• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
ZEISS DTI thermal imaging cameras. For more discoveries at night, and during the day.

Problem with pixels - CP 4500 (1 Viewer)

HelenB

Opus Editor and Expat from Cumbria
Opus Editor
A couple of weeks ago I noticed a green dot in the sky on a digiscoped image I'd taken of the moon (in the gallery under landscapes). On enlarging it, it appears to be a series of pixels - white in the middle, surrounded by dark pixels which are green below. So I checked back through older images and found that it is occurs as far back as Nov 02, but not so noticeable then. Haven't gone back farther as everything is saved on CDs.

See the 2 images below - the one from the moon shot taken on 16th June is on the left. I've had the 4500 since July 02, so thought I'd ask if anyone else has had this problem and will it be covered under warranty, as my year is almost up? Would appreciate any hints on what the problem is and whether is it fixable. It occurs in the upper right quadrant of the image, quite close to the centre so it is almost always visible after cropping the image.

Thanks in advance.
 

Attachments

  • greendots.jpg
    greendots.jpg
    45.7 KB · Views: 481
Hi Helen,
It's a dead pixel in the ccd of the camera... I've had one since I purchased my cp4500, last June. They tend to look far worse after you have sharpened the image.
I think Nikon has a duty to repair them, but few people bother with a solitary dead pixel (possibly a block of three)..... it's simple to clone/rubber-stamp the problem away with just a click in Photoshop or similar.
It's an annoying problem, that seems more common in the latest digital cameras.
Andy
 
That's interesting Andy, but if Nikon know about it, I wonder why they haven't fixed it yet?

But can I ask Helen, at what settings does this appear, ie, size, quality and so? Does it show on all shots or just those with a darker background? Having just bought a 4500 I wouldn't be too happy to know that there is a problem that Nikon know about and haven't sorted yet.

Ron
 
Ron

This is not a Nikon specific problem, but applies to all digital cameras. The CCD chips for all cameras come from a handful of chip manufacturers. With the number of pixels on each chip its actually surprising that there are not more dead ones.

I'm not sure what the position on cameras is but in the case of LCD PC monitors, a single dead pixel is well within manufacturing tolerances and will not entitle you to a swap out. I think you can have up to 10 faulty pixels with no more that 2 next to each other before the monitor will be classed as faulty. Luckily mine has none.

Paul
 
I hesitate to speak too firmly about cameras as my background is on the computer side, but I think I'm safe in making several points:

First, dead pixels are not a Nikon-only thing, nor (so far as I know) is there anything in particular about the CP4500 that makes it more prone to the problem.

Second, all CCD and CCD-like devices suffer from the same problem. A CCD, or a TFT screen, or anything else that is composed essentially of a very large series of identical micro-components (such as a RAM chip or a CPU) is prone to having a certain number of random flaws per unit of area. It is, among other things, the average number of flaws per unit of area that determines the maximum practical size of the commercial product.

As the manufacturers come up with ways to refine the production process (such as using ever-cleaner clean rooms and ever more highly-refined 99.999% pure raw materials), it becomes practical to make larger and larger arrays. This is why we now have 17 inch TFT screens and 4 to 5 MP CCDs available, and this is why we can buy cheap 512MB Compact Flash cards, and process the pictures on them using super-fast Athlon XP and Pentium 4 CPUs with 40 to 60 million transistors each.

The bottom line, from a manufacturing point of view, is that your cost per unit of area is more-or-less fixed. It doesn't much matter what you make on your 8 inch silicon wafer production line (RAM or CPUs or bits for mobile phones), the cost per wafer is about the same. If, on the average wafer, you have (say) 70 random defects, then you can't afford to be making physically large chips, as you will have to throw a great many of them away. (In general, remember, that a single flaw in a digital circuit makes the entire item unusable.) So you have to make a larger number of smaller chips.

As your technology improves, you get the defect count down, and this in turn makes it economic to produce larger chips - perhaps to switch from 1.3MP CCDs to 3.0MP ones, and so on.

As I said, in most products of this general nature, a single flaw ruins the entire chip. (CPUs and RAM chips are good examples: even a single bad bit makes them unusable.) But in some (TFT screens, smart card RAM, and CCD sensors, a certain, human-defined number of bad bits is tolerable.

Most people buying a cheap TFT screen will accept that they have one or two dead pixels. (That is, of course, why they are cheap - the big TFT wafer manufacturers test each one and sell the 100% good screens to the quality makers at a high price, and shift out the slighty flawed ones to the no-name makers at whatever price they can get. The badly flawed ones they simply have to throw away.

It is reasonable to assume that the same deal applies to the CCD market. Your top-class camera manufacturers pay top dollar for 100% good CCDs. The not-quite-right ones go into cheap cameras.

