Sign in with
Sign up | Sign in
Your question

how common are dead pixels?

Last response: in Computer Peripherals
Share
April 15, 2005 5:46:10 AM

how common are dead pixels?
i`m thinking about buying a lcd because crt`s are almost impossible to find but i`m afraid i will get stuck with an lcd with even 1 dead pixel.

More about : common dead pixels

April 18, 2005 5:13:14 AM

Dell 2005fpw #1 - Dead pixels and stuck pixels, sent back
Dell 2005fpw #2 - No Dead pixels or stuck pixels
Dell 2405fpw - Perfect

<pre><font color=red>A64 3200+ Winchester
DFI Lan Party NF4 Ultra-D
1GB Corsair 4400C25PT
WD740GD, WD2000JB, WD1200JB
ATI X800XL
Dell 2405FPW</pre><p>
April 18, 2005 8:35:16 AM

Im guessing that they are somewhat common. If you look at some of the return clauses it will say that unless you have 8 or more dead pixels that you cant return it.

How much are you going to spend? I bought a 37" plasma and love it.

Intel P4 550(3.4)<font color=red>Benchmarked 5Ghz</font color=red>
X800XL 474/609 <font color=red>#5 in 3DMark05 #3 in PCMark04
3DMark05-<font color=red>5965</font color=red>
3DMark03-<font color=red>13496</font color=red>
PCMark04-<font color=red>6170</font color=red>
Related resources
April 19, 2005 5:15:31 PM

how much was that?

i think at the most i would spend is 1000$

SUPPORT AMD. KICK INTEL OUT OF YOUR PC.
April 20, 2005 12:57:35 AM

I would guesstimate anywhere from 10-50%. It's hard to know for certain because manufacturers don't like to give out this sort of information.

In my experience, I've opened and tested about 10 monitors, and 2 of them had 1 dead pixel.

My best advice is to go to a store and try them out beforehand. Make sure to set the screen to white, and leave it there for a while, to let the components warm up. Then set the screen to black, and see if there are any bright pixels. Leaving the screen on white for a while is the key here though -- you want the pixels to warm up for a while to see if that shorts any of them out. Obviously, you can't do this if you order online.

Best of luck.

Chuck Hsiao
Amptron
April 20, 2005 10:01:29 AM

I have extended personal experience with two LCDs - One a an Acer that was on sale, and one is a Samsung. Neither have presented any dead pixels.

Women--can't live with them, can't have heterosexual same-species intercourse without them.
April 20, 2005 6:49:11 PM

Ive seen this Plasma go for as low as $1799.00 which is a steal. But probably more in the $2200.. range is where you would be. I know its alot, but after using a 37" screen I couldnt go back to anything less.


Intel P4 550(3.4)<font color=red>Benchmarked 5Ghz</font color=red>
X800XL 474/609 <font color=red>#5 in 3DMark05 #3 in PCMark04
3DMark05-<font color=red>5965</font color=red>
3DMark03-<font color=red>13496</font color=red>
PCMark04-<font color=red>6170</font color=red>
April 22, 2005 12:40:20 PM

You might find this link interesting:

http://www.newgen.ca/download/5981-6351EN.pdf

The nugget is:

"Bright sub-pixel defects: 3 maximum
Dark sub-pixel defects: 5 maximum
Total sub-pixel defects: 5 maximum
Full pixel defects: 0 allowed

As a result of these tighter specifications, an estimated 60 to 70 percent of all units are shipping with no pixel defects at all (per HP's internal audits), and another 10 to 20 percent are shipping with only a single sub-pixel defect. Most remaining units are shipping with only a handful of pixel defects that fall within the above listed range of acceptable pixel defects."

From statistical modelling for 17-inch and 19-inch monitors (1280 x 1024 x 3 total subpixels), it seems like either some areas of the factory have more dust than others (enough to throw this distribution off because defects will be more "clumped" rather than spread out), or their figures either understate the no pixel defects rate or understate the 1 pixel defect rate. If you assume 70% of all monitors have no defects, then 25% will have 1 defect, 4.5% will have 2 defects, and 0.5% will have 3 or more defects. If you assume 80% of all monitors have no defects, then 17.9% will have 1 defect, 2% will have 2 defects, and 0.1% will have 3 or more defects. The subpixel defect rate (based on those numbers) is between 1 in 11 million and 1 in 18 million. Since I've looked at about 10 monitors and 2 had 1 dead pixel each, right now Amptron's defect rate (based on this small sample size) is about 1 in 20 million, so I'm guessing that the next monitor I pick up will have a dead pixel or two =P so watch out. Amptron's pixel policy is basically HP's pixel policy by the way. I wonder if I can convince my boss to conduct an "internal audit" of my own, and just open up monitors all day and test them for dead pixels. On company time. Somehow I don't think he'd be favorably disposed to it.

If you're curious, the chance of having 6 or more dead pixels (and thus getting a replacement under warranty) is about one in a million, based on the modelling.

