APL windows Pro's and Con's
-
Now Available: Tech Talk Podcast with Scott Wilkinson, Episode 19 Click here for details.
1 - 13 of 13 Posts
Are you asking for use on your own display or just arguments in general for various display types?
I would say there's a strong argument for using them with any "dynamic backlight" LED LCD or any display with a dynamic iris.
With other displays, like plasmas, I still like the idea of using them, but I can't say with certainty.
I would say there's a strong argument for using them with any "dynamic backlight" LED LCD or any display with a dynamic iris.
With other displays, like plasmas, I still like the idea of using them, but I can't say with certainty.
Joined
·
3,458 Posts
One possible negative would be if the other on-screen levels contaminated the item you want to measure. This would not apply with a contact meter and displays like direct-view LCD, LED, and plasma. It would be more likely with non-contact meters and displays with lower ANSI contrast, such as rear-projection where there's a bit of internal reflection. On my RPTV any contamination from the other on-screen levels is generally not enough to consider, but I suppose it's possible.
I do not have enough experience with Plasma displays to really comment if the APL patterns might be useful for Plasma. What I would do is measure a typical set of windows, and I'd make another run with the APL patterns. I'd determine the relative Y values for all the measurements and compare what windows show against the APL measurements. Anyway, the main point of the APL measurement patterns is for relative Y, and I'm not sure how much plasmas typically vary.
I do not have enough experience with Plasma displays to really comment if the APL patterns might be useful for Plasma. What I would do is measure a typical set of windows, and I'd make another run with the APL patterns. I'd determine the relative Y values for all the measurements and compare what windows show against the APL measurements. Anyway, the main point of the APL measurement patterns is for relative Y, and I'm not sure how much plasmas typically vary.
Plasmas vary a lot in terms of picture level in my experience. Its either a flagging power supply issue (like CRTs) or its some manufacturer mechanism to prevent black detail getting washed out with higher level content. I recommend APL patterns over any others in pretty much every display environment and type. bear in mind there is very little chance of a complimentary background contaminating a meter , Meters read small areas and light does not turn corners very well.
Thanks for the replies. Chad, I guess I was asking in general terms but I see from your answer that you covered my type of display. I have a Samsung B8500 and as you already know (I see were you have calibrated and done a review on this display) it is a backlit LED LCD with local dimming. The last time I went through a calibration I used the larger APL windows (the first 30 times I did not) and maybe it is just my imagination but my end results as far as what my eyes see are the best to date. Now as far as what Calman and it's graphs say there was not a lot of difference?
So here is my next stupid question. I am using a PS3 with the AVSHD disc to display my patterns. So If I put up a 100% APL white window pattern it is going to give me an average picture level output. The question is, it is an average of what and how does this compare to a non APL window? Both windows have to put out a certain level, is there a standard for this? I'm not sure how this will effect the display you are calibrating seeing how it has no idea what type of pattern you are displaying. I'm sure that I am missing something simple here but I guess If I don't ask I won't know.
Thanks
So here is my next stupid question. I am using a PS3 with the AVSHD disc to display my patterns. So If I put up a 100% APL white window pattern it is going to give me an average picture level output. The question is, it is an average of what and how does this compare to a non APL window? Both windows have to put out a certain level, is there a standard for this? I'm not sure how this will effect the display you are calibrating seeing how it has no idea what type of pattern you are displaying. I'm sure that I am missing something simple here but I guess If I don't ask I won't know.
Thanks
Joined
·
3,458 Posts
Maybe that was a poor word choice. Primarily the point of my first paragraph was that the other on-screen levels could possibly have a small influence on the measurements for certain displays or situations. For example I believe on my RPTV the 0% and 10% will measure a noticeably higher Y with the large APL patterns than from windows. If you take black into account in the calculation of relative Y levels it's probably a near wash, but if you're doing something like trying to set brightness based off of a 10% measurement and target gamma while ignoring black it might be relevant. Generally I consider this a somewhat sketchy example, but mainly it's just food for thought on one situation where the pattern choice might not be ideal.
My RPTV uses a fixed light source. When the auto iris is turned off the TV delivers very little variation in measurements regardless of APL. I can measure the small APL pattern or I can measure a field, and they'll deliver almost exactly the same reading, although APL is vastly different for the two patterns. Because there's basically zero variation in my TV when the adjusting iris is disabled there is no reason to use the APL measurement patterns, but that is not necessarily how many current digital displays operate.
Maybe that was a poor word choice. Primarily the point of my first paragraph was that the other on-screen levels could possibly have a small influence on the measurements for certain displays or situations. For example I believe on my RPTV the 0% and 10% will measure a noticeably higher Y with the large APL patterns than from windows. If you take black into account in the calculation of relative Y levels it's probably a near wash, but if you're doing something like trying to set brightness based off of a 10% measurement and target gamma while ignoring black it might be relevant. Generally I consider this a somewhat sketchy example, but mainly it's just food for thought on one situation where the pattern choice might not be ideal.
My RPTV uses a fixed light source. When the auto iris is turned off the TV delivers very little variation in measurements regardless of APL. I can measure the small APL pattern or I can measure a field, and they'll deliver almost exactly the same reading, although APL is vastly different for the two patterns. Because there's basically zero variation in my TV when the adjusting iris is disabled there is no reason to use the APL measurement patterns, but that is not necessarily how many current digital displays operate.
