APSC2 (APS-C Squared) Rotate-and-Stitch Adapter

ProfHankD

Veteran Member
Messages
9,799
Solutions
32
Reaction score
6,251
Location
Lexington, KY, US
APSC2 (APS-C Squared) Rotate-and-Stitch Adapter is an open-source, 3D-printable, adapter that allows your APS-C Sony E-mount body to capture the largest square format images most full-frame lenses are almost certain to cover. I've been doing a lot of work with polar scanning (e.g., Lafodis160 ), but this adapter is much simpler and fully mechanical, working more like Fotodiox’s RhinoCam Vertex , but with a few important differences.

This is a four-part adapter for Leica M to Sony E -- except it is offset and rotates. The idea is simple: the diagonal of a 36x24mm full frame is approximately 43.27mm, so most full-frame lenses should be able to cover a square with approximately the same diagonal, which would be 30.59x30.59mm. Sony APS-C is approximately 23.6x15.6mm, thus by offset rotation in 90-degree steps it should be possible for four shots to cover up to 31.2x31.2mm. Of course, alignment is never perfect, but allowing for that, it's still possible to get about 30.6x30.6mm... allowing for minor stitching errors, etc., let's call it 30x30mm. Using a 24MP APS-C body, that means you'll get roughly 58MP worth of scene detail (although image stitching software might give you any number of pixels).

Why Leica M? For the same reason the TechArt Pro LM-EA7 uses it: nearly all old lens mounts can be adapted to Leica M, so this adapter is reasonably universal. For example, to use a Canon FD lens, simply mount a Canon FD to Leica M adapter on this and then mount your FD lens on the front. Not only are lots of adapters commercially available, but I've also designed and posted free 3D-printable M adapters for: Argus C3, Minolta/Konica-Minolta AF / SonyA, Canon FL/FD/FDn, Minolta SR/MC/MD, M42, and Kiev 10/15.

For example, I used my old Sony NEX-7 with APSC2 to capture various test images within the coverage of a Canon FD mount Vivitar 135mm f/2.8. For example, one capture with this looks like:

Reduced from 4000x6000 pixel JPEG capture
Reduced from 4000x6000 pixel JPEG capture

And four APSC2 capturtes stitched together using Hugin gives:

View attachment dab2703130e74e2297c7881c2a960b1c.jpg
The full-resolution JPEG stitch

To save bandwidth, here are a couple more stitched results, but scaled down:

Reduced resolution stitch
Reduced resolution stitch

Reduced resolution stitch
Reduced resolution stitch

Again, those are 58MP images from within the image circle of a designed-for-135-format-film Vivitar 135mm f/2.8 -- and all of those happen to have been shot wide open.

The adapter itself is:

aa193a51654846218d854ce9677447aa.jpg

And here it is in use:

f9749dbdb3074d5bbafde7bd102395a8.jpg

Note that this adapter doesn't have a tripod socket (if it did, it would bump into the camera grip when rotating), so your lens or adapter to M mount needs one. Some lenses and adapters have a "tripod foot" for mounting, but this FD-to-M and Vivitar lens didn't, so I used Parametric Lens Mount Ring to print one. There are lots of commercially available alternatives for under $25, but lens diameters vary, and this was easy to customize for a good fit on my Vivitar.

Complete instructions on making and using APSC2 are posted at Thingiverse: APSC2 (APS-C Squared) Rotate-and-Stitch Adapter .

It would be fairly easy to make a version for a Fuji APS-C body. However, Canon EF-M mount is a bit of a problem because Canon's sensor is undersized -- there would be gaps in the stitch. Canon EF-M sensors are 22.3x14.9, so maximum area covered would be 29.8x29.8mm with perfect alignment, and realistically more like 29x29mm... which would feel similar, but definitely requires a different rotation offset.
 
Hi ProfHankD,

I'm finally getting around to following this up after this recent thread that you kindly added you knowledge to.... https://www.dpreview.com/forums/thread/4614646#forum-post-65701757

I've been experimenting with the 4 shots rotated arrangement (without the rotating adaptor - just carefully handheld) to get a feel for the whole process, even if not the use of the whole image circle, and am keen to move onwards.

With your APSC2 adaptor being M-mount to E-mount and my needs being Pentax K to Fuji X, I'm thinking that I'd like to borrow the idea and do a cut and shut style modification to an existing K to X adaptor. Simply slice the adaptor midway along it's length somewhere and fit a rotator borrowed from somewhere else in the middle, obviously maintaining the FFD exactly and applying the required geometry in the rotator.

How do you feel about me doing that? I don't know if this rotated stitch method is your idea or someone else's, however it came to me through you so I'm checking with you if I'm jumping on your (or anyone else's) toes by borrowing the idea and making my own hardware to suit?

Cheers and thanks, Michael.
 
Hi ProfHankD,

I'm finally getting around to following this up after this recent thread that you kindly added you knowledge to.... https://www.dpreview.com/forums/thread/4614646#forum-post-65701757

I've been experimenting with the 4 shots rotated arrangement (without the rotating adaptor - just carefully handheld) to get a feel for the whole process, even if not the use of the whole image circle, and am keen to move onwards.

With your APSC2 adaptor being M-mount to E-mount and my needs being Pentax K to Fuji X, I'm thinking that I'd like to borrow the idea and do a cut and shut style modification to an existing K to X adaptor. Simply slice the adaptor midway along it's length somewhere and fit a rotator borrowed from somewhere else in the middle, obviously maintaining the FFD exactly and applying the required geometry in the rotator.
That's a lot more work than 3D printing, but sure.
How do you feel about me doing that? I don't know if this rotated stitch method is your idea or someone else's, however it came to me through you so I'm checking with you if I'm jumping on your (or anyone else's) toes by borrowing the idea and making my own hardware to suit?
This particular implementation is my idea, but there have always been people doing camera motions of all kinds to stitch images... certainly long predating digital imaging. There is also the commercial FF rotator RhinoCam Vertex for medium-format lenses. In short, nobody should have a valid objection even if you commercially marketed something.
 
Thank you.

Yes, I'm sure it is more work to do as I suggest, it is familiar and accessible work to me though, which I can't say the same about 3D Printing.

I greatly appreciate your generosity with this, even though I'm not (at this stage at least!) printing from your files, the generosity remains regardless.

It could take me a while to pull all this together but I'll be sure to let you know if and when I get a functioning unit.

Cheers!
 

Keyboard shortcuts

Back
Top