Initial thoughts: Rhinocam Vertex MF simulator adaptor

[DavidMillier]

I recently acquired a Fotodiox Rhinocam Vertex rotary adaptor to adapt Pentax 645 lenses to Sony full frame.

The purpose of the gadget is to make stitching 4 full-frame images into a 46mm x 46mm square medium format frame as painless as possible, and by using a flat stitching approach, to preserve the character of the taking lens.

From my perspective, the theory is it is a much cheaper way to dip a toe in medium format than going down the GFX route. The final frame is a little larger than a square crop from GFX.

I hope my initial impressions might be of interest to anyone thinking of a convenient way to simulate larger than full frame (relatively) cheaply. Read on to find out how I got on.

This is the thing:

How it works

The narrow end mounts on my Sony A7r2 (other mounts available) and the wide end accepts Pentax lenses for the 645 film format (other mounts available).

You can see from the pic that the camera mount is slightly offset from the centre line of the adaptor. This facilitates flat stitching using rotation rather than conventional sideways shifting.

Benefits

The main idea behind this gadget is to make stitching more convenient and reliable than traditional methods.

You take a picture, rotate the camera 90 degrees, take a second shot, rotate another quarter turn and so on for four shots.

I've done plenty of panos by re-pointing the camera and some flat stitching using shift adaptors but for some reason I find the vertex's rotational approach more natural. I can do it automatically without having to carefully rehearse and think through the stitching movements. I don't forget where I've got to in the middle of the process. From this perspective it is an instant success for me.

For framing I'm experimenting with a cheap Helios clip on hotshoe viewfinder; first results suggest very approximate framing!

Stitching process

To my annoyance, my version of Lightroom 6 refuses to stitch these images. Maybe too old?

Fortunately I have PTGui which works fine and is quick. The frustration with LR is that LR outputs the stitch as a linear DNG (which I want), while PTGui only offers Tiffs and Jpegs. Tiffs weigh in at 650MB a frame which is a bit much. I've modified my editing workflow to deal with this by exporting the edited file as a 100% quality jpeg and deleting the tiff and source files. Not ideal, but necessary for me for housekeeping.

Results

The final stitched output pixel count is 2.5x the full frame sensor and supposedly equivalent to a 46x46mm format. A square crop of a 645 film frame in effect. Using my A7Rii, the final output size is 100MP.

First light impressions with the 55mm: I've shot my initial tests at f/11 and f/16 and the frames stitch perfectly, with the final result sharp corner to corner.

First prints

The maximum size print from my P900 is about 16" square with a narrow white border. I've only printed a couple of shots to date but looking at them, they look no different from shots from full frame, aps-c or even m4/3. This actually seems to be the main drawback with high resolution larger formats, real or virtual: An A2 print isn't big enough to reveal the extra resolution.

Availability and cost

Fotodiox is based in the US but I was able to get mine from a German seller through Amazon UK. This avoided complications with taxes and customs often endured with imports. Shipping was a few days. Total cost to me, £325.

I purchased a Pentax 55mm f/2.8 645 lens and a 150mm f/3.5 from ebay for £175 and £110 respectively. Cheap! I already own a Mir 38B 65mm f/3.5 Pentacon 6 mount and I'm awaiting for an adaptor to make its way from war-torn Ukraine. The 35mm 645 lens is more expensive and harder to find. I may get one later.

Although much more expensive than traditional spin stitching (effectively free), I was still able to put together a reasonable outfit of adaptor and 3 lenses for about half the price of the cheapest GFX lens, and of course, I already have the camera.

p.s.

Sharpening question: I printed a lot over the years, but mostly small A4 or smaller, occasional A3. I have no real idea how much to sharpen for A2 prints. Should I really go overboard with sharpening in an attempt to make the prints look different from lower resolution files?



--
Website: http://www.whisperingcat.co.uk/ (2022 - website rebuilt, updated and back in action)
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

 

[JimKasson]

Sharpening question: I printed a lot over the years, but mostly small A4 or smaller, occasional A3. I have no real idea how much to sharpen for A2 prints. Should I really go overboard with sharpening in an attempt to make the prints look different from lower resolution files?

No. I see a couple of dozen images that are oversharpened for every one I see that's under sharpened. Print sharpening depends on the diffusion in the media and the marking engine. It's the same in absolute terms for the same printer and media independent of print size.