Chuck Hsiao
Amptron
April 29, 2005 4:54:23 PM

An update on dead pixels. From the statistical modelling, I figured dead pixels would happen about 70-80% of the time. One of my co-workers, however, figured they happened more than half the time. She figured I'd need to test multiple ones before I found one with no dead pixels. So we bet a meal over if I were to pick a random one, whether or not it would have any dead pixels.

I picked out one at random, a 19" V293. I had tested them before with no problems. I turned it on, let it set for a minute or two, then looked all over for dead pixels (using red, green, blue, and black screens). I found none. So I triumphantly went back and told her to come and check it out. She asked how many I had tested and I said it was the first one. So she came by and looked it over, and lo and behold, there was a dead pixel. Green. Actually it wasn't green, it kind of changed color depending on what angle you were looking at it from, but it was in a green subpixel's slot. Grrrrr. I later found a faint red pixel as well. Of course, what had happened was that I hadn't left it on long enough. I left it on when I went to go get her, which was then enough time for the monitor to warm up and for dead pixels to appear.

So I went back and got another monitor. It had a single dark red pixel.

And another monitor. And it had a single dark red pixel, as well.

Finally on the next one, I found one with no dead pixels. So I was finally able to show her one that didn't have any, on the 4th attempt. Just my luck. Had a streak of no dead pixels until I made a bet with her. Now I owe her lunch.

Moral of the story: If you want a monitor with no dead pixels, get an employee to bet that it will have dead pixels. If that employee is me, there won't be any.

Anyway though I wrote out which monitors I had tested, and could only come up with 9 (not 10). Adding these four makes 13 monitors that I've tested total, but a total of 6 dead pixels (4 of them on those four, grumble). So 6 dead pixels out of 13 monitors comes to a chance of 1 in 8 million per subpixel, which means 61% no dead pixels, 30% 1 dead pixel, 7% 2 dead pixels, and 2% 3 or more dead pixels. Until I make a bet with her that all next five monitors will have a dead pixel or two.

This fits amazingly well with my testing, by the way. So far, 8 out of 13 (61%) have had no dead pixels, 4 out of 13 (31%) have had 1 dead pixel, and 1 out of 13 (8%) have had 2 dead pixels. Now I just need to find someone to make a bet with.

Chuck Hsiao
Amptron
May 1, 2005 7:32:22 PM

in addition to the dead pixel rate refering to brand new monitors, is there a rate at which these pixels die over time also?



<b><A HREF="http://www.digitalgunfire.com" target="_new">DigitalGunfire-Industrial EBM</A></b>
ASUS P4S8X-P4 2.4B - 2x512M DDR333 - ATI 9500Pro - WD80G HD(8M) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW
May 1, 2005 7:52:16 PM

Usually, no. Dead pixels happen because of inherent defects in the panel -- piece of dust, electrodes placed too close, etc. The LCD panel doesn't really wear out over time though. So once it's been fully warmed up, that's pretty much all the dead pixels you'll ever have. The thing is, "fully warmed up" depends on the user. For example, if your contrast isn't set at 100, then the panel doesn't go through the full range of the voltage levels. So when I test monitors out, I always set the contrast to 100. Then I let it sit for a while at both dark and bright screens (it's actually the dark screens that full voltage is applied; a bright screen is the result of no voltage being applied) before inspecting for dead pixels.

I say "usually" because I haven't heard of any case where dead pixels come up over time, where the panel was fully inspected within a few weeks. After all, those things are hard to notice, so whether or not it seems to have appeared only because the user never noticed it before is debatable. But I suppose it may be possible. For all the LCD monitors I've used, after the first day, no other dead pixels have appeared.

Chuck Hsiao
Amptron
May 2, 2005 2:14:16 AM

thanks...

can i asume the same thing can hold true to LCD TV then?

<b><A HREF="http://www.digitalgunfire.com" target="_new">DigitalGunfire-Industrial EBM</A></b>
ASUS P4S8X-P4 2.4B - 2x512M DDR333 - ATI 9500Pro - WD80G HD(8M) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW
May 2, 2005 2:55:12 AM

Yup.

Actually, I'm not sure how much they differ, or how often it happens with LCD TV's, or the dead pixel policies for them. The thing is, they have roughly the same amount of resolution (about a million pixels), but are larger, so I'd intuitively think that it means they're easier to make. But I don't really know. But the same thing about dead pixels not appearing over time should hold true because they're the same technology, regardless of size.

Chuck Hsiao
Amptron
May 2, 2005 1:35:57 PM

If you're asking what percentage of monitors will develop dead pixels, then, theoretically, all monitors (given enough time) will eventually develop dead pixels. Personally, if your concerned about quality, pick an LCD from a reputable brand. For all displays, I recommend sony. Even if you develop dead pixels, don't worry... in 10 yrs LCD technology will be obsolete anyway!
!