PDA

View Full Version : Question for D200 owners.



DarkDTSHD
01-22-2006, 01:20 PM
Recently, another site did a quick hands-on "preview" of the D200. They mentioned one glaring issue some D200 owners were having with their cameras. And that was what this reviewer called the "cordoroy effect'. A series of verticle lines about 2 pixels tall. Only on specific types of shots. Here's the article (hope the moderators don't remove my thread because I mentioned another site):

http://www.imaging-resource.com/PRODS/D200/D200A.HTM

Have any of you noticed this with your D200's? Realizing you just got them...

I ask because I'm looking for my first DSLR and the D200 definitely caught my interest. And, do you guys think this issue is anything to be concerned about? Being as this only happens with certain types of shots. Such as the one mentioned in the article.

Thanks!!

erichlund
01-22-2006, 09:11 PM
Recently, another site did a quick hands-on "preview" of the D200. They mentioned one glaring issue some D200 owners were having with their cameras. And that was what this reviewer called the "cordoroy effect'. A series of verticle lines about 2 pixels tall. Only on specific types of shots. Here's the article (hope the moderators don't remove my thread because I mentioned another site):

http://www.imaging-resource.com/PRODS/D200/D200A.HTM

Have any of you noticed this with your D200's? Realizing you just got them...

I ask because I'm looking for my first DSLR and the D200 definitely caught my interest. And, do you guys think this issue is anything to be concerned about? Being as this only happens with certain types of shots. Such as the one mentioned in the article.

Thanks!!
That's a new one. I've mostly heard banding, and one reviewer is calling it striping. Frankly, no on really knows at this point how widespread the problem is, and even if there is only one problem. Some people have already had it fixed by Nikon. Some are claiming a complete fix, others are claiming it's only a partial fix. The bottom line is there's a lot of mixed information out there. It's not something to panic over. Nikon will fix the problem. It's just going to take a little time to fully sort out what the problem is.

Canon was able to fix a similar issue with a firmware update to the 20D. I suspect that a certain percentage of D200s have an actual hardware card that needs replacement. At least one person has had the problem completely eliminated by this solution. It's possible a firmware update will be required for the less aggressive versions of the problem.

To get the lesser problem, you really need to go out of your way to overexpose a shot that has very wide dynamic range. At least that's what's mostly been reported.

I know this really doesn't clear up the issue. I must say, my camera has not exibited any tendency to show banding, and I tend to believe the majority don't. It's a problem that Nikon does need to address, and several Nikon support centers have reported back to their customers that they are receiving training on the issue. However, I also believe that Nikon will do this in a rational manner, on a case by case basis.

Cheers,
Eric

DarkDTSHD
01-22-2006, 11:16 PM
That's a new one. I've mostly heard banding, and one reviewer is calling it striping. Frankly, no on really knows at this point how widespread the problem is, and even if there is only one problem. Some people have already had it fixed by Nikon. Some are claiming a complete fix, others are claiming it's only a partial fix. The bottom line is there's a lot of mixed information out there. It's not something to panic over. Nikon will fix the problem. It's just going to take a little time to fully sort out what the problem is.

Canon was able to fix a similar issue with a firmware update to the 20D. I suspect that a certain percentage of D200s have an actual hardware card that needs replacement. At least one person has had the problem completely eliminated by this solution. It's possible a firmware update will be required for the less aggressive versions of the problem.

To get the lesser problem, you really need to go out of your way to overexpose a shot that has very wide dynamic range. At least that's what's mostly been reported.

I know this really doesn't clear up the issue. I must say, my camera has not exibited any tendency to show banding, and I tend to believe the majority don't. It's a problem that Nikon does need to address, and several Nikon support centers have reported back to their customers that they are receiving training on the issue. However, I also believe that Nikon will do this in a rational manner, on a case by case basis.

Cheers,
Eric


Hello Eric,

Thank you for taking the time to reply! :)

I guess I'll wait a little longer and see what Image-Resource says on this matter. As you said, and as was mentioned in the article, you do have "...go out of your way to overexpose a shot that has very wide dynamic range". Still, I'm sure all D200 owners would feel better knowing that there is a fix or that a fix is in the works.

If you hear any more news please post.

Take care!

BigConig
01-23-2006, 06:50 AM
Both myself and my father have D200s. My father's was first batch and mine second. Neither of our cameras have the problem.

coldrain
01-23-2006, 07:15 AM
Both myself and my father have D200s. My father's was first batch and mine second. Neither of our cameras have the problem.
How did you and your dad test that?

DarkDTSHD
01-23-2006, 08:32 AM
Both myself and my father have D200s. My father's was first batch and mine second. Neither of our cameras have the problem.

