I just got my D-80 yesterday and did some test for the hot pixels, OMG I counted more than 20, and now on the way for the exchange. Hopefully will be better, I got some dust also on the sensors, is this the sign that not enough quality control on Nikon sides??
>Hi Alex, and welcome to Nikonians! > >Bad luck about the hot pixels - hopefully it will get sorted >out successfully with an exchange. > >Dust is a fact of digital life, though it can be minimised >with careful handling - it's hardly an indication of QC >problems
Brian thanks for your reply, I just get the exchange but I am so upset with this store, they said that after this if the problems still exist I have to bring it back to Nikon Canada directly, they don't even know how to check hot pixels, I think I am not gonna buy any camera from this store anymore, so painfull, I don't even used it, they also said that this is a common problems with digital camera, but my D-50 has no hot pixels at all. Oh well maybe this is my good luck )
>Forgive the simple question, but how do you check for hot >pixels? Visually or some other way? I copied it from other forums: Hot pixel test:
1. Set camera to manual mode. 2. Set lens autofocus switch to manual mode. 3. Set ISO to 1600. 4. Set shutter to 5 sec. 5. Set apperture to whatever you want. 6. Place lens cap on lens. 7. Take a shot. 8. Display image at 100%. You can see more hot pixels on 30 sec. Too bad I cannot attach the picture here.
Keep in mind though that much like pointing a D200 at a lightbulb, the lens cap test isn't necessarily a test of camera performance in real use conditions. Pixels that show up in the lens cap test don't always show up in actual images. I am not saying you're wrong - in fact I ran the test on my D200 when I first got it. However, people also do need to consider whether the pixels are visible when shooting real images. Just a suggestion....
>Keep in mind though that much like pointing a D200 at a >lightbulb, the lens cap test isn't necessarily a test of >camera performance in real use conditions. Pixels that show >up in the lens cap test don't always show up in actual >images. I am not saying you're wrong - in fact I ran the >test on my D200 when I first got it. However, people also >do need to consider whether the pixels are visible when >shooting real images. Just a suggestion.... > >Stephen Stephen, did you try this method on your D200??? just curious if you get any hot pixels, but how come my other camera don't have this problems at all (D50 & D70s), is this because the 10MP?? Hope you don't mind to answer this qusetions. Thanks in advance.
Using the lens test, my D200 has four or five visible pixels. One of these is also visible at 100% magnification in Capture in High ISO shots, but the others are not visible under any shooting conditions (including star photography). It takes less than 10 seconds to eliminate it in CS2 so I've never bothered with it.
The D200 is known for stuck and/or hot pixels so I would bet the D80 will have the same issue (of sorts). Some are worse than others, and many have none at all. Personally though I don't think there is anything wrong with the sensor. My impression is that Nikon has simply not mastered mapping them all out on the 10 MP sensor. Visible or not, all photo sensors do have them - they are just mapped out before you get the camera. As I mentioned in another thread awhile back, my D70 had no visible hot pixels with the original firmware or the version 2 firmware; however, it did have a couple with the intermediate firmware release. BTW, one thing I have noticed is that if I use Capture, NX, Nikon View, or Picture Project, to view a high ISO shot, the one pixel is visible, but if I use CS2/ACR, Lightroom, Rawshooter, or DX0, it is never visible (those raw converters interpolate it out as bad data). Since I use CS2 and ACR most of the time, I didn't even know the one pixel was stuck until I tried Capture.
I don't know whether any of this helps, and I am certainly not suggesting that you keep a camera if the pixels are a real world problem. BTW, there are a few very long and very informative threads on hot/stuck pixels in the D200 forum. They may be archived now but should be easy to find. Good luck and have fun with what otherwise looks like a very nice camera (too small for my hands though). I hope my (limited) experience is of some use....
