This is bad I think.. on my M8.2 .. HELP

[jaapv]

Advertisement (gone after registration)

I would imagine that it is more pronounced on an M8 due to the extreme angles that light hits the sensor edges compared to most DSLRs. With wider, or other lenses where the rear element to sensor distance is very small, I would think that adequate masking could be quite problematic. However, having said that, it's still not good, and I'm sure that it could be addressed by Leica if they had enough incentive to do so.

Hard to say, Nicole. Basically the problem lies with Kodak.

[Nick_Yoon]

I had this problem in May for the first time in 1.5 years of ownership. Sent a jpg to Leica, they said send it in. They had a look and said it was unfixable, checked out everything, cleaned it up and sent it back to me. Not the best solution, but it's only happened once so far.

[t024484]

Hans,

 

Let's just say we'll take Michael's word for it (and many other tech issues). He may not be directly connected to Leica but I would suggest you have a read of LFI some time where I think you'll find Michael's credentials more than credible.

 

Michaels credentials are indeed more than credible, not only in the Leica World but also in the Hasselblad Scene, but that is not at all the issue.

I was just curious to know if he had more acces to Leica Information than just an outsider.

 

 

Hans

[carcam]

I think it's a UFO

 

This is a well documented problem that occurs when you have a very strong light source just outside the left or right frame edge.

 

See also:

 

http://www.l-camera-forum.com/leica-forum/digital-forum/18270-green-stripe-berlin.html

[t024484]

Hard to say, Nicole. Basically the problem lies with Kodak.

 

Hi Jaap,

 

I am not so sure if only Kodak is to blame. I tend to agree with Nicoleica that the angle of incident might play a major role, plus a possible lack of adequate masking.

Pixels not supposed to receive light, that are there for black reference, obviously seem to be receiving light.

The dark reference pixels are shielded from receiving light with a coating on the Sensor itself, but that is probably not enough to prevent leaking of light with a steep angle of incident under this shielding.

For that reason there is a buffer zone of 20 pixels wide, or 0.14mm, on all sides around the active 10Mp sensor area, where masking already should start, to have a clear cut between active Pixels an Dark Reference pixels.

Maybe this mask is not tight enough or even completely absent, the M8 in that case just relying on the effectiveness of shielded dark reference pixels, which is obviously not good enough in some occasions.

 

Hans

[Julian Thompson]

Its a shame actually - that first frame was very nice :-(

[gravastar]

Advertisement (gone after registration)

I would imagine that it is more pronounced on an M8 due to the extreme angles that light hits the sensor edges compared to most DSLRs. With wider, or other lenses where the rear element to sensor distance is very small, I would think that adequate masking could be quite problematic. However, having said that, it's still not good, and I'm sure that it could be addressed by Leica if they had enough incentive to do so.

 

Incident angle of the light could well be a major factor. In many instances light "leakage" under the edge mask and onto the reference pixels is due to light piping in the transparent silicon dioxide/nitride layers that cover the sensor. Since piping relies on total internal reflection at the various interfaces between layers it's critically dependent upon the incident angle of the light. At normal (near right angle) angles of incidence the light bounces back out and doesn't get trapped, with shallow angles it totally internally reflects and is trapped. (I used to work with CCD sensors .....)

 

Bob.

 

ETA: I've not tried the experiment to see if the problem is less pronounced with a long focus lens where the angle of incidence would be more normal.