There was a retired physicist that refurbished an old aerial reconnaissance camera and painstakingly built a hybird film/digital workflow that resulted in claimed resolution equivalent to several gigapixels. I can't find his original site anywhere, but here's a Popular Mechanics write up of it:
It's too bad because the original site had lots of technical detail and was a good lesson in understanding resolution limiting factors at each step of the image chain (e.g. effect of air density changes on sharpness of distant objects when shooting landscapes).
Update: hooray for archive.org. Check it out here:
For the DSLR folks, the 16x20 Ambrotypes he's making are like a ~100+ MP image. If he had the lens quality and precision focal plane / standard mounting, he could project this onto a billboard at an unheard-of resolution.
I shoot large (4x5) and medium format film. I think that the 'resolution' is complicated by the entire workflow. He's using a very old lens with a very imprecise camera and imprecise film. The images should probably described as 'smooth' rather than necessarily high resolution. The old lens provides extraordinary sharpness in the center, but is soft on the edges.
That said, even with 6cmx6cm film negatives I feel like I often cannot really use the detail available. That is, unless I'm printing very large photos in a film-only workflow I don't really see all the detail. My scanner seems to be limited by the ability to keep the film flat rather than the actual scanning sensor. I don't usually bother scanning higher than 50MP.
This is absolutely the right answer. Too often, internet arguments about the merits of various formats devolve to more film/sensor == more resolution, when nothing could be further from the truth.
Case in point: I once mounted a 150mm large format lens into a cardboard box and taped some Provia 100 inside the box opposite the lens. The resulting image takes up a huge amount of film (roughly 6cm x 17cm), but despite having a decent lens, the resulting image isn't all that great... of course, a cardboard box can't keep the lens square to the box, sucks at keeping the film flat, and even if you managed to get a perfect exposure it's very hard to keep a large negative (positive in this case) flat enough to get a decent dry scan.
Wet mounting will only help if the scanner focus can be optimized. I've got a Epson V750, and _if_ everything is just right, you can get a very good scan, but still noticeably short of what a drum scan will give you.
When I print my MF or 4x5 images, I don't really worry about 'all the detail', since as you say, you'd need a pretty big print to really see them. I'm just happy if they look nice and sharp on the wall.
I have to give mad props to the guys still doing collodion--that is way above my pay grade, as they say. And I don't want to carry glass plates on my bike. I try to limit the toxicness of the chemistry I have in the house as much as can, but I still like doing the lab work--it is satisfying to me in a way that no digital process is ever going to be.
You might be amazed; printing above the industry-standard 300ppi can give you a picture you can never get close enough too, but even at more reasonable viewing distances you can tell the difference in acutance. 480 ppi is a good value for the current crop of Epsons (you'd need a RIP and a lot of farting around to do better, but you can get up to 720). It's sort of like the difference between a contact print and an optical print at the same size; there's no more detail as such at a typical viewing distance, but the edges of the details you can see are sharper.
As for the scanner, if you don't really need a car, you can always get a Flextight. (They really are that good, but they really are too expensive nonetheless.)
I'm aware of the affect of dpi on the print. Sure, I suppose I could also get a second-hand drum scanner, too; sometimes you can get them for cheap. A Flextight is nice but it still a CCD scanner, it is not going to match a PMT.
You can get 80 megapixel digital backs for Hasselblad cameras, where you actually get insane lens quality and a uber precise focal plane, so the resolution is not unheard-of. But the Ambrotype technique is really really cool.
Indeed, and other digital medium format cameras such as the Phase One are in the 100MP range. Of course, they're impractical for a lot of reasons - the $55k price alone puts them out of reach for most hobbyists.
Canon's 5Ds has been viewed as a game changer especially among aerial photographers. these guys used to convert six-seater planes and equip them with a catapult that pushes the camera back to get rid of motion blur during exposure time. In that industry, purchase price of these cameras wasn't really significant.