 

[DavidMillier]

Sharpening question: I printed a lot over the years, but mostly small A4 or smaller, occasional A3. I have no real idea how much to sharpen for A2 prints. Should I really go overboard with sharpening in an attempt to make the prints look different from lower resolution files?

No. I see a couple of dozen images that are oversharpened for every one I see that's under sharpened. Print sharpening depends on the diffusion in the media and the marking engine. It's the same in absolute terms for the same printer and media independent of print size.

Thanks, Jim.

What markers do you use to identify over-sharpening? For example, can a print be considered over-sharpened if there are no visible sharpening haloing around edges under the closest inspection? I'm also unsure about different sharpening methods. The software I use (darktable) recommends not using their USM at all and instead sharpen using their local contrast module. But they also offer a new module called "diffuse or sharpen" that I use but I have no idea what it is - it seems to do a multitude of things using methods I don't understand: https://docs.darktable.org/usermanual/3.8/en/module-reference/processing-modules/diffuse/

 

[Rob de Loe]

David, on the sharpening question, it might be worth your time to read through this thread: https://www.fredmiranda.com/forum/topic/1770487 A member (Tim) posted a RAW file and asked people to have at it with different sharpening approaches.

The takeaway for me is that some people like way too much sharpening. High resolution samples were posted by people who participated. Some of the early ones in particular seemed extremely over-sharpened to me.

 

[Rob de Loe]

Thanks for posting your thoughts on this tool by the way.

Michael and Paul are also using versions of this, but for GFX cameras, and have started some conversations. Clearly there's interested in bigger sensors. I'm on the fence about the benefits, but to keep the conversation going I put something up today for interested people to consider.

One thing I have learned as I've played around with this approach is how essential it is that the shift device is absolutely stable. The camera I built for my day-to-day work turns out not to be suitable for flat stitching because the shift movement is not perfectly linear; that's not an issue with single frames, but it messes up flat stitching. My Toyo is perfectly stable during shifts, but because the standards also swing, it's necessary to check and re-check that they're parallel; it's easy for them to not be perfectly parallel, so that introduces a failure point.

This is where shift adapters and your Rhinocam win the day: they're made to be stable along the shift axis.

 

[ProfHankD]

If you want something even simpler and much cheaper, I've created some 3D-printed adapters that allow getting more of the area covered by any lens adaptable to Leica M mount. For full frame cameras, it's Budgie (which now includes an Arcs Swiss compatible mount):


Sony A7II on Arca Swiss compatible Budgie with Minolta MD 50mm f/1.7 lens

You make 3 captures by sliding the camera on Budgie behind your lens to capture a 48x36mm stitched image... or just 2 captures if they are absolutely perfectly aligned. That gives an approximately 60mm diagonal, which is a tad more than most FF lenses can handle wide open at infinity, but you'd be surprised how many lenses designed for 43mm diagonal of FF can actually cover it well under ideal circumstances. Of course, 48x36mm is noticibly larger than the 44x33mm sensor in GFX, etc.

If you have an APS-C camera, my alternative is a bit more awkward (and more like the Rhinocam Vertex approach): APSC2 (APS-C Squared). This also takes any lens adapted to M mount and allows capturing 4 images offset-rotating the camera behind the lens:


The 3D-printed APS-C Squared M to E adapter.

Realistically, you get a 29x29mm stitched area (although it would be 29.8x29.8mm if you kept everything perfectly aligned while rotating).

The main annoyance with APSC2 is that you need to support everything by the lens while rotating and, unlike Budgie, the adapter doesn't help you with that by incorporating a tripod mount (it couldn't and still allow the rotations with an M mount). Thus, you'll need to use a separate lens tripod mount, like the 3D-printed one I used here:



Anyway, all the above are open source designs that can be 3D printed for a total of a few $ material cost. I also recommend using the open source program Hugin for stitching on your computer (I didn't write Hugin, but am a happy user of it).

 

[DavidMillier]

David, on the sharpening question, it might be worth your time to read through this thread: https://www.fredmiranda.com/forum/topic/1770487 A member (Tim) posted a RAW file and asked people to have at it with different sharpening approaches.

The takeaway for me is that some people like way too much sharpening. High resolution samples were posted by people who participated. Some of the early ones in particular seemed extremely over-sharpened to me.

There certainly are some crunchy preferences in there! Good debate, though.

The differences seemed most marked when viewing the full size file at 1:1. On my 32" monitor things smoothed a lot when viewing at about 15" high.

