black horizontal band

[gravastar]

Advertisement (gone after registration)

I posted this comment in another thread with reference to the edge pixels/circuitry. It would be a useful experiment to try:

 

"If the reference pixels are present they are probably protected from light by an opaque mask. It could be that due to the angle at which the light hits the sensor light is getting under the mask and causing the problem. If that's the case an experiment would be to see if the problem is worse with very wide angle lenses compared to a more telecentric design such as a 90mm."

 

Bob.

[khanosu]

Hi Robert,

 

I have three lenses, 28mm Summicron APO, 50mm Summicron, and 90mm Summicron APO. They all show this banding about the same amount (as far as I can detect). The strength of the banding depends strongly on placing the light source at the sweet spot. Since the frame of M cameras is not very precise it is a bit of a hit and trial to get the precise sweet spot. If we could precisely hit this spot then we could keep this as a constant and try out different lenses.

 

However you have a good point about the mask covering the peripheral pixels so the light has to go under the mask at an angle. Perhaps this is why it is more difficult to produce the banding from the sun (direct light rays parallel and 90 degrees to the lens) as opposed to house hold bulb filaments where the direct light comes out radially?

 

Furrukh

[khanosu]

Tom it took me some time to reproduce this with the sun. Have you tired this with a bulb at night? If you can not reproduce it with a strong light source at home (bright bare bulb) following the steps in this thread then perhaps they have fixed this problem in the latest batch of cameras. My camera came to stock at B&H about two weeks ago and delivered about a week ago.

 

Thanks for the input.

 

Furrukh

[ho_co]

If the reference pixels are present they are probably protected from light by an opaque mask.

Sounds logical, Bob, but my reading of the M8 sensor spec sheet doesn't admit of that.

 

Check KODAK Image Sensor Solutions - KAF-10500; from the "Specifications" tab download the PDF "Datasheet."

 

Although I don't understand everything going on in the sensor, it looks to me as if there is no baffle.

 

Your test is certainly worth a try, though. The more information you lucky M8 owners can present to Leica, the better.

 

--HC

[jedisdl]

Make that 5 sick ones following this forum - it'll be interesting to hear Leica's response.

 

I received my M8 last Friday (Dec. 22) - the horizontal band ends exactly midway when there is a strong light source near the edge of the image. It is somwhat stronger when using a 486 IR filter, if that provides a clue. I tested it against both tungston and halogen light sources, both of which produce the effect. The lens was a 28 f2 ASPH.

 

I've also sent Leica (USA) banding image files. Maybe it will be fixed with 1.10 (I can always hope!). I'm sure Leica will eventually get it right. The frustration is that the images produced are superb, when these infrequent problems are not present.

 

Steve

[khanosu]

Bob and Howard,

 

I guess to do the test one would mount a lens and then do a sequence of frames with small incremental movements of the camera with the light source moving out of the frame, and choose the frame with the worst banding. Do the same with different lenses and compare the worst banding case for each lens. The incremental movements have to relatively small for shorter lenses as compared with the longer lenses. If this is repeated enough number of times then a pattern might emerge out of this. Tedious but doable

 

Furrukh

[khanosu]

Advertisement (gone after registration)

Hi Steve,

 

In this thread alone there are examples/posts of 9 different cameras exhibiting this behavior. I am sure Leica will fix this. Hoping that this does not involve a trip to Germany. The good news is that it happens rarely – especially with the sun.

 

Furrukh