where there’s smoke there’s forum fire
psk31 said:
Well if this is true then I would be more inclined to call it a work around rather than a fix. A fix implies a resolution to the root cause of the problem.
My fix was to return the D7000 and move on. Worked for me.
From what Ive googled.. all the DSLRs with video have this, in the bigger badder canons, and the solution to correct is the same for all. 2 million video pixels, the eye wont notice.
OnTheBandwagon said:
From what Ive googled.. all the DSLRs with video have this, in the bigger badder canons, and the solution to correct is the same for all. 2 million video pixels, the eye wont notice.
Hi,
The issue in the D7K is something different than that.
These pixels are not 'typical hot pixels'. They are show up on a waveform scope as out of range for NTSC TV.
My best,
Mike
Does anyone here have a D7K serial # similar to mine -- 3028xxx?
The reason I'm asking is because the salesrep at Best Buy from whom I got the camera told me that, "these new batches (serial #s 3028xxx to 29xxx) are GUARANTEED by Nikon to be free from any dead/stuck/hot pixels." I need somebody to confirm this, and also, the boxes now have the BB "14-Day Return/This Item Requires A 15% Restocking Fee" seals on them. I feel this is BB's tactic to prevent future D7K buyers from exchanging their purchase for another one. Please I need your comments because I don't trust BB nor Nikon as they are only after sales.
Hi,
The seal has always been on the camera's box.
The stocking fee that _I'm aware of_ has always been there for any return of merchandise, but not _defective_ merchandise.
That's a different kettle of fish. You wouldn't pay restock a broken flashlight, would you?
As far as fixing the problem, if can look at the camera, you can check it out in the store. It's quick and easy.
Ask them to charge the battery for an hour or so, put it in the camera, and put in an SD card. Go to LiveView at 100ISO, shutter 1/60 with the lens cap on. Push the magnifier button several times to max LiveView (the "Red frame" will fill the screen) and go left-right, up-down, until you have examined the real estate of the viewfinder (a small representation will show at the bottom right). What's there is there. It should be black, if not it will show bad pixels.
To be clear, you could put the camera on a tripod, set it at ISO 100, shutter 1/60, point to some black velvet that has no light falling on it and do the same test.
Black is black. The lens cap trick is just easier to setup.
The magnifier button, "Qual" (lower left) magnifies the view in LifeView, just like the D90, and will help to spot any defective pixels easily.
If this is unclear, let me know and I'll try to explain again. The same tools are on the D90.
My best,
Mike
Mike Gunter said:
The stocking fee that _I'm aware of_ has always been there for any return of merchandise, but not _defective_ merchandise.
Which is why Nikon appears to be Janus faced.
By making their official position "Said hot pixels are not a defect" they protect their retail partners who now have an document to cite if they don't want the expense of taking returns.
By making their unofficial position "We're working on it and will address this non-issue with a firmware update" they quell the rancor of most of their customers.
Hi Testing...
No joke.
I'm loyal to the brand for thousands of reasons. And I want to shoot video with Nikon lenses. However, it's hard to feel the love...
Since posting a link is not something one should do here, Google "Ask Olympus: What is Pixel Mapping?" It's a good read and clears up a lot of junk talk about hot, dead, and stuck pixels - none of which is what is going on in the video of the D7K.
What it is, I don't know, but Nikon hasn't helped by straddling the fence.
The only thing that makes sense is that the company legal team wouldn't want them to open a door they couldn't close. Unfortunately for them, in doing what they have done, they have hinted that they can't fix the problem with firmware. Even their Janus-esque post suggests that. It inspires no confidence.
A lot has been made of the lens cap trick to check for the defective pixels, which is why I suggested another, albeit longer and more difficult, method. Just saying, black is black. What else should it be? Red, green, blue, white? Anything in a frame that a filmmaker decides to shoot should be the color he or she decides to make it. Any creative professional should accept no less.
As for the Best Buy sales rep, I truly doubt he know more than the sales rep at the Gap. I know I sound like a cynical old man, but I suspect that real news travels _slowly_ in that milieu, gossip and misinformation, however, is at the speed of light.
I'm waiting out the D7K to see what happens. I might get a D700 for my still work and get a Panasonic AG-101 for filming. It much more expensive route, but the Panasonic has more features - some I don't need since I would be doing other things - audio externally for example, but it depends on the length of the wait. I already have video equipment for the video side of my business, but I would like a larger sensor and would like to use my Nikon lenses. It just isn't happening now with the D7K.
My best,
Mike
Baraclava, mine's not from bestbuy but my SN is 3042### and has one stuck blue pixel in low light video.