Any comparison between that thread and my test images is complicated by lighting - I shot under softer, less directional light on a duller day and using a duller (tonally) subject. Also I'm printing on smooth matte paper with lower contrast and more ink spreading. When I size my image on the screen to match the size of the print (15% view), the print just generally looks duller and softer compared to the screen, especially under nighttime room lighting. And my test subject (gothic style chapel in the local cemetery) is not the usual kind of simplified, minimalist subject I usually favour and I get confused by there being so much in focus messy distracting detail.

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

 

[Rob de Loe]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

 

[JimKasson]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

I like that.

Let's keep in mind that there are three kinds of sharpening.

• Input sharpening is performed to correct for blurring produced by the camera, lens and sensor. There are ways to make input sharpening a fairly exact science.
• Artistic sharpening is performed to make the image look like the photographer wants it to look.
• Output sharpening is performed to correct for blurring produced by the output device.

I think that most of the oversharpening occurs during the artistic sharpening phase, or at the hands of the people who don't make the distinction between the three kinds of sharpening.

 

[JimKasson]

J

What markers do you use to identify over-sharpening?

If the image looks sharper than the world looks. I've been doing digital photography since 1989, so I've had a lot of experience, and I am fairly sensitive to how much sharpening is unrealistic.

Power tools are dangerous - the last word

When we worked in the darkroom, the tools we had for image manipulation were pretty crude by today's standards. With silver-based B&W, we could crop, change size, lighten, darken, change contrast, dodge, burn, bleach (like dodging, but affecting only the lighter areas). Sharpening required...

blog.kasson.com

For example, can a print be considered over-sharpened if there are no visible sharpening haloing around edges under the closest inspection?

Yes.

I'm also unsure about different sharpening methods. The software I use (darktable) recommends not using their USM at all and instead sharpen using their local contrast module. But they also offer a new module called "diffuse or sharpen" that I use but I have no idea what it is - it seems to do a multitude of things using methods I don't understand: https://docs.darktable.org/usermanual/3.8/en/module-reference/processing-modules/diffuse/

I've never used darktable.

I have stopped sharpening my scanned negatives at all. They look fine to me that way.

--

the last word the last word - Photography meets digital computer technology. Photography wins -- most of the time.

Photography meets digital computer technology. Photography wins -- most of the time.

blog.kasson.com

 

[DavidMillier]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

Nice one

It's difficult to compare my current limited set of test images because they weren't shot with remotely similar lighting but I noticed that least sharpened image in the thread wasn't yours, but the shot processed in darktable, my own software of choice. The poster noted that darktable dev's recommend against using their own USM sharpening tool (darktable has been updated a lot from its origins but they maintain deprecated tools for backwards compatibility with older images - one of the several reasons why it seems so complicated compared to Lightroom).

 

[Rob de Loe]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

Nice one

It's difficult to compare my current limited set of test images because they weren't shot with remotely similar lighting but I noticed that least sharpened image in the thread wasn't yours, but the shot processed in darktable, my own software of choice. The poster noted that darktable dev's recommend against using their own USM sharpening tool (darktable has been updated a lot from its origins but they maintain deprecated tools for backwards compatibility with older images - one of the several reasons why it seems so complicated compared to Lightroom).

It's definitely a moving target. I have some lenses that need very little sharpening, and some that need more. Some subjects do better with more, some papers need less and some more, and so it goes.

 

[Rob de Loe]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

I like that.

Let's keep in mind that there are three kinds of sharpening.

• Input sharpening is performed to correct for blurring produced by the camera, lens and sensor. There are ways to make input sharpening a fairly exact science.
• Artistic sharpening is performed to make the image look like the photographer wants it to look.
• Output sharpening is performed to correct for blurring produced by the output device.

I think that most of the oversharpening occurs during the artistic sharpening phase, or at the hands of the people who don't make the distinction between the three kinds of sharpening.

Yes indeed.

To add to this, I think one can become a bit path-dependent. Maybe a tool like Topaz worked well for one image -- so it gets used on every image, with every lens and camera (whether it's needed or not).

It's also possible to be carried along with the crowd you hang out with (whether that's a photo club, or the Instagram zeitgeist). This is where seeing work from all kinds of people is helpful. One can discover that there are other ways, other standards, other tastes, etc.

 

[DavidMillier]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

I like that.

Let's keep in mind that there are three kinds of sharpening.

