Reversed 1:2 Macro Lenses?

[ray_parkhurst]

I'm still wondering about solutions for 3-4x magnifications with very large image circles for camera-movement stitching, and started wondering if an older 1:2 manual focus macro lens in reverse would give good performance. Has anyone experimented with such a lens? My starting point for this work is the 105/4 Inspec.x L 3.5x, so to make sense the lens needs to have sharp performance at larger than f4. There are a few f2.5 Macros which are designed for 1:2, and these would have potential to fit the bill. I have a Vivitar 90mm f2.5 "Bokina", but it is abysmal for LoCA wide open, and not great at f3.3. It starts to look decent for CAs at f4, but that's too narrow.

Anyone have experience reversing a 1:2 Macro at wide aperture at 3-4x mag? What are the candidate lenses?

I am assuming the image circle will be huge at 3x for such a lens, perhaps >120mm. Is this a reasonable expectation?

[clarnibass]

Do you mean just the lens extended with tubes or bellows?

I haven't tried that, but I've used a reversed Nikon 55mm micro (manual) lens for coupling and it's excellent. Mine is f/3.5 and best at f/5.6, but still very good at f/3.5. There is a f/2.8 model available too.

I can't really say much about the image circle, but when reversed on top of the 105mm lens, or a 200mm lens, it doesn't cover FF camera sensor at all. It shows the circle with complete black around it. It does cover an APSC sensor with very little vignetting (i.e. darker at the corners but the frame is there, not just black outside the image circle).

I don't know if this helps at all if you just want to extend the reversed lens.

[ray_parkhurst]

Yes, just simple extension on bellows to get 3x in reverse. Image circle "should" be huge.

[Lou Jost]

I often use the Oly 60mm in reverse for that. It is very good but with some slight red fringing near the edges. Maybe these would go away at some sufficiently-large magnification. But for your use this would be hard, because you have to maintain the 17mm flange-sensor distance which becomes the working distance. You also would have to build an auto reversing ring set. I know you prefer lots of WD.

All reversed mass-marketed macro lenses with floating elements will have this WD problem, though I don't mind the Nikon or Canon flange-sensor distance as a WD.

[ray_parkhurst]

I often use the Oly 60mm in reverse for that. It is very good but with some slight red fringing near the edges. Maybe these would go away at some sufficiently-large magnification. But for your use this would be hard, because you have to maintain the 17mm flange-sensor distance which becomes the working distance. You also would have to build an auto reversing ring set. I know you prefer lots of WD.

All reversed mass-marketed macro lenses with floating elements will have this WD problem, though I don't mind the Nikon or Canon flange-sensor distance as a WD.

Yes, 17mm would probably be too short. The Canon/Nikon register is also pretty short for my purposes but the situation is very different for camera-panning vs lens- or subject-panning. When camera-panning, the lighting remains fixed, so can potentially be optimized even for shorter WD than I could use for lens-panning.

[Lou Jost]

I hadn't thought of that. The Cosina 65mm and the some of the latest Fuji macro lenses seem exceptionally good from reports.

What actual FOV are you interested in? That determines the type of lens you want. If your subject is less than 17mm across, a reversed MFT lens should provide more detail than an APS or FF lens. Total lens resolution (in lp/picture width) is the thing that matters when reversed, and for economic reasons, that total resolution does not increase linearly with the lens's designed format size. Consumer 6cm x 7cm medium-format lenses have more total resolution than consumer FF lenses, but not twice as much. Consumer FF lenses on the average have more total resolution than MFT lenses, but again, not twice as much. So I think you should use the smallest format lens you can get away with.

But this is not a law of physics, it's a law of economics, so there could certainly be industrial or scientific lenses that break it.

[ray_parkhurst]

The FOV needs to be ~21mm tall, so I'm in the range between APS-C and FF.

Edited to add: at 3.5x, this would seem to make the image circle requirement 104mm, but I don't need sharpness to the corners, only to the edges. This is one advantage to shooting round things.