>Hello: > >Using the lens test, my D200 has four or five visible >pixels. One of these is also visible at 100% magnification >in Capture in High ISO shots, but the others are not visible >under any shooting conditions (including star photography). >It takes less than 10 seconds to eliminate it in CS2 so I've >never bothered with it. > >The D200 is known for stuck and/or hot pixels so I would bet >the D80 will have the same issue (of sorts). Some are worse >than others, and many have none at all. Personally though I >don't think there is anything wrong with the sensor. My >impression is that Nikon has simply not mastered mapping >them all out on the 10 MP sensor. Visible or not, all photo >sensors do have them - they are just mapped out before you >get the camera. As I mentioned in another thread awhile >back, my D70 had no visible hot pixels with the original >firmware or the version 2 firmware; however, it did have a >couple with the intermediate firmware release. BTW, one >thing I have noticed is that if I use Capture, NX, Nikon >View, or Picture Project, to view a high ISO shot, the one >pixel is visible, but if I use CS2/ACR, Lightroom, >Rawshooter, or DX0, it is never visible (those raw >converters interpolate it out as bad data). Since I use CS2 >and ACR most of the time, I didn't even know the one pixel >was stuck until I tried Capture. > >I don't know whether any of this helps, and I am certainly >not suggesting that you keep a camera if the pixels are a >real world problem. BTW, there are a few very long and very >informative threads on hot/stuck pixels in the D200 forum. >They may be archived now but should be easy to find. Good >luck and have fun with what otherwise looks like a very nice >camera (too small for my hands though). I hope my (limited) >experience is of some use.... > >Stephen Thanks for sharing your experience Stephen, I will try take some pictures today and will check the results.
I was also victimized by a hot pixel on my D80. It was showing on images at ISO 100 and up. Fortunately, I purchased from Ritz.com, with their liberal return policy. I was also able to order a second D80 body from Ritz and have been using it for a few days with no hot pixels or other problems. Btw, this is my first digital SLR and I have high hopes for this camera, which I suspect will be satisfied. I'm a former N90s user.
My first D80 had HORRIBLE hot pixels. There were more than 70 of them at ISO 1600 1/2sec, and most of them were in 6 big clusters. They were easily visible in full frame images at normal shooting ISOs and speeds.
I called Ritz about it, sent the camera back, and had a replacement in hand the NEXT afternoon. Way to go Ritz.
I can still find a few "warm" pixels in my new D80, but they don't show up in normal shooting.
By the way, I was surprised to find that 1/2 sec exposures revealed the most hot pixels in both of my examples. Counterintuitively they disappear with longer exposures; I don't know why.
I intend to use the camera for a while, test again, and if things are worse I'll ship it back to Nikon for re-mapping if necessary.
Dang it. I just checked and I have a cluster of about 18 in one location. I shot some night images two days ago and thought the red dot in the sky was a chopper or airplane. Nope, it was this cluster of hot pixels.
I, as you did, spotted my hot pixel cluster in an ordinary photograph and THEN ran some tests. I'm not a pixel peeper, but this hot pixel thing, especially the ones in large clusters, is apparently not an uncommon problem.
Out of curiosity I tested my old D70 for the first time and couldn't find any hot pixels until I got to ISO 1600 at 30 sec, a place where I never shoot. Even then there were only two and they were what I would call "warm" pixels.
It will be interesting to see what Nikon does about this. My replacement D80 shows a few hot pixels, but they aren't in clusters and so far I haven't had them jump out at me in any of my normal shooting. However, they're in there and I reckon a re-mapping might be in order one of these days.
Upon receiving my second body, I immediately shot the same series of tests I used on the first body. The comparison was favorable to my second body. Hot pixels did not show up until ISO 400 at 30 seconds, but not nearly as many as the first body at this same setting. ISO 1600 at 30 seconds very faintly showed 4 pairs of warm/hot pixels, each pair in very close proximity to each other. The first body looked crazy at this setting. However, I doubt I'll ever shoot anything at these extreme settings so I'm pretty satisfied with my replacement.
Bottom line..I am not seeing any hot pixels in my normal, day to day images. I've yet to do any nighttime, long exposures. We'll see how that works out.
