FWIW, I've noticed a number of different effects on the 42HDS69 that could probably fall into the category of "flickering issues", including floating/fluctuating blacks, clamped/limited white levels (ie contrast), pixel orbiting/shifting, and various interlacing and deinterlacing effects.
If I'm interpreting remarks in
this thread correctly, it sounds like the floating/fluctuating black levels may be the effect that bothers some people the most. The problem is worst with the Dynamic Contrast Mode enabled, which produces a pretty unattractive and overly-enhanced picture, so it's not much use anyway. Once the Dynamic Contrast Mode is disabled, all that's left (for the most part) is some relatively minor floating of blacks which is only really visible (to me anyway) with rather low or no surrounding illumination. FWIW, this seems to be a fairly common issue on plasma displays, and not something totally exclusive to just Hitachis. The floating of blacks in the Normal Contrast Mode is sort of similar to the way blacks float on a CRT. So this may be a feature deliberately designed to give the Hitachi plasmas a more CRT-like contrast response. As long as there's a decent level of ambient illumination though, it should be less noticeable.
If you wanted to stop the white level/contrast clamping as well (which is probably not as beneficial), you should be able to locate the clamp point by displaying a high APL (ie bright) image, and then raising or lowering contrast until you find the setting where the changes in picture intensity stop. Setting Contrast below this point should stop the clamping. The downside is that darker imagery will appear less dynamic.
"Clamped" and "crushed" contrast/white levels are two different things btw. There are no signs of white
crush on the display that I can detect. The
clamping is simply designed to keep the contrast in the picture from becoming too overpowering to your eyes in bright scenes.
The clamping is dependent on the average picture level (APL) of the image. IOW brighter images are clamped at lower contrast levels than darker images. On tests with DVE the clamping seems to work roughly as follows. (This is for an HDMI player w/unexpanded palette. More details
here):
CONTRAST CLAMPING RESULTS:
Code:
-----Contrast Clamp Point-----
APL Day-Dynamic Day-Normal Night
20% 100% 77% 77%
40% 100% 70% 70%
60% 59% 54% 59%
80% 33% 29% 32%
100% 20% 17% 19%
This is with Brightness adjusted close to DVE spec. Since the clamping is dependent on APL, the clamp points will likely change depending on how other controls (eg Brightness, Black Enhancement, Day/Night and Contrast Mode) are set on the display or input device.
The floating blacks may also work a little differently depending on which Day/Night setting is used. See some tips
here for more.
If you see the whole picture shift position by a small amount every few minutes, that's probably the pixel orbiter/shifter which is part of the SCREEN SAVER functions in the SETUP menu.
That brings us to interlacing/deinterlacing issues, which is sort of a complicated subject. First thing you should know (in case you already don't) is that the 42HDS69 is an interlaced display, unlike the vast majority of other flat panels which are progressive. Rather than updating all pixels on the screen in every refresh, an interlaced display alternates between updating just the odd and even lines with every refresh. Hopefully that's not a new concept to most folks on this forum, since interlacing has been with us since the 50's on standard NTSC 480i TVs. The 42HDS69 has 1080 interlaced lines rather than a mere 480 though.
Both technologies (interlaced and progressive) have their advantages and disadvantages. The main advantage of a display like the 42HDS69 is it can display 1080i in it's native interlaced format. This can be particularly beneficial with field-based 1080i video which is more difficult to deinterlace to progressive frames than film/frame-based content. An interlaced display shoud retain more of the original look and feel of field-based content, motion and detail-wise, because it doesn't have to "invent" information to fill in the missing lines the way progressive displays do with field-based 1080i. Interlaced displays may also give film-based content a somewhat "live-er" feeling as well, due to less repetition of information on the screen (since the odd and even lines are separated in time).
The downside is that interlaced displays may exhibit slightly more noticeable flicker, jitter or twitter on pictures with alot of fine vertical detail, especially thin horizontal lines. A slight amount of blurring is often applied to video in a vertical direction to minimize the interlaced twitter on horizontal lines and details. High sharpness and edge-enhancement on the display may restore some of the twitter though (this can be seen in DVE's multiburst vertical pattern at Title 13, Chapter 3). Some people are bothered by the twitter/jitter/flicker. Others less so. From what I've seen looking at various test patterns, especially via a computer input, the 42HDS69 applies some edge/detail-enhancement in all of the "NR" modes, except the High setting (with lower Sharpness settings). It might be more appropriate to call the NR modes "detail-enhancement modes", since that's more the way they seem to behave. The Off NR setting seems to have the highest edge/detail-enhancement, and the High NR setting appears to have the lowest (getting close to none at about 10% Sharpness). So if you want to minimize the slight overall flicker/jitter/twitter as a consequence of interlacing alone, the High setting, as some others have already concluded, is likely your best bet. Whether that will look best to you is your call. On some sources (such as computer input or game graphics), it may be difficult to remove ALL the twitter or jitter on horizontal lines or edges, without adjusting the antialiasing or softness on the source itself.