Joined
·
3,458 Posts
In a typical window pattern you have two on-screen video levels. You have the center measurement area, and you have the black background. The pattern marking indicates the video level of the center area.
If someone really wanted they could calculate an average brightness, or average level, for all the on-screen pixels. There really is no reason to do this, because typical video material offers a vast range of video levels, from a full screen of black to a full screen of white. The only relevant item to note about window patterns is that a 100% window will overall be brighter than a 10% window. A 100% video level should output more light than a 10% video level, and since the areas are the same in a window, the 100% window has a higher average screen brightness than the 10% window.
The APL measurement patterns are intended to output the same amount of light across the entire screen for a series of measurements. If someone actually wanted to add up the couple million pixels they would find that both the 10% and 100% APL patterns have the same overall screen brightness. While typical window patterns vary overall brightness along with the measured video level, the APL measurement patterns hold screen brightness constant. Only the measured video level changes in the APL measurement patterns as the overall light output remains the same. The primary difference with the APL measurement patterns is that the average brightness of the screen does not vary, unlike typical windows that vary average light output as the video level changes.
Some displays will vary light output depending on the signal they receive. For example with some displays you may find that the 100% video level will measure a different Y depending if you measure a window or if you measure a field. Both the 100% window and the 100% field share the same video level, but the overall brightness of the pattern is lower for the window (because of the black background) than the field. It's already been noted that overall screen brightness also varies when measuring a series of typical windows. Where you may run into issues is if you are using a procedure that assumes Y remains constant on a display that tends to vary Y depending on overall screen brightness. While there are some displays that can measure a consistent Y regardless of the overall screen brightness, many digital displays (and certain settings) tend to vary Y for a video level depending on overall screen brightness. The point of the APL measurement patterns is to hold overall screen brightness constant in order to defeat variation from the display.
In a typical window pattern you have two on-screen video levels. You have the center measurement area, and you have the black background. The pattern marking indicates the video level of the center area.
If someone really wanted they could calculate an average brightness, or average level, for all the on-screen pixels. There really is no reason to do this, because typical video material offers a vast range of video levels, from a full screen of black to a full screen of white. The only relevant item to note about window patterns is that a 100% window will overall be brighter than a 10% window. A 100% video level should output more light than a 10% video level, and since the areas are the same in a window, the 100% window has a higher average screen brightness than the 10% window.
The APL measurement patterns are intended to output the same amount of light across the entire screen for a series of measurements. If someone actually wanted to add up the couple million pixels they would find that both the 10% and 100% APL patterns have the same overall screen brightness. While typical window patterns vary overall brightness along with the measured video level, the APL measurement patterns hold screen brightness constant. Only the measured video level changes in the APL measurement patterns as the overall light output remains the same. The primary difference with the APL measurement patterns is that the average brightness of the screen does not vary, unlike typical windows that vary average light output as the video level changes.
Some displays will vary light output depending on the signal they receive. For example with some displays you may find that the 100% video level will measure a different Y depending if you measure a window or if you measure a field. Both the 100% window and the 100% field share the same video level, but the overall brightness of the pattern is lower for the window (because of the black background) than the field. It's already been noted that overall screen brightness also varies when measuring a series of typical windows. Where you may run into issues is if you are using a procedure that assumes Y remains constant on a display that tends to vary Y depending on overall screen brightness. While there are some displays that can measure a consistent Y regardless of the overall screen brightness, many digital displays (and certain settings) tend to vary Y for a video level depending on overall screen brightness. The point of the APL measurement patterns is to hold overall screen brightness constant in order to defeat variation from the display.
To be honest I can't come up with a pressing reason as to why apl patterns shouldn't be the norm for every display type. I suspect the only reason we had window patterns on black in the first place is they were comparatively easy to generate at the time.
Quote:
For the life of me, my Mits 65833 could never get a flat or close to flat gamma and greyscale. Once I used the APL windows, I was able to obtain a flat response in greyscale that I never achieved by simply using a window based only patter and my gamma was all over the place. The IRIS is not defeatable, so using the APL window seems to take care of this issue and now I have a flat gamme of 2.22 through the entire range...Im sticking with APL windows..
Originally Posted by Mr.D  To be honest I can't come up with a pressing reason as to why apl patterns shouldn't be the norm for every display type. I suspect the only reason we had window patterns on black in the first place is they were comparatively easy to generate at the time. |
I recently calibrated my TC-P50G20 and noticed the other day that the grayscale changes fairly drastically in the lower end (20%-30%) and get much more red when there's other content, such as the TV menu, on the screen. Because of this behavior does it make sense to recalibrate it using an APL window or have other content on the display at the same time as the normal window?
I'm thinking this may be what's making my shadow detail look like heck anytime they have reds or flesh tones in them.
I'm thinking this may be what's making my shadow detail look like heck anytime they have reds or flesh tones in them.
Joined
·
3,458 Posts
I read this and your other question. My question would be if the grayscale actually measures xy any differently with the APL patterns. The xy is the part of an xyY measurement that represents color, and the Y is lightness or brightness. On most displays the xy wouldn't be expected to vary much with the APL patterns. The APL patterns are primarily intended for relating how different Y measures compare.
1 - 13 of 13 Posts
