Over the weekend, PetaPixel published an article encouraging Fujifilm to make a digital XPan camera. Their idea was to make an X100VI-like camera with two 40mp sensors side-by-side, for 80mp 65:24 (XPan) aspect ratio pictures. If Fujifilm ever did that, I would buy it no questions asked. Within the first minute of the preorder opening, I’d purchase it. That would be a dream camera. But this isn’t Fujifilm’s only—or maybe even best—option, should they desire to pursue a digital XPan camera.
For those who don’t know, XPan was a collaborative project by Hasselblad and Fujifilm beginning back in the late 1990’s. The cameras used two frames of 35mm film to make a panoramic picture. While only moderately successful at the time, the cameras are now much beloved with a cult-like following. A digital version would certainly be niche, but it would also become iconic very quickly.
First, Fujifilm doesn’t necessarily need to make a dedicated XPan camera. The GFX cameras already have the XPan aspect ratio, and you could use the GFX100RF as an XPan-like camera, producing 50mp panoramic pictures. This is probably the closest you can get right now with Fujifilm, and it really is close-enough. Someone let me use their GFX100RF, and I loved using it for XPan pictures. Fujifilm could (and should!) add the XPan ratio to all of their 40mp cameras, which would produce 22mp panoramic pictures. There’s no good reason why I can’t make in-camera XPan pictures with my X100VI, X-T5, or X-E5, and I would if I could (unfortunately, it has to be cropped after-the-fact using software).
The Hasselblad and Fujifilm XPan cameras were interchangeable-lens models, so for a truly accurate XPan experience you need an interchangeable-lens camera. I would be surprised if Fujifilm ever made an interchangeable-lens XPan camera, but if they did, one option would be to use GF lenses. Perhaps make a version of the next X-Pro with a wide sensor and GF mount. More likely, in order to do a wide sensor (or two sensors side-by-side), Fujifilm would use a fixed-lens body, which was PetaPixel’s idea. If the WCL and TCL lenses also work with an X100-like XPan camera, I could see this as being a good compromise (I would be mildly surprised if the conversion lenses actually work for it, though).
There is another option. It’s not perfect, but it would certainly simplify everything. This could be pretty easy for Fujifilm to do. It’s not as exciting, but it would work. Fujifilm could simply do anamorphic de-squeeze in-camera. For example, Fujifilm could make an XPan (1.8x) anamorphic lens that screws onto the future X100VII (like the WCL or TCL does), and the camera would then de-squeeze it, and show the wide image in the LCD or EVF. This would produce a 40mp XPan image. If Fujifilm wanted to do this for the interchangeable-lens models, they would need to make 1.8x anamorphic lenses (no such lens exists for X-mount). This, of course, comes with some negative side effects, namely a reduction in sharpness and “stretched” bokeh.
There isn’t a perfect solution. For those wanting a digital XPan camera right now, the GFX100RF is your best bet. If Fujifilm were to ever make a digital XPan camera, it would likely be a fixed-lens model; however, they could use GF lenses on an interchangeable-lens camera. An anamorphic lens with built-in de-squeeze is another option that would be easier to implement, but has downsides. What Fujifilm could and should do more than anything else is simply add the XPan aspect ratio to the 40mp models via a firmware update, and also include it on all sixth-generation cameras. With that said, if Fujifilm ever did make a dedicated digital XPan camera, I would be right at the front of the line to buy it.
See also:
10 Frames in New Mexico — Fujifilm GFX100S II XPan Photographs
XPan III — The Digital Panoramic Camera That Should Exist
That is essentially how I have set up one of my GFX50R cameras, using the 24:65 ratio panorama, three Zeiss manual focus FF M-mount lenses in 18mm, 25mm and 50mm to nearly match the original lens coverages on the Fuji/Hasselblad camera. Now I am only getting about 25mb out of that combination, but that works for me. If you had a GFX100 you would get 50mb, and the upcoming GFX180R would get 90. Without a special camera or lenses.
I wish that Fujifilm would make a 100mp followup to the 50R. Seems like it would be a winner.
Nah. I want the 180mp version. Been waiting this long, should get the next best thing.
For XPan, a maybe someday 180mp version would be best, especially for very large prints. I’m pretty satisfied with 100mp GFX, couldn’t imagine the file sizes for the full 180mp images from a higher MP sensor.
As a side note Fujifilm made a panoramic camera that took 6 x 17cm images on 120 film. (Four frames per roll!) As far as I know Linhof was the only other manufacturer to make a camera in this format.
Although not in the same league I had a Fujifilm DL Super Mini compact which had a panoramic mode activated with a small lever that physically masked the viewfinder and film.
Many cameras of course have a panoramic mode which is supposed to stitch several frames together as you pan across. It’s tricky to get the knack of doing it right, or at least I find it so, and if there is anything moving it can lead to curious “glitch in the matrix” effects.
Yeah, the Fujica GX617, right? I saw one in person not very long ago. It was really large.
Bought a Panasonic S9 today for that very fact with an 18-40 lens for a very reasonable price.Still if Fuji got their act together I would be number 2 in line.Go on on then,take me money….
One smart thing that Panasonic did for the S9 was including the XPan ratio; however, 13mp (after the crop) is borderline for printing a decent-sized panorama. It makes no sense that Fujifilm doesn’t offer it on the 40mp cameras, which is more ideal.
I would be second in line Ritchie and if it took selling a lot of my gear to get it I would… I love the XPAN ratio.
Awesome, we’ll stand in line together 😎
I’m the TX-1/XPan buzzkill. I think there’s ZERO reason to build a camera with an XPan shaped sensor.
