L90D+ LCD Misleading specs

[djotai]

I've examined the Hyundai L90D+. The color depth on their site
is advertised as 16.7 mil. Well, I've used it and am certain that
it is a TN+Film type - the viewing angles are extremely limited.
This type of panel is 6 Bit + 2 bit dithered, with all the shortcomings
of such a panel. Representing it as 16.7 mil is misleading.
This doesn't mean it's a bad panel - but one should be aware of
what he is buying and a company should not mislead customers.

 

[ChuckHsiao]

Just wondering if you've tested it out. 6-bit panels can only produce 253 colors (as opposed to the 8-bit's 256), which is why they end up with 16.2 M instead of 16.7 M. If I understand correctly, the three missing colors are 253, 254, and 255 (on a scale of 0 to 255). So you can just open up paint, and make blocks with 252, 253, 254, and 255 color values all next to each other, and see if you can tell them apart. You should make sure Windows is working at a 32-bit instead of 16-bit color depth of course. Unfortunately, I haven't had the chance to test this out personally (all the monitors I've worked on have been 8-bit, so they all passed this test), so the missing values might be at the beginning instead (i.e. 0, 1, 2) but I'm fairly sure they're at the bright end. If you can see the different shades (at both the bright and dark ends), it means that the monitor is 8-bit. If you can see the different shades at only one end, it means that it's 6-bit, with the other end being the missing (i.e. unsupported) values. Hope it works.

Chuck Hsiao
Amptron

 

[rise]

when this was the monitor to get this issue was brought up and a member emailed hyundai about it and got the official word. its somewhere in one of the "hot deals" threads on the monitor but if you think i'm going back in there and getting it youre crazy

 

[cockeyed]

HERE is an article that sheds some light on 6-bit / 8-bit color and the math used. When I see an LCD listed faster than 16ms with 16.7m colors, I become suspicious.

 

[ChuckHsiao]

Thing is, there's been lots of advances since then. Nowadays, it's possible for an 8-bit monitor to have good response times (for example, if the controller circuitry includes overdrive).

If you got the monitor, you can test for if it's 6-bit or 8-bit here:
www.amptron.com/chuck/6bit8bit.bmp
www.amptron.com/chuck/bittest.bmp

Simply see if you can see the border between the boxes. This is most easily seen by looking at the monitor diagonally, i.e. not from above nor from the side but both (i.e. from the corner of the monitor), where the monitor's viewing angles are most atrocious.

Chuck Hsiao
Amptron

 

[xtknight]

Originally posted by: ChuckHsiao
Thing is, there's been lots of advances since then. Nowadays, it's possible for an 8-bit monitor to have good response times (for example, if the controller circuitry includes overdrive).

If you got the monitor, you can test for if it's 6-bit or 8-bit here:
www.amptron.com/chuck/6bit8bit.bmp
www.amptron.com/chuck/bittest.bmp

Simply see if you can see the border between the boxes. This is most easily seen by looking at the monitor diagonally, i.e. not from above nor from the side but both (i.e. from the corner of the monitor), where the monitor's viewing angles are most atrocious.

Chuck Hsiao
Amptron

OK..I can see a very small difference between the two images in those pictures. Does that mean I have a 6-bit panel with high-quality 2-bit FRC or does that mean I have an 8-bit panel? Actually I am 95% sure I have a 6-bit panel but I'm asking for the sake of everyone else. I guess in that case, my 6+2-bit panel has such good emulation that it looks like an 8-bit anyway?

 

[ChuckHsiao]

Whoops I should've been more clear. If you can see a difference between the shades, it means you have an 8-bit panel. If you can't see the difference, it means you have a 6-bit panel.

I should say as a caveat that I've yet to hear someone say they can't see the difference with a 6-bit panel, so I don't know if it really works -- my information is the same as your information, it's off the Internet (so it must be true!). All my monitors have been 8-bit, so they've naturally all worked. It's possible that the three missing colors of a 6-bit panel are 0, 1, and 2 (instead of 253, 254, and 255) but you can just see if you can see the black shades in bittest.bmp. Unfortunately the borders for black are harder to see than the borders for white. Oh well. For that one, if you can see shades in both black and white, then you have a 8-bit for sure. If you see shades in only one of those, then you have 6-bit. You will be able to see at least either white or black, so if you can't see any, you're not looking hard enough. Heh. I didn't see them when I looked at first.

Oh yeah, I also forgot to mention, makes sure that Windows is set to 32-bit rather than 16-bit. I don't know if it really affects things, but might as well as eliminate unknown variables (i.e. how Windows selects shades in 16-bit, which doesn't cover all 8 bits per color).

Chuck Hsiao
Amptron

 

[cbehnken]

Dithering makes it possible for 6-bit monitors to simulate the missing shades. Often it is difficult to detect, but easily seen in colorful movies.

