Tons o Glass
Veteran Member
I picked up a Canon J15x9.5B at my local swap meet a while back figuring it would be fun to mess around with for $5. Can you blame me? A 9.5mm to 143mm f/1.8 zoom sounds beyond impressive even if it's only meant to cover 2/3"-type sensors. I whipped up a 3D-printed adapter and gave it a try on my A3000.
Oops! There are major color fringing issues on a single sensor wide open, and it's even still noticeable when the lens is stopped down to f/11. Thinking there could be an issue with the lens I then did the research I should've done a little earlier in the adapting pipeline, hehe.
As it turns out, this is by design (or lack thereof?), as I'm sure some of you already know. These B4-mount zooms are intended for use with 3CCD broadcast cameras that employ a split prism and three sensors (one for each color channel) that each have slightly different flange distances appropriate for their respective colors - the sensor positioning corrects for longitudinal CAs - what a nifty system (sounds like it would be a chore to calibrate/align the three sensors though)!
Here's a quick test I did to loosely emulate using three sensors to give the lens a more fair shake. I just varied focus a bit with three images and stacked the best-matching color channels (one red, one green, one blue) from each image. If I had some old R/G/B front-mounted filters that fit it, it would have been easier to get more optimally focused channels to stack.

One of the three images used for stacking. 143mm f/1.8. The color-fringing issues are several orders of magnitude worse on the wide end of the lens, which makes sense.

A stack. You'll notice I didn't use the green channel from the image above - the red channel from this one matched best for a clean stack of subsequent frames.
LoCA cleaned up pretty well doing this, methinks, but it would be quite the chore to use the lens like this with a single sensor out in the field for actual images.
So on to my public service announcement: Avoid using or adapting B4-mount lenses unless you have a proper 3CCD camera (anyone know of an affordable 1080p one, BTW?)!
Oops! There are major color fringing issues on a single sensor wide open, and it's even still noticeable when the lens is stopped down to f/11. Thinking there could be an issue with the lens I then did the research I should've done a little earlier in the adapting pipeline, hehe.
As it turns out, this is by design (or lack thereof?), as I'm sure some of you already know. These B4-mount zooms are intended for use with 3CCD broadcast cameras that employ a split prism and three sensors (one for each color channel) that each have slightly different flange distances appropriate for their respective colors - the sensor positioning corrects for longitudinal CAs - what a nifty system (sounds like it would be a chore to calibrate/align the three sensors though)!
Here's a quick test I did to loosely emulate using three sensors to give the lens a more fair shake. I just varied focus a bit with three images and stacked the best-matching color channels (one red, one green, one blue) from each image. If I had some old R/G/B front-mounted filters that fit it, it would have been easier to get more optimally focused channels to stack.

One of the three images used for stacking. 143mm f/1.8. The color-fringing issues are several orders of magnitude worse on the wide end of the lens, which makes sense.

A stack. You'll notice I didn't use the green channel from the image above - the red channel from this one matched best for a clean stack of subsequent frames.
LoCA cleaned up pretty well doing this, methinks, but it would be quite the chore to use the lens like this with a single sensor out in the field for actual images.
So on to my public service announcement: Avoid using or adapting B4-mount lenses unless you have a proper 3CCD camera (anyone know of an affordable 1080p one, BTW?)!