Foveon Green ...

xpatUSA

Forum Pro
Messages
26,773
Solutions
25
Reaction score
10,185
Location
-, TX, US
... sometimes you see it, sometimes you don't - but from the same capture?!

Starting quite off-topic, I went out and shot some candidate "Merrill" clouds - intending compare various converters. First one I tried (RT) showed some green in a corner of the frame. Next, I tried SPP - no green. Then RawDigger - green. Then extracted the embedded JPEG - no green.

Mira:

See captions for which is which ...
See captions for which is which ...

Looks like a ray angle effect which Sigma knows about and can fix. Images above are crops from the top left corner of the scene.

Tried the same shot later with the DP2 - no obvious corner coloration. I remember that the F13 sensor in the DP does have different microlenses than the SD14/15; I also remember wondering here if the DP Merrill F20 sensor is similarly different than the SD1M.

I now wonder if the F20 DP sensor is the same as the SD1 but is corrected by software/firmware?

--
Ted
 
Last edited:
That's some serious green Ted.
 
That's some serious green Ted.
Yep. Today, I fired up Kalpanika and got a DNG and a TIFF output. Both had green up in that corner, so one can only conclude that SPP and the in-camera JPEG engine do have secret sauce at work.
 
That's some serious green Ted.
Yep. Today, I fired up Kalpanika and got a DNG and a TIFF output. Both had green up in that corner, so one can only conclude that SPP and the in-camera JPEG engine do have secret sauce at work.
I'm guessing that the Merrill cameras will really suffer over time, as new lenses are made, but they won't be updating their capability to fix those color casts in-camera. I guess that would only apply if they keep on making new model DSLR lenses though, which I have predicted is not likely for long (and the more I think about it, the more I think there might not even be one new model after the ones they have just introduced, though I hope that's wrong, and there are at least a couple good new DSLR lenses introduced in 2019, like a 105mm f2.8 OS Art and a 24-135mm f4 OS Sport).
 
This is the infamous Merrill color cast caused by the varying angle of incident of the light rays on the sensor. SPP corrects for it, as well as the in-camera jpeg engine.

For Kalpanika, X3F Wrapper can be used to insert corrective opcodes in the generated DNGs. ACR and LR understand these and create images withour color cast.

Could not test this with RT, unfortunately: on my machine RT 5.5 crashes whenever I try to open a DNG generated by Kalpanika, with or without the corrective opcodes.
 
This is the infamous Merrill color cast caused by the varying angle of incident of the light rays on the sensor. SPP corrects for it, as well as the in-camera jpeg engine.
Indeed, as I said in the OP "Looks like a ray angle effect which Sigma knows about and can fix."
For Kalpanika, X3F Wrapper can be used to insert corrective opcodes in the generated DNGs. ACR and LR understand these and create images withour color cast.
Interesting - is there a link explaining more about these "corrective opcodes"?
Could not test this with RT, unfortunately: on my machine RT 5.5 crashes whenever I try to open a DNG generated by Kalpanika, with or without the corrective opcodes.
Mine too. The solution is to run the Kalpanika DNG output through the Adobe Raw to DNG Converter first. The bonus is that the file-size about halves.
 
That's some serious green Ted.
Yep. Today, I fired up Kalpanika and got a DNG and a TIFF output. Both had green up in that corner, so one can only conclude that SPP and the in-camera JPEG engine do have secret sauce at work.
I'm guessing that the Merrill cameras will really suffer over time, as new lenses are made, but they won't be updating their capability to fix those color casts in-camera. I guess that would only apply if they keep on making new model DSLR lenses though, which I have predicted is not likely for long (and the more I think about it, the more I think there might not even be one new model after the ones they have just introduced, though I hope that's wrong, and there are at least a couple good new DSLR lenses introduced in 2019, like a 105mm [f/2.8] OS Art and a 24-135mm [f/4] OS Sport).
Seems like you're talking about the SD1 Merrill with all these oracular lens-mentions. My thread is about the DP1 Merrill.

--
Ted
 
Last edited:
This is the infamous Merrill color cast caused by the varying angle of incident of the light rays on the sensor. SPP corrects for it, as well as the in-camera jpeg engine.
Indeed, as I said in the OP "Looks like a ray angle effect which Sigma knows about and can fix."
For Kalpanika, X3F Wrapper can be used to insert corrective opcodes in the generated DNGs. ACR and LR understand these and create images withour color cast.
Interesting - is there a link explaining more about these "corrective opcodes"?
Could not test this with RT, unfortunately: on my machine RT 5.5 crashes whenever I try to open a DNG generated by Kalpanika, with or without the corrective opcodes.
Mine too. The solution is to run the Kalpanika DNG output through the Adobe Raw to DNG Converter first. The bonus is that the file-size about halves.
X3F wrapper correcting Merrill color cast (take the 0.57 beta):


DNG opcodes are part of the DNG specification, and essentially describe parameters to be used in various raw-decoding stages for optional post-processing. One of these operations would be flat-fielding, and the corresponding parameters can be included under the "OpcodeList3" tag in DNG raws. Some background information is here


and here


Regards, Georg
 
BTW, just checked: RT 5.5 really does not crash when opening Kalpanika-generated DNGs which have been post-processed by Adobe DNG converter. But it does not get by well with the opcode information and shows even more weird color casts.

To get around this: select Camera Raw compatibility "4.6 and newer" in Adobe DNG converter's options. This way, DNG converter directly applies the opcodes to the raw values.
 

Keyboard shortcuts

Back
Top