Now that canon is pushing 50-160 MP on 36x24, digital medium format starts crumbling, and the idea of a "surveying plane" dies along with it. Drones lack the coverage, but the combination of unmanned flight and camera prices that dropped an order of magnitude creates a textbook disruption.
not really, for filming, yes. I was there when it exploded onto the advertising stage. But for stills not really. It was an alright camera, but there were more sensitive cameras even from canon. The nikon d700 was easier to use, and I think it was faster with better optical res (I could be wrong on the last bit)
For high end surveying the 5d sucked because it has rolling shutter(and no finegrained speed control), so you couldn't use it in film mode to get 3d imagery. (RED suffers from this as well, along with being blind to colour, and having a rabid fanbase.)
But then there was a trade off, one time resolution vs speed of acquisition. Film for large plate photography has better dynamic range even now. But it is slower and depending on the stock has less optical resolution.
Most survey planes are all digital. For the kind of thing that's required its just not practical to have 3 miles of 70mm film.
For peering out the side of a helicopter, just choose any high end dSLR, unless you want night shots, then you can't really use a digital medium format.
I have the 5DsR, which is the same as the 5Ds but with an anti-aliasing cancellation filter. I can confirm that at the pixel density on this camera, it is VERY challenging to get perfectly sharp photos if your shutter speeds drop too low. Pixel-level blur is a real problem on these cameras. Conversely, on medium format, you have a little bit more leeway for tripod shake as the individual pixels are larger.
I have no experience with analog photography, So I'm not sure whether these same considerations are present.
However it depends on the solution, process, timing, temperature and skill of the operator.
Wet plate looks amazing, because the depth of field is much smaller, and the colours that the process is sensitive to is very tight. (I think its blue/UV but I could be wrong. I also suspect its very dependent on the process.)
Decent 35mm film is 8 mp (in movies at least) 70mm (IMAX) is about 32 mp (the 80mm panavision on hateful 8 is possibly more, depending on the film stock)
so scaling that up, nievely you can see its possibly much much more. But, modern film is light years removed from wet plate.
Depth of field is a function of the sensor size, the size of the aperture and focal length.
Film tends to have bigger image area, which means longer focal length, which gives shallower depth of field. (the bigger the film, the more resolution you have also)
Because large CMOS/CCD sensors are expensive most digital sensors are smaller than film. most dSLRs are APS-C sized, which is notable smaller than 35mm or "full frame" which is again smaller than medium format or 120mm those digital cameras are still notably expensive ($45k+) and less sensitive compared to film. 35mm DSLRs destroy their film counterparts now. Especially the nikon d800/600 and the canon equivalents.
It is, but when you take cropping into account it's like this: If I shoot 2 photos, one with a 50mm lens at f/2 on 35mm film, and one with a 35mm lens on APS-C, I need f/1.4 to achieve the same depth of field, because everything is smaller. The image will however be brighter, because I shot at f/1.4 (assuming ISO and shutter speed are constant).
It takes less light to illuminate a smaller sensor using a smaller lens (because the image projected on the sensor is smaller), so I need to open the aperture less to get the same f-stop number, but I have achieved less bokeh (more depth of field) at what is the equivalent f-stop of a larger lens in doing so. The bigger you need to blow up the image to fill the sensor, the dimmer it gets, so more open aperture to get the same f-stop, but your DoF is shrinking when you do so.
It took me a long time to wrap my head around it this far, let me know if I'm making sense. I'm rambling a bit because I've not had to put this into words before, and could be way off base.
Large format photography delivers amazing detail. Normally you'll see folks making their own plates past a certain point, but this DIY 14x32-inch camera on the Make Magazine site uses large format X-ray film, which you can get for cheap, since hospitals have been converting to digital equipment: http://makezine.com/2011/05/17/gigantic-diy-ultra-large-form...
http://www.popsci.com/scitech/article/2005-10/sharpest-image
It's too bad because the original site had lots of technical detail and was a good lesson in understanding resolution limiting factors at each step of the image chain (e.g. effect of air density changes on sharpness of distant objects when shooting landscapes).
Update: hooray for archive.org. Check it out here:
https://web.archive.org/web/20060218113103/http://www.gigapx...