Thanks BigConig! Perhaps the "Cordoroy effect (aka "banding") are isolated insidents. On select units that got overlooked by the quality control people. (shrug)

Hope to hear from more D200 owners and from Image-Resources. As IR is planning to look into this for us. Or at least keep us up to date on this issue.

Have a good week! :)

coldrain
01-23-2006, 08:52 AM
Thanks BigConig! Perhaps the "Cordoroy effect (aka "banding") are isolated insidents. On select units that got overlooked by the quality control people. (shrug)

Hope to hear from more D200 owners and from Image-Resources. As IR is planning to look into this for us. Or at least keep us up to date on this issue.

Have a good week! :)
When you read that article you link to it gets clear it is not a quality control issue, more a design problem/part specification issue. I doubt BigConig has actually tested his dads and his camera, it is not easily tested as you can read on that page. It does not surface frequently, and as such it probably is more accurate to say BigConig has not NOTICED it on any of their photos.

It seems to be not a huge issue, more of an annoyance IF you happen to make a photo where that weird phenomenon shows up. Same as the odd pink that has popped up in different ways on D50 photos. Or the odd maze artifacts that pop up sometimes in D70/D70s photos.

DarkDTSHD
01-23-2006, 04:08 PM
When you read that article you link to it gets clear it is not a quality control issue, more a design problem/part specification issue. I doubt BigConig has actually tested his dads and his camera, it is not easily tested as you can read on that page. It does not surface frequently, and as such it probably is more accurate to say BigConig has not NOTICED it on any of their photos.

It seems to be not a huge issue, more of an annoyance IF you happen to make a photo where that weird phenomenon shows up. Same as the odd pink that has popped up in different ways on D50 photos. Or the odd maze artifacts that pop up sometimes in D70/D70s photos.

Hello coldrain,

Thanks for your reply. But, I'd have to say I'm still inclined to go with my guess that it might be a "quality control" issue. As it was mentioned that this might not occur in all D200 cameras even in the worst conditions. Again, this is just a guess and nothing more.

Image-Resource will do a followup on this issue in time. I assume after hearing from the engineers at Nikon. So, till then...it's any one's guess as to what is the gremlin...

Well, on to other things...keep discussing as you please guys. Time for me to bow out. And thanks again for all your comments people. Much appreciated!!

erichlund
01-24-2006, 02:13 PM
I had to create a pretty extreme test to create it, but I've included some links here to show what it looks like. The first shot is the control, a normal exposure (ISO 100, matrix metered) shot of an halogen lamp. If I blow this up to 400% in NC, I can see either banding or pixelation in the margin between the light and dark. Other than that, clean as a whistle. The dark body of the lamp is admittedly just a bit underexposed, but that's not surprising considering the subject.

http://eric-lund.smugmug.com/photos/53753933-M.jpg

The second is the same lamp, but with ISO400 and +3.3EV.

http://eric-lund.smugmug.com/photos/53753929-M.jpg

And a 100% crop of the banding on the second.

http://eric-lund.smugmug.com/photos/53753927-M.jpg

Finally, here's a link to the gallery (http://eric-lund.smugmug.com/gallery/1108651/2/53753927), so you can take a closer look if you wish.

Moral of the story. I'm beginning to be convinced that the issue has to do with the multichannel method of getting the data off the chip. I suspect the channels are not quite balanced correctly. This does not appear with normal exposures, but does raise it's ugly head when severe overexposure occurs, resulting in blown highlights adjacent to dark regions. This would jive with reports from Nikon Sweden where they are being trained on tuning the sensor output to remove banding (not simple hardware replacement, as was done with one camera). Being a tuning issue, there are likely stronger and weaker cases of the issue.

There are other, more severe cases of banding, but I have not been able to reproduce those on my camera. There may be other problems, such as those discussed previously.

I believe Nikon is working on solutions to all the forms of banding, and the best thing we can do is keep them informed of issues we are finding, with sample photos. The more evidence they have of the problem, the faster and more likely they will come up with effective solutions.

So, what I'm doing is sending the information to Nikon for their evaluation. I'm not going to send my camera at this point, since under normal conditions, it's working just fine. However, I suspect I will send it in when there is sufficient evidence that they have come up with a high confidence fix.

Cheers,
Eric

Addendum: If I had to guess, I'd say this problem is actually related to the red channel issues we see with the D70. In the bayer grid, there are two green sensors for each blue and red. It's likely that the blue and red are tuned slightly more sensitive than the green, so when a channel gets blown, the blue and red blown channels are overloading and bleeding into the dark areas, causing the banding. Of course, this is purely guesswork on my part, and may be totally out to lunch.

coldrain
01-24-2006, 03:00 PM
Try the same shot with ISO 400, apparently it will get even stronger then...