Mike, you can post links here. The mods will delete links posted by new members to commercial sites, but once you've been here a while they're not as strict—at least they don't delete mine. Anyway, here's what Mike was referring to (thanks for the link, Mike):
http://olympusamerica.com/crm/oneoffpages/ask_oly/crm_e_ask_oly_03_09.asp
They are stuck pixels, though. The reason they show up as a certain color is that each stuck photosite has its own respective color filter over it as part of the Bayer pattern filter. White (full charge on the photodiode) is not white on the final output for each pixel because the camera/raw converter interpolates the values from the working pixels of the other colors nearby.
Hi Johnny,
I wasn't sure about posting to site and wanted to be sure. Olympus is a commercial site, but the information is useful.
Clearly you know more about diodes and photosites and the like than I do. I was unaware that there was a RAW converter at play during the video process. Of course, for me, it's all magic.
I do know that after several years of producing a TV series for PBS on consumer grade cameras, I had hoped to upgrade the sensor to an APS-size and use Nikon lenses as many of my colleagues in other forums (fora?) have also voiced.
We can only hope that Nikon sorts it out soon, and that the firmware update takes care of it.
My best,
Mike
Anytime you have information from all three color channels (red, green, blue) in an image from a bayer-pattern sensor (which means basically everything except sigma right now), it means there has been a demosaic algorithm run to get that info, including for video. That just means that because each photosite only records either red, green, or blue light (but never all three), you guess at what the other two channels would be based on neighboring photosites that were measuring light of those colors.
All the data coming straight off the sensor is raw and not interpolated. Cheap point-and-shoots might have the sensor and demosaic hardware on the same module, but when you save a raw file you're putting off the final processing until you get it on your computer. Of course, the camera still does its work, anyway, so that you have an image on the camera LCD and a thumbnail image, but the sensor data gets stored as captured, which is why raw is better for adjusting white balance and/or exposure after the fact.
I'm really sorry the D7000 hasn't worked out for you, Mike. Maybe the upcoming firmware fix will be able to get it back to within spec for your work. Here's hoping...
FNJ2C said:
I'm with mike nikon needs to get a fix now or recall what is out there.what is takeing so long.And thanks mike for your posts.
Hi FNJ2C,
Let's keep our optimism up (grandpa was down on my meds yesterday / :-0 /), and hope that Nikon gets its collective head out... ah, in the right place.
Truly a week for me to wish my best to all, Seasons' Greetings!
Mike
So folks, has this been fixed with the update? There's 19 pages here of comments on the pixel problem. And less than one page of comments on the Firmware update thread. With as many people complaining about this issue, I'd think there'd be more talk about whether or not this is a real fix. What's the word? I'm looking to invest in a DSLR and the D7000 is the top of my list except for this issue.
Hi Bill,
Welcome to the forum. From what I saw more people seemed to make comments on the homesite (nikonrumors.com) when the firmware fix was announced than in this forum thread. As far as I could tell most if not all comments were favorable although I didn't read everyone's comment. I did notice that several of us here commented that the fix was a success.
Some people had to run the firmware update more than once for it to "catch". As for me, I had one stuck blue pixel that showed on video when recording in dark areas and using ISOs greater than 200 and High ISO Noise reduction ON. The firmware fixed that although most of my recording is in good light anyway. Whether it's a true fix vs. a bandaid; I don't know and sorta don't really care. I'm happy with the result. Others may chime in here with further comments.
Welcome aboard.
Bill Tuttle said:
So folks, has this been fixed with the update? There's 19 pages here of comments on the pixel problem. And less than one page of comments on the Firmware update thread. With as many people complaining about this issue, I'd think there'd be more talk about whether or not this is a real fix. What's the word? I'm looking to invest in a DSLR and the D7000 is the top of my list except for this issue.
I agree. The firmware update fixed my problem in one-pass. Since the upgraded firmware, the stuck-pixel problem is no longer an issue with the D7000.
Thanks guys. If the results work, it's a fix in my book. :) I appreciate the responses. Unless I find any other showstoppers, I think this puts it solidly at the top of the list again.
---Bill
Forum folk,
I'm not a Nikon noob, and yet I'm having a noob problem. I just upgraded tot eh D7K, and though I've installed Nikon's questionable native software, my camera won't connect to my computer (macbook.) Possibly connected is the fact that my batter won't read as fully charged even though I've charged it fully. Anybody have similar problems with a new D7000? Is there a very obvious step I am missing somewhere?
Hi all,
Here's an interesting dilemma, not sure if it's the same sensor issue but it's truly horrible. I upgraded D90 to D7000 in December, loved the D7000's clean output even at 3200. But last week I took some night shots in Singapore with 10-30s exposures at ISO 100. When I looked at the first shot on the camera, I thought, "weird, the sky is completely hazy, but the camera is picking up faint stars!?" The noisy pixels were that bad. Even in the (black) water and on moderately lit buildings, I could see the noisy pixels, which stayed in the same spot even when the scene was recomposed.
There are a ton of "warm" pixels... ranging from barely visible to a gray spot. Longer exposure makes it even worse. I experimented with lens cap on later, no change, some are visible as colored spots (nearly fully stuck on?) even at low ISO. Firmware upgraded the next day; no change.