Or maybe I’m the Pentax 67 superfan. I’ll stick the TL,DR up front: if you had the choice between an XPan or a Pentax 67, with unlimited film and the same size of camera… you ought to build the Pentax 67 instead. With digital, it’s wiser to make the sensor use as much of the image circle as possible and crop.
First of all, I bet you probably can’t just put two small sensors next to each other and call it good. If you could, folks would have been doing that with full-frame sensors ages ago, just for quality control reasons (much less sensor area wasted to flaws in manufacturing). Having any gap in there is going to ruin images.
Second, XPan made sense because it used film, and film size is pretty fixed. You can use some length of 135 film, or some length of 120 film. 135 film is the easiest to buy, the easiest to have developed. Just practically, it made a lot of sense. Wasting film is a real issue, so cropping a 6×7 medium format image has a significant cost. Plus, the camera bodies are meaningfully bigger. With digital, the waste not an issue–just make the sensor size you need. Meanwhile, the camera body size isn’t going to be meaningfully impacted.
The key is that the biggest limiting factor will be the image circle of the lens. That’s what will drive the size of the camera. The lens size is going to be what it needs to be, so why not use as much of the image circle as you can? An XPan shaped sensor uses a tiny SLIVER of the image circle. Making a camera with an image circle that covers just the 65:24 portion of a GFX 100 sensor, you’ll get a 50mp image for 65:24 sensor, but a 42mp image if you’d gone with a 16:9 shaped sensor. Tiny loss. HOWEVER. The 16:9 camera has close to a 60mp image if shooting 3:2, and the XPan shaped sensor has 28mp or so. For a less than 10% drop in resolution for XPan shaped images, you gain MORE THAN TWICE the resolution of 3:2 images. It just doesn’t make any mathematical sense to build an XPan shaped sensor when you can just crop and get more quality in essentially every format. Getting a sensor a bit wider than 3:2 is a great boost, but any sensor aspect ratio wider than 2:1 makes so poor use of the image circle that it’s just a mistake.
That said… I firmly believe that a Fuji optimized for XPan has a lot of non-sensor-shape optimizations that make sense. A wider aspect ratio viewfinder, for example, is really going to make things feel nicer to use.
The Nikon D810 has two sensors side by side. It is so well done that it is almost impossible to tell. Only in very low light long exposures can you sometimes tell a slight visible difference in exposure between the two sensors.
Fascinating.
Doing a little reading, and it’s a pretty complicated process that can go wrong.
I wouldn’t say that there are zero reasons. If there are people willing to buy, that’s a pretty significant reason, probably the most significant reason of all the possible reasons. I guess if you exclude that number one reason out of the list of reasons, then it might be close to zero (but it’s still probably more than zero). I have no idea how many would line up to buy a dedicated pano digital camera, but at least me and a few others. I’m sure it would be super niche. It might not be a long-enough line to justify the development of it, I have no idea. Some people say there’s zero reason for a monochromatic-only camera when you can just convert a color image, but a couple companies are making them and people are buying them. I see this similarly.
As far as 6×7, it’s pretty close to the GoPro 13’s sensor shape, right? So it certainly can be done. It might be pretty expensive on a medium-format camera, but maybe reasonable-enough on an APS-C model?
Regarding two sensors side-by-side, I have no idea if that’s possible (that was PetaPixel’s idea), but I assume it must be, since there are cameras that use multiple sensors (360 cameras, for example). I do know that Sony can make the sensor to be whatever shape it wants to; whether they would make an XPan-shaped sensor or how much such a thing might cost, I have no idea, and that could kill the project right there. Sony might not be willing to make 5:4 aspect ratio sensor, either, who knows?
That might be right that marketing and buzz might overwhelm practical concerns. Folks probably would get excited of an XPan shaped sensor at the same price and same size as a 16:9 sensor that could crop to XPan with the same resolution, just because of buzz and hype. I find that depressing.
Folks say “I don’t want to waste pixels by cropping” all the while they’re pre-cropping the image circle. In film, it is waste. In digital, it really isn’t. In terms of capacity and manufacturing… just making a camera designed for XPan-style shooting with a taller-than-XPan-ratio sensor makes more sense for anything other than hype.
Put a physical switch on there, market it as a “multiple aspect sensor” like how some Panasonic cameras have been marketed. And optimize the camera for XPan usage. I think having an XPan shaped viewfinder is kind of essential to such a camera, to maximize your experience while shooting. Go all in on such an finder. Make it the height of a 5.7M dot viewfinder, and then go extra wide. Dang, that could be as much 10.4M dots. It would actually be worth it, however.
The camera has to FEEL XPan in use, and the viewfinder is the most important point of contact for that. But the sensor itself? More practical use of the image circle just makes it a better photographic tool.
//
Oh, 6×7 *is* 5:4, actually. One of the quirks of medium format size descriptions is that centimeter values are somewhat approximate, 67 is 56mm by 69.5mm, or about 4×5 aspect ratio–less accurate of a aspect ratio description because it’s more accurate as a film dimension.
The reason I was saying a Pentax 67 is just because 56x69mm is a film dimension that contains the whole of the XPan frame. With unlimited film and developing and scanning, with a viewfinder that suits the image, a taller sensor is better.
I don’t agree with the “waste pixels” argument (like in the PetaPixel article), I think it’s a bit silly. The argument should be a camera that is optimized for the experience, and let the chips fall where they may for such a camera. I think the experience is the critical aspect more so than the sensor itself (although it might dictate the sensor).
Since a 5:4 sensor has been done before (by GoPro via Sony sensors), it’s more likely to be done again than something that has never been made previously.
Thanks for the input!
x100 anamorphic WCL would be sweet, too.
Kind of a yes-and, rather than an either-or.
I think it would be super cool, especially with built-in de-squeeze.