[Medex]

Monochrome converted sensor / camera family grows up. K100D, K200D, K-5, GR II, K-3, K-30 and now 36 mpix FF sensor (D800/ K-1).
Conversion was difficult but sensor is alive and shows no scratches. Some cleaning is needed due to tiny specks of CFA left and replacement filter as well because now original filter is left for protection and as I know it has anti-aliasing properties.
Sensor is from Nikon D800, but we know that our beloved K-1 has same sensor. D800 camera is partly invalid due to missing arrow / navigation button and I got quite old Nikkor 50/1.4 D for testing purposes. So, F1.4 picture is a little softish, bet another two at F5.6 are really sharp.
All pictures shot in live view mode because mirror is misfunctioning.
RAW files in LR looks sharp, converted DNGs to avoid demosaicing are even more sharp. And this is not the limit due to AA filter left 🙂.
Full resolution JPEGs are uploaded to Flickr.

[CedrusMacro]

Thanks for keeping us updated Medex, it's awesome to see what potential is available with a monochrome variant if these sensors. The FF look very crisp.

[Medex]

Thanks for keeping us updated Medex, it's awesome to see what potential is available with a monochrome variant if these sensors. The FF look very crisp.

Thank you. As I said, this is not the limit, because anti-aliasing filter still sits on the sensor. It should be removed for maximum sharpness. Will do it soon.

[UncleVanya]

It would be nice to see comparison shouts with and without CFA - I realize you probably don’t have two d800’s - do you have a K-1 to compare with? It’s not clear how much the gain is but I assume it is a gain larger than the removal of the AA filter provides alone (and I understand the AA filter remains in place for now.)

Really interesting work.

[ChristianRock]

Fantastic work, as always! Hopefully removing the AA filter won't scratch the sensor.

[Medex]

Fantastic work, as always! Hopefully removing the AA filter won't scratch the sensor.

No for sure, because removal of AA filter + UV-IR cut filter is very simple - just 4 screws to unscrew.

---------- Post added 02-11-22 at 12:14 AM ----------

It would be nice to see comparison shouts with and without CFA - I realize you probably don’t have two d800’s - do you have a K-1 to compare with? It’s not clear how much the gain is but I assume it is a gain larger than the removal of the AA filter provides alone (and I understand the AA filter remains in place for now.)

Really interesting work.

I have now Nikkor 50/1.4D for Nikon D800 and DFA* 50/1.4 for Pentax. Can do comparison but I think DFA* 50/1.4 will be sharper. Maybe f4-5.6 will be my choice.
Theoretically removal of Bayer CFA should give pixel shift effect, just handheld (no tripod required). And K-1 pictures really differ when shot in normal mode and Pixel Shift mode, even K-1 has not AA filter.

[UncleVanya]

No for sure, because removal of AA filter + UV-IR cut filter is very simple - just 4 screws to unscrew.

---------- Post added 02-11-22 at 12:14 AM ----------



I have now Nikkor 50/1.4D for Nikon D800 and DFA* 50/1.4 for Pentax. Can do comparison but I think DFA* 50/1.4 will be sharper. Maybe f4-5.6 will be my choice.
Theoretically removal of Bayer CFA should give pixel shift effect, just handheld (no tripod required). And K-1 pictures really differ when shot in normal mode and Pixel Shift mode, even K-1 has not AA filter.

If you can grab a manual focus Nikon lens it will mount and focus on the Nikon and Pentax bodies.

IMGP1217 by -vanya_42nd-

Nikon 50 mm f/2 on Pentax K3 by -vanya_42nd-

It won’t be locked in place unless you add a lock slot for this purpose. I would only do it with lightweight small lenses.

[Digitalis]

The results you're getting are interesting to see, the image quality you are getting seems to me in line with what I'm getting with my Leica Monochrom and Phase one Iq3 Achromatic both of which come with monochrome sensors. Phase one with their achromatic digital backs mention that the sensors they use have to be of an unusually high standard with minimal defects to be used in these cameras. Leica also makes similar claims with the sensors they use in the Monochrom - taken together these statements imply that there is a lot going on behind the scenes with bayer sensors, and that monochrome sensors can't simply have hot pixels and other defects simply mapped out quite so easily. I think with the Leica monochrom - which shoots DNG there is a flag the camera uses in the metadata to avoid unnecessary de-bayering of the files.

Considering you are starting with a bayer sensor, the results you are getting are impressive.

[ProfessorBuzz]

I don't understand what you are doing. Have you removed the optical cover glass on the CMOS APS sensor itself, then removed the microlens array and Bayer matrix filters above the silicon?

[Medex]

I don't understand what you are doing. Have you removed the optical cover glass on the CMOS APS sensor itself, then removed the microlens array and Bayer matrix filters above the silicon?

Yes, exactly what you mentioned. I did "nude" sensor without microlenses and Bayer color filter array. Sensor now fixes luminance information only.

[BenwayB]

Congratulations �� !!

Did you do it chemically or mechanically this time?
I got stuck with my K10 at the moment, as I don't manage to remove the glued glass from the sensor... ;-)

[Medex]

Congratulations �� !!

Did you do it chemically or mechanically this time?
I got stuck with my K10 at the moment, as I don't manage to remove the glued glass from the sensor... ;-)

I used same method as with K5, GR II, K-3 and K-30. Mechanical sratching is the main option for me. But some chemistry is used to make scratching easier and not so agressive.

---------- Post added 02-12-22 at 12:12 AM ----------

The results you're getting are interesting to see, the image quality you are getting seems to me in line with what I'm getting with my Leica Monochrom and Phase one Iq3 Achromatic both of which come with monochrome sensors. Phase one with their achromatic digital backs mention that the sensors they use have to be of an unusually high standard with minimal defects to be used in these cameras. Leica also makes similar claims with the sensors they use in the Monochrom - taken together these statements imply that there is a lot going on behind the scenes with bayer sensors, and that monochrome sensors can't simply have hot pixels and other defects simply mapped out quite so easily. I think with the Leica monochrom - which shoots DNG there is a flag the camera uses in the metadata to avoid unnecessary de-bayering of the files.

Considering you are starting with a bayer sensor, the results you are getting are impressive.

Thank you.
I try now different Pentax cameras I can buy for low price but in good shape. This is step by step process until I will be prepared for K-1 monochrome. Nikon was just experiment, but successful. I think I will sell this sensor for after I finish with cleaning and filter replacement

[Joetitch]

Wish I had a spare K1 I could get you to convert accepting that if it went wrong it would be my problem not yours. The idea excites me even though I have no experience in monochrome. A whole new world to explore.

[Medex]

Wish I had a spare K1 I could get you to convert accepting that if it went wrong it would be my problem not yours. The idea excites me even though I have no experience in monochrome. A whole new world to explore.

Some people says that fujifilm gfx 100 has pixel shift mode that gives 400 mpix resolution. I am not sure about 4x increase in resolution due to my comparative experience with monochrome sensors but at leats 2x increase is real. So Pentax K-1 with 72 mpix mono sensor excites me as well. Images look more punchy, objects in focus stand out more. And 36 mpix sensor is one of the best IMO.

[Digitalis]

The Leica M10 Monochrom has 40 Megapixels which is plenty considering the small buffer size on the M10.

Lenses such as the Leica 50mm f/1.0 noctilux get pushed to their limits on this sensor.


Leica M10 Monochrom - Leica M-Noctilux 50mm f/1.0 @ f/1.0 ISO 160 1/4000th + Hoya 25A filter



Leica M10 Monochrom - Leica M-Noctilux 50mm f/1.0 ISO 160 1/500th -probably stopped down to f/4 where it performs better.