 

[xtknight]

ok...thanks for the info...well I can just BARELY discern the difference in the black image. I certainly wouldn't know there was a transition if I wasn't looking for it though. Either way I already know I have a 6-bit panel so that doesn't help me I guess. I have a 6-bit panel and I can differentiate (though barely) between the two...but what I'm really asking is: how in the world would I tell the difference in real life between a 6-bit and an 8-bit panel? If I have to look at these for 5 minutes at all different angles to finally decide if there's a difference, there's no way I'm going to see that in a motion picture, like cbehnken said?

i heard the eye is most sensitive to green, so would it help if this image were variations of green +/- 2? would it be easier to tell the differences?

 

[ChuckHsiao]

Uh oh. 6-bit panels aren't supposed to be able to show the final colors in the spectrum (either the brightest whites or the darkest blacks). It's possible that those guys have gotten the controllers to wise up to simulate them though (it just requires modifying the circuitry and the dithering algorithm at those ends). Hehe so you got a 6-bit monitor where you can see the individual boxes on the first and third rows for bittest.bmp? Maybe it's really an 8-bit monitor.

I didn't really consider color, since fundamentally LCDs are just black-white shutters (the color comes from the color filter). It might be possible, but alas, my coworker's already gone for the day so I won't be able to upload a new one until tomorrow. People say that they can distinguish 6-bit and 8-bit by looking at the color quality, but I'm not one of them, so I can't tell.

6-bit monitors can actually only show 262k colors. But through dithering or frame rate techniques, they can simulate 16.2M colors. However, they can't simulate one of the extreme ends of the scale (either the blackest black or the whitest white, I'm not sure which; I think the latter though). This is why this test works. Or is supposed to. Grumble.

Oh yeah: did you set Windows to 32-bit before looking for the boxes?

Chuck Hsiao
Amptron

 

[daveybrat]

I had the L90D+ and going from my 19" Samsung 950P CRT was a horrible experience.
The colors in games were night and day to me. The dithering effect in Call of Duty was terrible and nowhere near as vibrant and colorful as my CRT.
Movies also looked much worse.....

the only redeeming quality it had was that the text was crisp and the screen was very bright while surfing the web. I returned it though as i'm not spending that much on a new monitor to surf the web.

 

[dfloyd]

Speaking of fast response times and 8 bit color there is a monitor out that does both.

In fact it does both and throws in a bag of chips as well. Or so the saying goes.

The new Viewsonic VP191B.

An 8MS Repsonse time 16.7 Million Color, MVA panel that never goes above 16 ms response according to the french review site that tested the new model (Viewsonic did something dumb, they updated the model with a new panel but did not change the model number, so the majority of reviews on the net are still for the old panel).

Like the person above stated, its all in the new tech, overdrive and such.

 

[djotai]

BEFORE buying an LCD, do yourselves a favor and read the article
mentioned in Cockeyed's post above. Most LCDs today are good - but
they differ in performance according to the type of panel used.
The point of my post was that the manufacturer should not mislead
customers by providing inaccurate or ambiguous data on their
sites.
Study the characteristic differences in performance among panel types -
plenty of information available online - then get an email statement
from the manufacturer of the LCD you want to buy as to which type
of panel they use for that LCD.
This way you'll be able to know in advance what you're buying and
what to expect, so the chances of dissapointment diminish.

 

[ChuckHsiao]

Hmm did your 6-bit monitor show borders for both the first and third rows on bittest.bmp? If so, it's possible that they've upgraded the controller circuitry to cover up those missing colors now, and that this is no longer a valid test. In which case, a better test might be to produce a continually varying shade spectrum.

Unfortunately I've used nothing but 8-bit LCD monitors, so I have no idea how much the quality suffers when using 6-bit -- it's just that people have mentioned the difference, that 8-bit has better colors but worse response time, while 6-bit has worse colors but better response time.

Oh yeah, having Windows in 16-bit does destroy the test completely, so this test has to be done in 32-bit Windows.

Chuck Hsiao
Amptron

 

[cockeyed]

Visit BenchmarkHQ and download the TFTtest program at the bottom of the page. It has some very good test patterns that I think might show dithering effects. This is a Russian website in English that has many test programs for monitors. Maybe one of these would be useful for checking 6-bit, 8-bit, color display.

Edit: In the top row of test patterns, use #2 (Diagonal fade), #3 (Color bars), # 5 (circle fade), these should all reproduce with smooth color. Try setting the video properties to 16-bit color and run these again, the dithering effect will jump out at you. My 16ms Acer AL1912 LCD, rated as 16.7m colors, reproduces these patterns with smooth color when the video properties are set to 32-bit color.