D70FAN
01-24-2006, 04:42 PM
I've come to the conclusion that Nikon needs to spend some quality time in Beta testing, as this glitch was discovered pretty quickly in the field.

For future reference, if Nikon developers are reading this... I volunteer. My fee as a Beta site would be a production model of the Beta camera.;)

Eric, correct me if I'm wrong but isn't multi-channel data transfer a CMOS design methodology?

coldrain
01-24-2006, 05:40 PM
Eric, correct me if I'm wrong but isn't multi-channel data transfer a CMOS design methodology?
According to the author of the above linked "the corduroy effect' article, they transfer the data via two sides, to make the speed of the D200 possible. His explonation is a very interesting read, and he explains it much better than i can repeat it, so read the article, a very good read.:)

BigConig
01-25-2006, 09:37 AM
How did you and your dad test that?

I took pictures of my halogen floor lamp at ISO 400 +4 EV. My father hasn't tried to find banding as far as I know but when I saw him last weekend the counter on his D200 was around 3500 shots and he said he hadn't seen it yet.

coldrain
01-25-2006, 09:50 AM
I took pictures of my halogen floor lamp at ISO 400 +4 EV. My father hasn't tried to find banding as far as I know but when I saw him last weekend the counter on his D200 was around 3500 shots and he said he hadn't seen it yet.
It is not there in most photos, and hard to spot in others. I think one might not notice it for a long time when the effect is only moderate, and you do not print in big sizes or look at photos at 100%.

How do you like your D200? And how do you like the colours it produces?

erichlund
01-25-2006, 10:26 AM
Try the same shot with ISO 400, apparently it will get even stronger then...
Read again. The shot showing banding was ISO 400

BigConig
01-25-2006, 10:29 AM
It is not there in most photos, and hard to spot in others. I think one might not notice it for a long time when the effect is only moderate, and you do not print in big sizes or look at photos at 100%.

How do you like your D200? And how do you like the colours it produces?

I love it. metering and color are great. The one thing that still amazes me everytime I shoot is how good metering and white balance are when I shoot with the SB-600 flash. I' (not-so) patiently waiting for the days to get longer and the weather to get nicer so I can give it a real work out. but the snap shot type stuff I've done to date has been great.

Here's an example. Now I know there is nothing special about this shot content-wise, but shooting a white dog with a flash before I always either had badly blown highlights or underexposure. I also used to get a lot with a greenish hue. This was just basicly a point and shoot in aperature priority with the speedlight and auto WB. It really captured the color and texture of Bandit's fur.

http://www.pbase.com/bigconig/image/55231002/large.jpg

Anyways, I hope to have some better samples to share at some point (there are a few in my pbase gallery) but when I looked at this shot (especially full size) I realized the camera's true capabilities.

erichlund
01-25-2006, 10:54 AM
I've come to the conclusion that Nikon needs to spend some quality time in Beta testing, as this glitch was discovered pretty quickly in the field.

For future reference, if Nikon developers are reading this... I volunteer. My fee as a Beta site would be a production model of the Beta camera.;)

Eric, correct me if I'm wrong but isn't multi-channel data transfer a CMOS design methodology?
I'm not exactly privy to Nikon's engineering data, but their literature talks about it a bit, and you can probably make some assumptions from that. There are four pixels in the bayer pattern, two green, a red and a blue. Nikon processes the data off the processor in four channels, two green channels, a red channel and a blue channel. Put these two things together and it seems logical that they extract the pixel data directly from the sensor into the appropriate channel. No sense combining then then splitting them apart, as that would create a choke point. So, I would assume they would extract all green1 pixels into the green1 channel, all green2 pixels into the green2 channel and so forth.

I'm not sure where I picked up the info on the red and blue pixels being tuned more sensitive, but I've read it more than once. It does tend to make sense, to balance those against the two green pixels.

OTOH, all this could be a load of... well, you get the point. To a certain degree, we're just guessing here.

I do want to emphasize that this is not a huge problem. I would never create an exposure like that as a keeper. I tried the same thing taking a picture out of a window, hoping the bright window would blow out, creating the effect, but the light colored curtains framing the window did not show any signs of banding. Apparently, unless you have a very severe case, which may include bad hardware, you have to go to pretty extreme measures to create the banding. For cameras like mine, it requires light against dark and extreme over exposure.

I haven't tried every combination to determine the minimum point at which I get banding, and frankly, I'm not going to. I not planning on sending my camera in until I feel that it would be beneficial to do so. Right now, I think Nikon is still in the learning stage on this problem. Some cameras have been reported coming back better, but still able to produce banding. I suspect that what I have is what they are left with.

