Recipe for Banding (for Bjorn and others)

[bendheim]

After reading all this banding stuff, I decided to give it a try.

Yes, I think banding can be reproduced, if you want to. I've seen nothing on normal shooting.

Simple. Set the camera at ISO 400. Aim it at a bright flourescent light source..it seems to be more noticeable than incandescent. Set your meter EV at between +1 and +2.

Make sure that you have left a good bit of space around the source. Try the ends of a flourescent fitting, where the struts and bits stick out.

Meter only off the light source, so the areas surrounding it are darkish.

In Photoshop, pull up the curves considerably to brighten the dark areas around the source. Sharpen at say 250/1/1

View at 200 percent. Chances are you have banding in the pulled up (formerly dark areas) around the light source.

It's much, much harder to reproduce this shooting a naked incandescent bulb...Bjorn I wonder why? Just traces of banding at 200 percent immediately adjacent to the light source...

Personally I'm leaving well alone unless this becomes a standard shooting issue. I don't intend photographing these sorts of things, besides, I would expose them differently.

Peter

 

[nfoto]

You probably get a "beat" pattern with the AC source. This recipe for banding doesn't work when you shoot into the sun, I've tried and failed (or rather, D200 failed but yet didn't fail).

I've spent several wasteful days exploring alleged this banding issue and have had my fill of the darned thing.

 

[bendheim]

I've spent several wasteful days exploring alleged this banding issue and have had my fill of the darned thing.

LOL ! Yip, I'm personally not concerned, it's just scientific curiosity. Like you, I'm naturally curious, but from a distance...

 

[nfoto]

However, unlike you, I'm expected to provide a review ..... Hence my slightly abated curiosity.

 

[PJohnP]

After reading all this banding stuff, I decided to give it a try... Make sure that you have left a good bit of space around the source. Try the ends of a flourescent fitting, where the struts and bits stick out... It's much, much harder to reproduce this shooting a naked incandescent bulb...Bjorn I wonder why? Just traces of banding at 200 percent immediately adjacent to the light source...

Peter :

I haven't had any banding as yet, but what I've seen of the other discussions on this is that it does seem to be occurring within artifical light, not daylight conditions.

At a guess, I'd hazard that there is something occurring with the fluorescing effect (which runs at either 50Hz or 60Hz depending where you are in the world or in some cases multiples thereof). Incandescent lighting doesn't show the same cycling as the filament is excited continuously (not truly theoretically correct, but eminently correct from a practical engineering perspective).

One could test this more rigorously by using a light source connected to a controlled cyclic AC system to determine if this is related to cycle speeds. Not an easy measure, as the power source is not something many people would have handy, and most fluorescent light systems will not function outside certain cycles. A low level strobe could be used with a slightly longer exposure where the shot is metered against a blown illuminated white subject or something, but that's going to be a pain to set up; again, the cycle speed would be a critical variable to control.

I suppose it's theoretically possible that the sensor in the camera is sampling at some harmonic to the light source cycle rate, and developing a shadow noise pattern based on "misfiring" with some segment of the spectrum. If that's the case, it's going to be yet harder to reproduce the issue, especially as we can't test the D200, as readily in IR or UV spectra ranges (one could especially rule out/rule in one issue if we simply had a hot iron as an IR source - it wouldn't cycle noticably).

I note from your discussion that you (and some other examples I've seen) are metering more for the light source, and not a balanced exposure for the image, which means that you're drawing out effects from the shadows with your processing.

Interesting that the sensor is responding in this fashion, but it doesn't seem to be the "D200-killer" that so many people are positing. I'm personally far more concerned about the additional batteries that I have coming, and hoping that I'm not going to get the ones on the left tail of the quality assurance curve like Geno did.

And how is your New Year, Peter ?


John P.

 

[bendheim]

Hi John P

The New Year finds me fine, thanks, and glad the bloated season has passed. I hope it finds you well and energised for a great year of photography.

