to try to get various displays to look somewhat similar, and in order to do this it's sort of difficult to make absolute rules, since again various displays can perform differently.

- In theory 17 is expected to represent information that is probably imperceptibly lighter than black. Ideally you would not want to cut off this information, but on the other hand the difference against black is likely expected to be smaller than possible on most digital displays.

- Also in theory 20 is expected to be slightly lighter than black. It's rather difficult to show just how much lighter than black using video, and some displays would also vary depending on the exact image displayed.

Realistically many LCD displays come out of black much more quickly than a CRT display would with the exact same input. With some LCD displays you may be able to use room lighting to help mask the relatively poor near-black performance, and in such a case it might be fine to set 17 to flash when viewed near the screen in a completely dark room. Other displays have even worse near-black performance, where the first level above black is clearly visible even with some room lighting, and in such a case it may be a reasonable compromise to try to get the display to look more like other displays by cutting off a bit of information near black. My basic point is that there are not really absolute rules, because different displays do not perform exactly the same. To get two displays to look sort of similar in a side-by-side comparison you may need to set them slightly differently.
Edited by alluringreality - 2/21/13 at 4:59am

Thanks for the "mini" guide. :-)

Is there someone who can tell how this is done using HCFR?

I haven't followed the HCFR fork after Version 2.1, but on the old version you can look at the measurement information in terms of RGB to find where a display might clip near white. On the right side of the HCFR Measures screen there is a "Display" setting that has pickboxes, which has choices like xyY or RGB. The measures screen also has a "Selected color" area that also displays R, G, and B numbers. There are also the "Target" bars that show how red, green, and blue relate to each other as a percentage (%), but I suggest looking at the R, G, and B numbers that generally are comparable to Y.

If you measure grayscale you would want the R, G, and B numbers to be approximately equal. As you turn up the contrast control you may find that one of the colors might stop increasing along with the other primaries. For example say a TV runs out of red, so R will get to a point where the R number might go no higher, but the B and G numbers can get much higher. Since R is unable to get any brighter, it would be necessary to set G and B to that same maximum value in order to display gray.

I think this has been brought up many times over but here it goes .....again
On the black clipping pattern i set 17 so that it is barely visible with my nose right up on the screen . One click down and it completely disappears this is too low for 17 , 16 is invisible. If you go one click up and 17 looks obvious this brightness setting is too high. Just get it to where 17 looks "barely visible" this is the guide i use. I have a samsung d series plasma btw. Buzzard who I have spoken too in private mentiioned to go for 17 being barely visible if you can't make it appear barely visible then go with a setting that is one lower .

on various Samsung LED/LCDs I have, there is one setting where 17 is obvious and one click below that makes it disappear altogether... so on these sets, I can either have 17 obvious or invisible, nothing in between

If it can't be made barely visible then go with the setting that is darker.