I think the main point that I would like to make is that this is not something to panic over. The camera is capable of taking great pictures, as it is right now. Now I have to get my own skill level up to what this camera can do. If I start seeing banding in regular photos, then Nikon will get to fix my camera. Until then, I'll just keep using it. If they appear to come up with a fix for my minimal case that still requires that I send in the camera, I'll make that decision when the time comes. We aren't there yet.

Cheers,
Eric

erichlund
01-25-2006, 11:10 AM
How do you like your D200? And how do you like the colours it produces?
I've been very satisfied with the colors. White balance is generally good, with the usual caveat that I believe Nikon's test lab has a different version of incandescent light than the rest of the world. I've just never had much luck with their incandescent setting. Auto white balance works much better, as does preset white balance or setting color temp directly.

In addition to my opinion, Bjorn Rorslett's friend, the studio photographer, decided to switch entirely from his S3 to the D200. We are all, by now, familiar with the fact that the S3 is considered to have the best color and dynamic range in the dSLR world. Otherwise, no one would put up with it's horribly slow performance. The D200 was good enough for him to drop the S3 entirely. That doesn't mean it's got better color and dynamic range. Just close enough that all the other stuff makes the D200 the better choice.

D70FAN
01-25-2006, 11:58 AM
I'm not exactly privy to Nikon's engineering data, but their literature talks about it a bit, and you can probably make some assumptions from that. There are four pixels in the bayer pattern, two green, a red and a blue. Nikon processes the data off the processor in four channels, two green channels, a red channel and a blue channel. Put these two things together and it seems logical that they extract the pixel data directly from the sensor into the appropriate channel. No sense combining then then splitting them apart, as that would create a choke point. So, I would assume they would extract all green1 pixels into the green1 channel, all green2 pixels into the green2 channel and so forth.

I'm not sure where I picked up the info on the red and blue pixels being tuned more sensitive, but I've read it more than once. It does tend to make sense, to balance those against the two green pixels.

OTOH, all this could be a load of... well, you get the point. To a certain degree, we're just guessing here.

I do want to emphasize that this is not a huge problem. I would never create an exposure like that as a keeper. I tried the same thing taking a picture out of a window, hoping the bright window would blow out, creating the effect, but the light colored curtains framing the window did not show any signs of banding. Apparently, unless you have a very severe case, which may include bad hardware, you have to go to pretty extreme measures to create the banding. For cameras like mine, it requires light against dark and extreme over exposure.

I haven't tried every combination to determine the minimum point at which I get banding, and frankly, I'm not going to. I not planning on sending my camera in until I feel that it would be beneficial to do so. Right now, I think Nikon is still in the learning stage on this problem. Some cameras have been reported coming back better, but still able to produce banding. I suspect that what I have is what they are left with.

I think the main point that I would like to make is that this is not something to panic over. The camera is capable of taking great pictures, as it is right now. Now I have to get my own skill level up to what this camera can do. If I start seeing banding in regular photos, then Nikon will get to fix my camera. Until then, I'll just keep using it. If they appear to come up with a fix for my minimal case that still requires that I send in the camera, I'll make that decision when the time comes. We aren't there yet.

Cheers,
Eric

Yeah, makes sense when you talk about color channels. I was thinking of a different transfer function. Thanks.

Like the phantom moire' issue on my D70, I don't think the banding issue would prevent my buying a D200. It's still a money thing.;)

ridewya
01-27-2006, 09:00 PM
This is taken with My Nikon D200 and Tokina 28-80mm F2.8 at F11 I cropped it 100 % on the sun but there is no banding on mine:D

erichlund
01-27-2006, 11:22 PM
This is not the type of shot that would normally produce banding. You need something dark next to the blown highlights. That's where the banding occurs. In most cameras, even then it is minimal. You have to use an extraordinary overexposure to create it, then view it at least at 100% magnification.

There are exceptions. There are some cameras with faulty parts that are being replaced. These generally show much more severe cases of banding, but the quantity of these cameras I believe to be well within normal manufacturing tolerances. You are always going to have a few bad apples.

The banding that only requires a balancing (Nikon calls it tuning) of the channels to fix the banding can occur in varying degrees, but is mostly not severe, and if not for pixel peepers (God bless them), it may have been months before we discovered there was an issue at all. I mean, who would have seriously taken a way overexposed picture of a light bulb, just to see what the camera could handle. I never would of thought of doing it.

Cheers,
Eric

D70FAN
01-28-2006, 07:50 AM
According to the author of the above linked "the corduroy effect' article, they transfer the data via two sides, to make the speed of the D200 possible. His explonation is a very interesting read, and he explains it much better than i can repeat it, so read the article, a very good read.:)

Thanks for that pointer.