RETURNED MY D-80 AND GET ANOTHER D-70S BECAUSE I SOLD IT BEFORE TO GET THE D-80 BUT GET UPSET WITH PIXELS PROBLEMS THAT SHOWED UP ON THE REAL PICTURES ON ANY SETTINGS OF ISO OR SHUTTER SPEED, IT SHOWED AS A RED DOT, IN FACT MY D-50 AND THE NEW D-70S HAS NO HOT PIXELS AT ALL, ALL CLEAN 30 SEC OR 5 SEC TESTED ISO 1600, BUT MAKE SURE TO TURN ON THE LONG EXPOSURES NR, SO I THINK NIKON NEED MORE QUALITY CONTROL BEFORE THEY PUT THEM ON THE MARKET. I AM USING THIS SOFTWARE TO COUNT THE HOT PIXELS. http://www.starzen.com/imaging/deadpixeltest.htm
I'm sorry to hear about your situation and I hope your D70s works out for you. I am certain the new camera is going to be a great tool from many people, but if you have multiple normal shooting hot/dead pixels (or clusters) that is more than I would be willing to tolerate as well. I do need to insert a disclaimer about the software you mention though. I can run that on different images from my D200 and get anywhere from zero to 90 dead pixels and 1,000,000 to 9,000,000+ hot pixels (never the same twice for either). Obviously those are not exactly accurate results. Dead pixels should always be dead and hotpixels should not register in the millions. Just thought I shoudl pass that along....
FORGOT TO MENTION THAT I ALSO GOT 1 DEAD PIXEL ON THE LCD SCREEN, I WROTE THIS THREAD DOES NOT MEAN I WANT TO MAKE THIS ISSUE BIGGER, LIKE OTHER PEOPLE SAID WHY DON'T JUST IGNORE IT AND JUST KEEP SHOOTING, BUT WHY SHOULD I PAY FOR SOMETHING THAT IN FACT HAS 1 RED PIXEL THAT I CAN SEE THROUGH MY EYES, IF I DON'T SEE IT DIFFERENT STORY, BUT I ALREADY TRY MY BEST TO DO SOME TEST WITH DIFFERENT SETTINGS ON THE REAL LIFE PICTURE, AND THE RED DOT STILL THERE BUGGING MY EYES, SOME PEOPLE SAID THAT WE CAN DO SOME TOUCH UP WITH PHOTOSHOP OR OTHER SOFTWARE, BUT I AM NOT GONNA DO THIS THING EVERYTIME I TOOK PICTURES, PP IS OK BUT NOT TO ELIMINATE THIS PIXEL ON EVERY PICTURES.
I final went digital, with the D80. I do mostly macros and find being able to easily crop 50% is fantastic. Problem is when I do I notice some of these hot pixels! I understand this can happen at high ISO and long exposures - But I notice them at ISO 400 between 1/500 upto 1/2 consistantly see one. At ISO 400 at 1/2 photo has 5! Since this is at lower ISO and normal shutter speed, is this considered 'defective' and camera should be replaced?
This Hot Pixel Thread is so ridiculous! It is as if Nikon Owners are obsessed with finding SOMETHING wrong with a Nikon. The Canon forums simply do not do this. Why is that? Canons have Hot Pixels, but you don't see THEM running tests.
I tested my Canon XT and found over 30 Hot Pixels. So what? It is a fact of life, like dust in a lens. Ran the same test on my D50. 21 Hot Pixels. OMG, what do I do Now?? Nothing.
Go take pictures and stop trying to "catch Nikon" at QC issues.
I'm going to have to agree with the previous post. Hot pixels when shooting at iso 100 in daylight is one thing. This business with noticing FOUR hot pixels when shooting at ISO 1600 and THIRTY SECOND EXPOSURES is ridiculous. If you can find reasons to shoot 30 sec ISO 1600 exposures often enough to make this a problem, you can also take a minute in PS CS2 and paint the four pixels you're seeing back to black.
>I'm on my second D80. Both had hot pixels that were visible >at ISO 200 and 1/10 sec. > >Nikon USA says hot pixels are not normal for the D80 and I >should exchange at place of purchase. Local retail store >says every D80 has hot pixels. > >I'm okay with ISO 1600 and shutter speed = several seconds >having hot pixels, but ISO 400 or less with a reasonable >shutter speed should not reveal any sensor flaws. > >Steve
I completely understand. Under those circumstances, it needs to got back. But, 1600, 5-30 secs., lens cap on... is ridiculous.