On now to the
deinterlacing artifacts... You might be thinking to yourself: why would an interlaced display need to
deinterlace video in order to display it? Well there are two cases where this can apply-- when feeding the display a different interlaced resolution than it's native display format of 1080i, and when displaying it's native rez either magnified or reduced (especially in a vertical direction). Case #1 = 480i. Case #2 = 1080i with AR modes other than Standard 2.
Case #2 is (relatively) straight-forward. If you want to see a 1080i source with the fewest scaling/interpolation and deinterlacing artifacts, then using the Standard 2 AR setting should map to the screen's vertical resolution at 100% or 1:1. This may be more crucial on field/video-based 1080i content than on film-based content, since the latter can probably be deinterlaced and then reinterlaced more effectively. (Note that the AR modes may work differently on other Hitachi models with a native rez of 1024i rather than 1080i.)
Case #1 is slightly more involved. The display's internal processing on 480i signals probably goes something like this:
Step 1: 480i deinterlaced to 480p.
Step 2: 480p scaled to 1080p
Step 3: 1080p re-interlaced to 1080i.
The most crucial step from a user's standpoint is the first one. If the 480i source isn't deinterlaced well to begin with, then bobbing, or weaving/combing, and more jagginess are usually the result. To complicate matters, the deinterlacing can take place either in the display or the source (if you're using a progressive or upconverting DVD player for example, or upconverting 480i with a receiver).
Since 480i comes in alot of different flavors (film or video-based, field or frame-based), the component doing the deinterlacing has to be very dextrous to avoid unnecessarily dropping into "video mode" and simply doubling up the lines, which is what produces the more pronounced bobbing/flicker and jagginess on horizontal and diagonal edges and lines.
The display will simply look at the picture information in the 480i signal and try to make an educated guess how to perform the deinterlacing based on that. This is known as cadence detection. Some displays are better at this process than others. The quality of the signal may also be a factor.
Progressive and upconverting DVD players rely primarily on flags, which works pretty well on most maintstream DVDs, but is not 100% reliable either. Better players may use both flags and cadence detection, and/or more sophisticated motion adaptive techniques. (There's a good article on this
here.)
If you want to see what bobbing/jaggies/line-doubling look like, feed your 42HDS69 some anime/or 2D animation at 480i and turn the AUTO MOVIE MODE in the Video settings off. This puts the display's deinterlacer into "video mode". Horizontal lines should flicker or bob up and down noticeably, and diagonal lines should have a stair step effect. The effect is similar to interlaced twitter, only more ugly and pronounced.
When AUTO MOVIE MODE is turned on, the display will search for the distinctive 3-2 cadence in film-based 480i content, and use a process called
reverse telecine to deinterlace it. Under normal circumstances this should produce a much smoother-looking image, free of deinterlacing artifacts associated with the simpler line-doubling technique employed by the video mode.
If you see noticeable bobbing, jaggies or other line-doubling type effects with AUTO MOVIE MODE enabled, then the 480i source may be video or field-based rather than film-based, or the display may simply be having difficulty keeping track of the film cadence, and is switching from reverse telecine to line-doubling until it's able to pick the film cadence up again. I've noticed this with various film-based DVDs, including opening scenes in SW:ANH and some animated titles. Sometimes you'll see the line-doubling come and go, as the display tries to lock onto the film cadence. If this starts happening try letting the player do the deinterlacing instead, by switching it's output from 480i to 480p (or 720p or 1080i if it upconverts). It's possible the display may have an easier time with digital 480i inputs (ie HDMI), than with analog as well. That's a theory I've yet to test though.
Using a 480p input generally seems to produce less in the way of motion artifacts on the 42HDS69 as well (though that may vary with the player). So I try to avoid using 480i whenever possible.
If you're using fluorescent light or dimmers for your background/room lighting that could also contribute to the sensation of flicker on the display or eye fatigue.
More here.
I think that about covers most of the flickery type stuff I've seen. The only issues above that are probably unique to Hitachi's ALiS displays are the interlacing issues. All of the others (including the
deinterlacing artifacts) also can occur on progressive displays.