I agree that it only happens when exposing for the light source. It doen't bother me personally -- it's a matter of good vs. bad technique...But it is still there. Take a look at this post I did over at DPR

http://forums.dpreview.com/forums/read.asp?forum=1021&message=16579763

using some deliberately bad examples. BTW, I like your theory, except the daylight shot at this sample set kinda rules out the artificial light source as an issue...?

Kind regards to you
Peter

 

[PJohnP]

I agree that it only happens when exposing for the light source. It doen't bother me personally -- it's a matter of good vs. bad technique...But it is still there. Take a look at this post I did over at DPR

http://forums.dpreview.com/forums/read.asp?forum=1021&message=16579763

using some deliberately bad examples. BTW, I like your theory, except the daylight shot at this sample set kinda rules out the artificial light source as an issue...?

Peter :

Hmmm... Well, that does indeed spoil my grand theories. :Smart: :Angry: :Unsure:

How much adjustment did you make in PS to the shots in that thread ?

As for me, I spent the entire holiday period with a bloody awful 'flu bug, and I'm just now starting to get over it, in time to head out for business trips. Nonetheless, I'm always grateful to wake up in the morning...


John P.

 

[bendheim]

Hi John
Sorry bout the flu bug, glad you nearly are over it...

Adjustments - just pulled up the curve in PS from the middle as far up as it could realistically go..that's all.

Except for the daylight image, where I used D-Lighting in NC and no curves in PS.

Hmmm...

 

[PJohnP]

Adjustments - just pulled up the curve in PS from the middle as far up as it could realistically go..that's all.

Except for the daylight image, where I used D-Lighting in NC and no curves in PS.

Hmmm...

Peter :

Well, D-Lighting and the old DEE introduced a certain amount of shadow noise, but in my experience, never banding. That this comes up in both PS and NC, though, would seem to indicate that it's independent of the software, unless the RAW handling from Nikon is at fault (I'm assuming that PS is using the Nikon plugin). It's possible it's the RAW handling from Nikon, but...:Whistle:

If I can get a few spare moments in the next few days, I'll see if I can run a test or two myself. Without more data, I'm unsure if this is a flaw specific to some cameras or if its endemic to the line (D200). We (D200 shooters) probably need to develop some kind of systematic test for this, perhaps with specific over/underexposure criteria, as well, as making specific directions on processing to provide repeatability. Bjørn's made some initial steps in this arena, but we'll all likely have to perform tests to see if it's indeed camera related and/or software. Only with a systematic test format can we get reasonably comparable data, as Bjørn's or my or your camera could be anomolies.

It's all something to ponder.

In the interim, I've had some passable opportunities shooting with the D200 where I was holding exposures close to where they needed to be (Taos), and I'll just have to suffer with a great camera that works well with proper care ! :biggrin:



John P.

 

[bendheim]

In the interim, I've had some passable opportunities shooting with the D200 where I was holding exposures close to where they needed to be (Taos), and I'll just have to suffer with a great camera that works well with proper care ! :biggrin:
John P.

John, those are fantastic shots, well composed and beautifully exposed. I'm sorry I initially missed that thread, but I'm glad that I got to see it. That seems like a really wonderful part of the world, maybe I'll make it there someday...Thanks for pointing my way to your wonderful images.
Best

Peter

 

[Elf_8]

Join the banding wagon

I followed the "recipe" and got it all right.

Subscribe to see EXIF info for this image (if available)

To all that I've read on the topic, I can only add that the effect is visible directly in the LCD at max zoom in the affected area. That rules out the effect of softwares used in postprocessing.

 

[bendheim]

Hi Christian

Christian, I now strongly suspect that it may well be pretty much universal given the right conditions.

But those conditions need to be severely "manufactured"...not in general normal use. I wouldn't over stress about it.

 

[nfoto]

I'm no longer in a position to test these aspects of the D200

However, shooting into a partially cloud-obscured sun gives much more than 2 stops overexposure, and I couldn't get the banding to appear. I could with underexposure, however. So this entire banding business is utterly confusing.

I do agree with Bendheim. Banding may need severely abuse of the image and good photographers should know better.