These people need to go take pictures and stop trying to CATCH a flaw in a Nikon, which exists in all brands.
I wasn't trying to find something wrong with my Nikon. I just want a $1000 camera to work. I don't want to have to add another step to my workflow for every image I capture at ISO 400 or more.
I ran the tests down to 200 ISO and still had a cluster of hot pixels. Then I took some normal images at f4, 1/60 sec at ISO 400 and because I knew where to look I could SEE the red cluster at 100% magnification. I went back and looked at a representative number of images I had taken on my D80 last week and found that I could locate the hot pixels on most of them.
That's enough for me to exchange the body for a new one. I have no beef with Nikon. All I want is a camera that works to spec.
Rest assured they are not all bad. After testing mine with the lens cap technique, I do notice very random hot/warm pixels at 25" and ISO H 1.0. However, I'm sure these will get lost in the mix of an image very easily.
Still loving the camera though....and I kept my D50 for reasons just like you folks are mentioning. If the D80 ever has an issue at 25" and Iso 1600, I'll just pull out the D50 and the problem is solved .
>This Hot Pixel Thread is so ridiculous! It is as if Nikon >Owners are obsessed with finding SOMETHING wrong with a >Nikon. The Canon forums simply do not do this. Why is that? >Canons have Hot Pixels, but you don't see THEM running >tests. > >I tested my Canon XT and found over 30 Hot Pixels. So what? >It is a fact of life, like dust in a lens. Ran the same test >on my D50. 21 Hot Pixels. OMG, what do I do Now?? Nothing. > >Go take pictures and stop trying to "catch Nikon" at QC >issues. > >Take pictures!
More time shooting pictures and less time trying to be lab rats. If you look closely at anything you will find a flaw.
I got to sit in a brand new ferrar once that had been driven about 70 kms.... I got out of the car to look it over and to my surprise, there was a stone chip on the hood....A flaw? on $100,000 vehicle???
Point is...unless your photos are showing up with dots all over the place then forget about it.... every sensor will have hot pixels.
>This Hot Pixel Thread is so ridiculous! It is as if Nikon >Owners are obsessed with finding SOMETHING wrong with a >Nikon. The Canon forums simply do not do this. Why is that? >Canons have Hot Pixels, but you don't see THEM running >tests.
Why run tests? Canons are perfect. Just ask all the fatory-paid independent bloggers.
First let me say that I admit to having only read through this long thread very quickly. I don't mean to offend anyone; I'm sure some of the posters do have problems that should be addressed by Nikon, while others do not.
I know a great deal about this subject.
Lens cap tests are definitely bogus. What makes a photosite bad is it's non-linear response to light and so a test without light is meaningless.
High ISO tests have low Signal to Noise Ratios that make distiniguishing between noise and a bad photosite very difficult. Many people misidentify noise as a "hot pixel".
The practical test is whether low ISO images in ordinary lighting show bad photosites at reasonable enlargement.
I know of no proper (IMO) program or testing methodology for detecting bad photosites that is widely available to the "public".
However, I have a testing method that is based on a particular type of exposure and a statistical analysis and I perform these anaylses as time permits for Nikonians. At the bottom of my web site there is a list of post to analyses that showed bad photosites. (Not all files submitted to me actually have bad photosites.)
If you feel you have a "bad" D80 feel free to contact me by e-mail for further instructions.
Hi, I am new to Nikonians and to digital SLR photography. I have a question.
I have no idea what a hot pixel looks like. I was reading this thread and decided to try the lens cap test on my own new D80. When I looked at the pic, it was of course black, but there were a ton (I am talking thousands) of very tiny "sparkles" sprinkled across it fairly evenly. White, red, blue, and green. When I zoomed 100% as suggested, the tiny sparkles became large (rather pretty)star shapes.
I guess I am worried, whether or not I should exchange the camera.
All my photos taken so far look great to my unprofessional eye. I am just wondering if those are hot pixels, if I should worry, and are they likely to get worse?