Of course, they can only test for flaws that are already there. There is no certain way to predict the development of flaws later in the life of a product, except insofar as a batch that has produced few immediate flaws can also be assumed to be likely to produce a smaller number of grown defects.

Finally, this brings us back to your Nikon. There would seem to be two main possibilities:

(a) You just got unlucky. Nikon bought the best available CCDs but you drew the short straw and got the one in 100,000 that develops a post-inspection defect.

(b) Nikon cut things a little too fine, squeezed a supplier a little too far on price, and wound up with a batch of CCDs that are not as good as you have the right to expect. (Taking into account that we are talking about Nikon here, and that you and I pay Nikon's high prices precisely because we expect a top-quality product.)

Either way, an obvious dead pixel is not something that we expect in a Nikon product, and if the company is half as good as their PR would make out, then they will immediately replace your CCD under warranty. This will cost them a heap of money (they might even just give you a whole new camera because that works out to be cheaper for them) but that's why they are Nikon: because they go the extra mile.

In short, I'd be sending the camera in for warranty repair. I expect that Nikon will fix it for you, no questions asked. They have a great reputation (and the pricing structure to match), so let's see if they live up to it.

I'm betting they will.
 
Ahh, I see that while I was typing that mini-essay above, Paul has said what I took 700 words to say in about three sentences. ;(

One thing I would add to Paul's point is that dead pixel policy is not fixed, it varies a great deal between manufacturers and models. On a $AU 450 15 inch TFT screen, you can expect them to tolerate quite a few dead pixels. But on an $AU 650 15 inch screen from a top name maker, then you have every right to expect a zero or near-zero dead pixel policy. You get what you pay for.

Please be sure to let us all know how you go, as this sort of issue is the litmus test of a quality maker. As I said, I'm betting that Nikon will look after you no problems. But if by any chance they don't, I want to know about it so as to be sure to take my buying dollars to another company next time.
 
Tannin

I think you are correct there, but its difficult to find anything in writing on the web to what individual makers policy is in this area, with one exception (that I could find) CTX's S530 which comes with a 100 day zero faulty pixel guarantee
 
Last edited:
There is a reason for that too, Paul - or at least, in my cynical old 20+ years-in-the-industry mind, I think there is a reason.

Replacing screens (or CCDs, whatever) with dead pixels costs a lot of money. So if you advertise your zero dead pixel policy far and wide, you get a good reputation but you spend a heap on replacing things under warranty. Not good.

On the other hand, if you don't have a replacement policy, you wind up with customers saying bad things about you and lose sales. Not good.

So what you do is you have a replacement policy, but you don't tell anyone about it. If a customer compains about a dead pixel, you replace the unit. That way the customer doesn't tell all his friends how bad your company is. And if a customer doesn't complain about it, you do nothing and you save anything up to $1000 or so on your warranty service costs. If, later on, that customer says "hey, I had a bad pixel" you can just say: "Well, we would have fixed it while it was under warranty, of course, but you didn't let us know about it, and it's out of warranty now. Sorry." So the customer winds up feeling that it was his own silly fault and doesn't get too upset about it, and your good name is protected.

And, if you are down in the sales department, and someone walks in saying "the shop up the road has a similar product only it's Brand X and they advertise a zero-defect policy" you can say - "oh yes, we have that too, and we are $50 cheaper".

And you are $50 cheaper, because most of your customers don't know about the dead pixel policy, and that means you only have to fix, say, one in every four faulty ones, so your warranty cost base is lower than the guys making Brand X.

Last of all, you do a little fine tuning. You never actually refuse to fix a dead pixel problem, but you drag your feet a little. You keep quiet about the policy, you don't encourage your retailers to take advantage of it, you look doubtful and hope that the customer decides it isn't really worth bothering about. (And let's face it, there are a hell of a lot of people out there who couldn't tell a dead pixel from a dead parrot.) Only if the customer seems to know what he or she is on about, and to be pretty sure that this is something that ought to be fixed under warranty do you actually go ahead and replace the unit.

The net result is that you have a prodct that is every bit as saleable as the Brand X product, none of your customers are more than just slightly anoyed with you, and you only have to replace a small number of cameras (or TFT screens, or whatever).

Seeing as this is the obvious way to maximise your sales and minimise your costs, almost everybody does it.

Except, of course, for the hordes of no-name merchants who only sell on price in the first place - for they are well aware that their customers buy on price and price alone, and, not being very bright in the first place, the "best-price, best-price" buyer is the easiest one of the lot to make large slabs of money from. You won't have any reputation left, but that doesn't matter - there are always plenty more price-buyers to flock to your shoddy in-store specials like moths to a flame, and they never seem to learn from their mistakes.