• Input sharpening is performed to correct for blurring produced by the camera, lens and sensor. There are ways to make input sharpening a fairly exact science.
• Artistic sharpening is performed to make the image look like the photographer wants it to look.
• Output sharpening is performed to correct for blurring produced by the output device.

This is the Photokit team's (Bruce Fraser?) sharpening philosophy, isn't it?

I think that most of the oversharpening occurs during the artistic sharpening phase, or at the hands of the people who don't make the distinction between the three kinds of sharpening.

I'm on a journey to change my editing habits (which were more or less randomly arrived at) more towards Bruce Percy's methods. He counsels avoiding editing using the usual global tools in favour of very carefully considered and deliberate selective local edits. He makes heavy use of masks to limit the scope of his curves edits so they become very finely applied dodge and burn to delicately alter tonal relationships and to draw the eye where he wants it to go. No crude Lightroom-style global adjustments for him. darktable is powerful enough to be a parametric raw editor equivalent of photoshop for these purposes, including selectively applying sharpening.

In a similar vein, Keith Cooper of Northlight images has some fascinating guides to using shift lenses for subtle artistic effect, not just as correction tools. I found his guide to using diagonal shifts to re-point cloud streaks to act as leading lines eye opening. There is still lots to learn about using technical stuff for artistic purposes even after 45 years of photography

The trick of course, is to turn appreciation into an effective practical skill...

 

[Rob de Loe]

In a similar vein, Keith Cooper of Northlight images has some fascinating guides to using shift lenses for subtle artistic effect, not just as correction tools. I found his guide to using diagonal shifts to re-point cloud streaks to act as leading lines eye opening. There is still lots to learn about using technical stuff for artistic purposes even after 45 years of photography

The trick of course, is to turn appreciation into an effective practical skill...

I just read that article of Keith's too. Even though I've been using movements for decades, like you, this hadn't occurred to me! I do use tilt selectively, so not just to get more things in focus. But this was a new one.

See this is why I like to hang around here. Lots of great ideas.

 

[ProfHankD]

Clearly sharpness is part artistic preference rather than simple technical right or wrong but it would still be good to have some basic reference points as to when you have definitely got it wrong...

I agree that there's an element of artistic preference. But I think there's also a threshold beyond which it's only artistic preference if the person made it hideous on purpose!

Interestingly, the ones I thought were most over-sharpened were the ones that people ran through special third-party sharpening tools. Those were the crunchiest.

For what it's worth (which is not much), I try to live by what I call the "toupee rule". In other words, like a toupee, if you can see it, it's too much.

I like that.

Let's keep in mind that there are three kinds of sharpening.

• Input sharpening is performed to correct for blurring produced by the camera, lens and sensor. There are ways to make input sharpening a fairly exact science.
• Artistic sharpening is performed to make the image look like the photographer wants it to look.
• Output sharpening is performed to correct for blurring produced by the output device.

I think that most of the oversharpening occurs during the artistic sharpening phase, or at the hands of the people who don't make the distinction between the three kinds of sharpening.

My rule for most such image processing is the "half what you think it should be" rule: process to taste and then 50% average that with the original. ;-)

 

[DavidMillier]

J

What markers do you use to identify over-sharpening?

If the image looks sharper than the world looks. I've been doing digital photography since 1989, so I've had a lot of experience, and I am fairly sensitive to how much sharpening is unrealistic.

https://blog.kasson.com/the-last-word/power-tools-are-dangerous/

For example, can a print be considered over-sharpened if there are no visible sharpening haloing around edges under the closest inspection?

Yes.

I'm also unsure about different sharpening methods. The software I use (darktable) recommends not using their USM at all and instead sharpen using their local contrast module. But they also offer a new module called "diffuse or sharpen" that I use but I have no idea what it is - it seems to do a multitude of things using methods I don't understand: https://docs.darktable.org/usermanual/3.8/en/module-reference/processing-modules/diffuse/

I've never used darktable.

If you feel the urge to read the unbelievably complicated page in the darktable manual link above that blathers on about Fourier partial differential equations and diffusion models and explain the gist in layman's terms, I'd be eternally grateful

I have stopped sharpening my scanned negatives at all. They look fine to me that way.

p.s.

I'd just like to say for the record that your "power tools" article is superbly written. Not just the sentiment, but the construction. Having read it again, I'm left with the feeling that changing a single word would be like vandalising a Shakespearean sonnet. Perhaps you could volunteer your services to the darktable team as the manual technical editor...kidding of course but not entirely...