You're describing noise, not hot pixels. And as a previous poster noted, the lens cap test is very unrealistic. It all comes down to this. Look at the photographs you normally take. If all appears well, stop worrying about hot pixels because you're A-OK. If a single pixel or cluster is screaming out at you with an inappropriate color (mainly with out zooming in significantly), you may have a hot pixel or group of pixels.
I apologize for misunderstood in regards of hot pixels after visited Bill website and see the samples of bad pictures with hot pixels, but I did have 1 red dot with + sign on it that keep bugging my eyes to not to see it that's why I have decided to return this camera and maybe worth wait for a while for the next batch of D80 perhaps will be better without this hot pixels thing. The reason I started this thread is just curious why on my D50&D70s has no such colourful marks on the "Bogus" test as Bill said, but on my new D80 has lots of them, some people said that they were noise, so it means older camera has no noise?? this made me confused as some people said D80 has less noise, if Bill read this can you please give me more explanation, because everyone said that you are the expert about this matter. Thanks for your time to read.
For what it is worth. Although I never use ISO above 400, my D80 shows no hot pixels with a 90 second exposure at ISO 100. The Canon 5D I recently sold under the same conditions showed about 15-18 hot pixels. (That's a guess as I never actually counted) Not bad considering the D80 is 1/3 of the price...
Granted, I haven't done the lens cap test, and likely will not.
The + sign shape is a classic side effect of a bad photosite; so it's likely you have one there.
The D70/D70s/D50 use the same sensor. The D200 and D80 use very closely related sensors. For whatever reason the D200/D80 sensors show more noise in lens cap tests. They also appear to be more prone to bad photosites. FWIW, I own a D70 and a D200 and have no problems with the noise performance of my D200.
Bill, I thank you for the explanation and yesterday I gave a try for another D80 and this one was perfect no red dots at all and also no dead pixels on the LCD, I feel ashame with the thread I started here. Image quality was perfect. I saw some dust on the sensor but I have tried to use a blower still could not removed, so I try to zoom in and out my 18-200 VR lens and it's gone by itself, hmmm this is strange. My D80 eats lots of battery power, so I have to get a spare battery this time, D50 and D70S battery last much longer.
So far two D80 owners have sent me files. In one case no bad photosites were found. In the other, initally no bad sites were found, but sites appeared at longer exposures. Probably there is a current leak at that site so over time more electrons accummulate than should. In fact, this is the worst site I have ever seen from a Signal to Noise Ratio standpoint. At ISO 100 it was fine at 1/40s but it was terrible at 1/2s(!)
Here is the report on the really bad one (38.6 standard deviations):
Ritz said I could return my D80, because of the hot pixel problem at low ISO and normal shutter speeds - BUT it must arrive at warehouse (this is a quote from customer service) " in pristine condition, so it can be restocked". Knowing that they will just resell it (which may be how I recieved mine, with pixel issues) - I cannot in good conscience return it. So it will go to Nikon repair, assuming they think the problem qualifies for repair under warrantee!
Mine was sent back as a warranty repair. My D80 also had stuck pixels visible at low ISO and normal shutter speeds. Given the very long lead time associated with Nikon Service, I would really like to see a CCD remapping feature installed in-camera or available as a routine within their Nikon Capture program.
While I'm waiting for the D80 to return from Nikon, I'm going back to using my very reliable Nikon film cameras.
Thanks for mentioning your experience. This was my entry into digital, from the N70. I too will go back to film till I get D80 returned from service (have not sent in yet). I found the transtion from N70 to D80 to be very smooth and feel the D80 is an excellent camera!
I, too, have hot pixels on my new D80. 6 to be exact. 4 were clustered together and 2 were scattered. The first thing I saw in every picture on my computer screen was the damn green dot in the upper left hand corner. I took the camera to Wolf (where I purchased it). The manager saw the pixel and agreed to exchange it when a new one arrived--they have evidently been selling like hotcakes.
I also printed the pictures that had the hot pixel cluster. The prints were pristine. The manager still said he would change it out.
I suggest printing some photos to see if they actually show up. I'm still trading mine in because I'm afraid that an enlargement may show the pixel.