Anyone have bad experiences with long exposures? Could someone who has the "hot pixels video mode" problem try this out, 10-30s exposure with lens cap on / varying ISO? 30s with ISO 800, lens cap on looks like a checkerboard of colored dots even at low or no magnification!! SN in 3046xxx range, bought from BB.
Debating what to do. Unfortunately I bought my camera over Xmas in the US, but I live in Germany, so probably no warranty service available here :-/
-Dave
Hey, Dave. Welcome to the forum. Do you have long exposure noise reduction on or off? Also, what do you use to post process? I can try some test shots for you, but I'd like to have the same settings you did.
Hi Jonnyapple, I tried it without the long-exposure NR, in "Normal" .jpg mode, when doing the lens-cap-on tests (though results were similar in RAW mode). Standard camera settings, no post-pro. So, try 30s exposure, ISO range like 100/400/1600/6400, no NR, lens cap (or even body cap) on.
Afterwards I tried it *with* the long-exposure NR, which does nearly eliminate the effect. BUT there are a few problems with that: 1) we're not just talking about noise, but hundreds of hot pixels (and with low ISO at that); 2) I see the problem (faintly) starting at 1/4s exposure, very bad by a couple seconds, and long-exposure NR only starts working at 10s+; 3) who wants to wait double the exposure time before starting the next shot.
Just tested my D90 (also without NR), it has about 2 pixels with slight problems at high ISO; at low ISO it's black as a clean chalkboard. Whereas the D7000 has hundreds bad even at low ISO.
I'll probably call US Nikon Support today and discuss with them, let's see how that goes. Curious to hear your results - thanks!
Dave
Here's the series you asked for, Dave. I have to admit it's the first time I've shot my lens cap. I'm not going to make a habit of it. ;-)
Settings: 2s shutter speed (and f/11 for anyone who's counting...), jpg normal large straight out of camera, standard picture control, auto WB, D-lighting off (that may be something to check), ISO NR off, long exposure NR off, sRGB (can't see how that would make much difference).
I'm really impressed by this performance—I expected more noise/hot pixels than I see. I saw no hot pixels at ISO 1600 but after seeing a big one at ISO 6400 center-right (a group of two red ones), I went back and it is visible. Amp noise is starting to be visible in the lower left corner of the frame at ISO 1600, and there are (expectedly) some hot pixels at ISO 6400.
ISO 100
ISO 400
ISO 1600
ISO 6400
edit: fixed broken links
Can you show images of what you're seeing at low ISO?
Thanks, Jonny... looks like you have less problems, though at only a 2s exposure it's harder to see.
Check out these two, if I get the links right:
30s, ISO 1600
20s, skyline view, ISO 100 (!)
Zoom 100% on the water (brigher parts) in the skyline pic, you will see "stars..."
Waiting now for Nikon's response on those images.
Dave
A side note, I noticed some red fringing with long exposures / high ISO and my 18-200mm lens. But with body cap, instead of a lens, that disappeared. I guess it was slight light leakage around the lens cap or elsewhere in the lens body.
I see what you're talking about, Dave. It would be interesting to know whether dark frame subtraction (long exposure NR) would take out what you're seeing. You won't be able to effectively remove those bright spots easily from jpegs in post. Did you try reupdating the firmware after using your sensor for a while? Put it in liveview for 5 minutes or so and then rerun the firmware update. My first time updating didn't get rid of the two stuck pixels I had in video but the second time did.
It seems to me that your options are to either use long exp NR or post-process with a raw converter that has a demosaic routine that will watch out for this issue and automatically remove them (I use Lightroom 3 and it does—at least here in the US it has a 30-day trial).
See my 2nd post about the long exposure NR, it works, but only at >10s exposures and it's a pain. I did only update the firmware once.
Well, of course my other option is repair ;-). Actually I did get a response from Nikon US that I need to send the camera in for warranty service. Now I need to determine if I can send it here in Germany, or only back to the States.
Anyway I think it's good to know that if the sensor is *this* bad, Nikon will do a repair! I am curious to see in the next weeks if others on the forum have a similar long-exposure issue. Good to see that your results are better than mine, so it's not all D7000s that are affected, only some.
one51:
The Nikon USA support page includes this statement: "If these spots are seen on images photographed under normal conditions (bright light with exposure times shorter than 1/4 second) then the camera may need to be sent in to a Nikon Service Center for evaluation." If you can produce bad pixels at 1/4-second, then Nikon is on record for "fixing" it. I've actually seen worse from a D7000 at another forum (his photos looked like pictures of a snowstorm). I haven't tested my D7000 at exposures longer than 0.8 seconds, but up until then it seemed fine (I have already upgraded my firmware). Yours appears far worse.
My advice: demand a sensor replacement from Nikon. According to others here, they've been known to replace sensors for less.
You must log in to post.