And also for the hardy few that advertise their zero dead pixel policy far and wide, then grin and bear the cost. These are usually former no-name companies trying hard to move on up into the quality market. I imagine that it's as good a strategy as any.

(Excuse rant - I plead guilty but insane and in mitigation profer my half a lifetime in the computer trade, which is enough to drive anyone around the bend!)
 
Ron - check out the moon photo in the gallery here:

http://www.birdforum.net/pp_gallery/showphoto.php?photo=6294&password=&sort=1&cat=500&page=1

You'll see that dead pixel as a green spot on the right. For the example shot I posted here, I zoomed in on the original at 100%and cropped it at 250 pixels width. If you then check my last competition shot, you'll see that pixel just to the right of the martin's beak - not quite so noticeable because the DOF background is green:
http://www.birdforum.net/pp_gallery/showphoto.php?photo=6297&password=&sort=1&cat=500&page=1

Andy - I have to admit I'm still learning everything that I can do with Photoshop - will check out if I can do the clone/rubberstamp with the Elements version.

I'll call Nikon CS today and let you all know what they say. Thanks for the prompt responses.
 
Hi,
This is a common problem. I have a CP4500 and ran a test I got from an astro forum. I found I have two pixels that cause this problem.

If you want to try the test. Switch your camera to maunal mode and then take a series of photos, with the lens cap on, at 1/60 and slower. Saving in tif format. Using photoshop you can look at the results and decide how bad the problem is.

Apparently Nikon tend to average pixels with the adjacent ones. I presume this is to even out pixel brightness.

My conclusion was that I know where they are, and in most photographs they are not a problem.

Regards,
david
 
I found my hot/dead pixel within a day of purchase..... I had no inclination to take the camera back, despite the formality of an exchange for another one.
Didn't one camera company offer a dead pixel remapping programme/firmware?
 
I cannot believe I actually have something to contribute on this topic but you can apparantly remap your camera. If it is still under warranty, I wouldn't try this and if it isn't bothering the hell out of you I wouldn't try it either. But... here is a link that explains all about remapping the camera. Use at your own caution!! It apparantly has been tested on the 4500.

http://e2500.narod.ru/ccd_defect_e.htm
 
Oh thats scary, dont think I would like to try that, especially as that document seems to have been translated from Russian (some of the words are still in Russian).
 
Well done K.C. for finding that snippet of info.
Yes, it's risky! That's why camera manufacturers are reluctant to give this power to Joe Public..... In fact they don't even trust us to do firmware upgrades ourself.
 
DavidL,
I tried your test and I found a total of 3 dead pixels. A pink one in the bottom right hand corner, which has been there since I got the camera. And another greenish one towards the bottom left which is not very noticeable, but both this and the one I first wrote about were not there when I checked the first 10 images that I had taken last year.

Does this mean, I wonder, that more dead pixels will develop in time? Haven't had a chance to phone Nikon yet, but will let you all know what they say, asap.
 
HelenB,
Just did the test again following your comment. I found one more making a total of 3. The first test was done mid september 2002. I understand for long exposures in astro the ccd's are cooled to stop hot pixel problems. I will be interesting to see if they reduce in the winter. (No I'm not putting my cp4500 in the fridge or freezer to test this).
Regards,
DavidL
 
raw (diagnostic mode) cpix

KCFoggin said:
I cannot believe I actually have something to contribute on this topic but you can apparantly remap your camera. If it is still under warranty, I wouldn't try this and if it isn't bothering the hell out of you I wouldn't try it either. But... here is a link that explains all about remapping the camera. Use at your own caution!! It apparantly has been tested on the 4500.

http://e2500.narod.ru/ccd_defect_e.htm

By sheer coincidence I stumbled across a reference to this site concerning a raw diagnostic mode for cpix cameras. Now this has fantastic potential for control over the ccd data. However.... when i visited the narod site it i found it daunting, a scary proposition. I am cautious soul and whould therefore like to know if anybody has enabled this raw diagnostic mode and did the camera survive?
 
I finally called Nikon and they said to send it in for service while its still under warranty. So its gone. I printed out some examples of the problem and included them with my letter of explanation. I'll let you all know what happens........

Thank goodness I still have the Oly C700 for non-digiscoping, or I'd be feeling withdrawal symptoms!
 
Nikon have received my CP 4500 and are repairing it under warranty, so I hope I did the right thing by sending it in.

I did try the clone tool in Photoshop, as suggested by Andy, and I was surprised at how easy it was to get rid of the green dot.
 
Warning! This thread is more than 21 years ago old.
It's likely that no further discussion is required, in which case we recommend starting a new thread. If however you feel your response is required you can still do so.

Users who are viewing this thread

Back
Top