One of the most loaded questions I get, often in the form of criticism based on assumption rather than observation, is about the "detail" of the ObscuraFlex digital image compared to an analog large format film image.
While I haven't yet been able to claim that the ObscuraFlex system can beat, or even match, the quality of analog film... the difference between film and an ObscuraFlex image from a diffuser screen at the film plane shot with an iPhone or digital camera... is often far less extreme than people expect.
How to Measure Detail
Systems to measure detail have been around since the birth of photography. Since the 1950s the preferred method has been "line pairs per millimeter" (lp/mm). It's a simple idea, shoot a chart with a series of ever shrinking groups of lines, then look at the result and ask "How tiny can a pair of lines be before they become a blur?"
A higher number is better. It means you can see more lines, aka finer detail.
This is useful because "lines" or "detail" per millimeter is the same whether you're discussing an 8x10 sheet of film or a relatively small 35mm frame. When we talk about the level of detail in large format film being greater than smaller formats, it's simply because it's bigger, there are more millimeters available to reproduce the image.
What Level of Detail is Good?
While film manufacturers rarely publish average lp/mm measurements today, a medium grain black and white negative like Ilford HP5 (used in this test) would be expected to resolve detail as fine as 80-120 lp/mm under highly controlled laboratory conditions, shooting perfect charts, with modern scientific grade lenses, and using microscopic analysis to view the results.
But we aren't talking about laboratory conditions, we are talking about photography, created by photographers who love vintage cameras and lenses.
What we need to determine is the level of detail a typical photographer might expect to achieve when shooting film using standard equipment. In the case of ObscuraFlex we are talking specifically about antique cameras and vintage lenses used by hobbyists, enthusiasts, and fine-art photographers. And like it or not, in the 21st century a large format photographer is more likely to process at home, then scan and ink-jet print their images rather than use a large format enlarger or professional lab.
So that's what I did.
The Experiment
I set up my Graflex Crown Graphic with my best lens, and pointed it at a standard USAF-1951 chart at a scale that would allow me to measure the point of failure. I shot the chart with a sheet of Ilford HP5, and again with a Ricoh GR IV Monochrome digital camera via an ObscuraFlex 4x5 adapter. The Graflex aperture and focus remained untouched; only the capture medium changed.
Let's take a look...
On the left is an HP5 4×5 negative scanned at 3200dpi* with a desktop scanner. The resulting file shows detail down to about 35 lp/mm. Pretty good!
On the right, the same scene exposed using an ObscuraFlex and a Ricoh GR IV Monochrome camera shows detail down to about 21 lp/mm. Also pretty good!
Film wins, but not by as much as I was expecting.
*One could scan the film at a higher resolution, but at 3200dpi this scan is about 185 megapixels, which is far beyond the resolution needed to resolve line pairs down to ~60 lp/mm if the detail were there. If the combination of camera, lens, negative, and scanner was capable of providing more detail, we would see it in the file.
On the other hand, while the Ricoh GR IV Monochrome does an amazing job, it has a hard resolution limit based on its 25.74 megapixel sensor, and within that the image area must be cropped to match the 4x5 negative. The file ends up being 5,016 x 3,965, or a hair under 20 megapixels. Mathematically, the measurement we are seeing might be the highest lp/mm those pixels could ever represent.
So the bottleneck might be the megapixels, not the ObscuraFlex itself.
As I expand the ObscuraFlex Pro line to include higher resolution digital cameras I will be watching closely to determine the true limit of the diffusion screen. The grain structure will also be there, but I am curious where the "detail" bottleneck really is.
Conclusion
Ultimately, neither the film scan or the digital image would look out of place blown up to 3 or 4 feet wide on a gallery wall. Viewed from a few feet away they would appear very similar.
Without a doubt the film image would appear sharper when viewed up close, so I wouldn't recommend hanging a 4ft wide print from an ObscuraFlex next to an Ansel Adams in a museum.
But if you are shooting for fun, or for your online portfolio, or if you're printing images for your wall at home, or even a gallery, say 16in x 12in, I think the results will be surprisingly good, even compared to the same image, from the same camera, shot on film.