 

[PJohnP]

John, those are fantastic shots, well composed and beautifully exposed. I'm sorry I initially missed that thread, but I'm glad that I got to see it. That seems like a really wonderful part of the world, maybe I'll make it there someday

Peter :

Well, that's settled. We both have to visit the other someday ! :biggrin: And thanks for the kind thoughts on my shots - coming from a fine photographer like you, that's a high compliment.

I did run a quick test with my D200 and the 28-70mm AFS lens, and I can get very minor effects at 200%. It might be that I need to override the meter even more (I ran it with +2EV) to make this happen or that I needed a darker surround than the tongue-and-groove wooden ceiling in my office. I can see more effect at 400%. I ran D-Light full out to get some effect. But this has to be worked at for such an occurrence - increase ISO, exposure management out of normal bounds, grossly enlarged, etc.

And that echoes a remark of yours to Christian, we have to make this happen. Based on your, Christian's, and Bjørn's findings on this, I don't consider it a fatal flaw. By comparison, I can induce digital noise in my D100 much more easily ! :tongue:

And now, back to writing reports on potential industrial disasters ...


John P.

 

[Elf_8]

Maybe someone can come up with another recipe for revealing aural lights. That would be far more exciting.
Right about the extreme conditions : hyper-contratsty scene, blatant overexposure (in my case +2 EV), shadows severely opened (D-lighting in my case), 200 % view.
Using something similar, one could well reveal the canvas underneith a master painting.

 

[Iliah]

Aim it at a bright flourescent light source..it seems to be more noticeable than incandescent.

For D200 with fluorescent lights blue pixels in Bayer pattern are heavily underexposed. Blue channel amplification for white balancing reaches 6 times, especially if you shoot the ends of the bulb, and your exposure is not 1/30, 1/60, or 1/125 (60Hz mains, more difficult for 50Hz, but 1/30 is pretty much safe). This alone can result in banding. As Bjorn mentioned, shots made outdoors (worst case amplification is 2 times) are less prone to banding. Worst case for incandescent is 4 times amplification.

 

[Elf_8]

I just found this, and found it interesting enough to share.

It is from http://ltpwww.gsfc.nasa.gov/IAS/handbook/handbook_htmls/chapter7/chapter7.html#section7.7.
Banding :"Banding or "scan-to-scan striping", is a sometimes visible noise pattern caused by memory effect. After scanning past a bright target such as clouds or snow, detector response is reduced due to memory effect. Thus if the region past the bright target is uniform, data values obtained from the sensor will be slightly lower than corresponding values obtained on the following scan (since the following scan is in the opposite direction and therefore, has yet to encounter the bright target.) The result is that scans in one direction will be noticeably darker than adjacent scans in the opposite direction. The banding pattern is very small in intensity, typically on the order of 1 to 2 counts."
Memory Effect : It is manifested in a noise pattern commonly known as banding. It can be observed as alternating lighter and darker horizontal stripes that are 16 pixels wide in data that has not been geometrically corrected. These stripes are most intense near a significant change in brightness in the horizontal (along scan) direction, such as a cloud/water boundary. Because of this, it was formerly termed 'Bright Target Saturation' or 'Bright Target Recovery.' ...
ME is known to be caused by circuitry contained in the pre-amplifiers immediately following the detectors in the instrument electronics. It is primarily due to a portion of a feedback circuit that contains a resistor/capacitor combination with a time constant..."

 

[PJohnP]

Christian :

Well, that would fit the conditions for the phenomenon so far observed, even to the change in direction for portrait mode shooting.

Good catch, sir.



John P.

 

[Iliah]

You can look at colour channels to see what is happening. It seems to me that nose is generated by one sensor channel only. Would be nice to look at nef...

 

[jgrove]

Wow this issue is so confussing! Anyway sure enough if i follow the first post i get some banding on my D200.

I also tried my D2hs to see if the same happend and it doesnt.

I really cant see this as a problem becuase people are only seeing it because they are now trying to find it.

If it is a problem then i am sure Nikon will fix it.

For me i am just happy, as the D200 produces some stunning images!