If this is the tool to use for evaluating a scanner, then can we also use it to evaluate a "camera scanner"?
With my current camera scanning setup, I'm pretty sure it's not worth looking for much more than around 2,666 ppi (as a function of all the limitations of the system). I can get 2,666 ppi with 12 frames. To get 3,000 ppi takes 20 frames -- so for a modest increase in resolution, a lot more work.
I've been judging by comparing "scans" of actual 4x5 negatives at different resolutions. But if I wanted to be more precise, would this tool do the trick?
As an aside, either Edmunds Optics is ripping off the language they use to describe their much more expensive USAF 1951 glass slide from the Chinese eBay sellers, or the other way around... I suspect the other way around.
This has already been talked about in this thread. Yes, you can use a USAF1951 target with a camera-based scanner. The real problem is correlating the results with real-world film scans. When I developed my DSLR scanner, I tried all sorts of lenses, from an old 55 micro Nikkor to a 5x Nikon Measuring Microscope lens. The higher magnifications gave much better results with an Edmunds chrome-on-glass high resolution USAF1951 target, but I did a comparison using a 35mm Kodak Technical pan negative. The image was shot using a very good prime lens at it's ideal aperture. I used a cable release, mirror lock up, and a 20lbs tripod. This system should've given higher resolution than that achievable with large format film, at least without a NASA-level budget. The result, with my system, shooting at 2x magnification might've given a tiny bit more detail than at 1x....but I couldn't be sure there even was a difference, even at %400 on my screen. Yes, I used lenses optimized for each magnification. So, the added hassle of going to 2x did not make sense, at least with my system, even though the test slide showed clearly better results.
Last edited by Peter De Smidt; 18-Jan-2019 at 15:41.
“You often feel tired, not because you've done too much, but because you've done too little of what sparks a light in you.”
― Alexander Den Heijer, Nothing You Don't Already Know
Yes... of course. https://en.wikipedia.org/wiki/1951_U...ion_test_chart
You need an slide resolving Group 7, or less if your scanner is not a drum. For a drum you may make a contact copy of the glass slide on Adox CMS 20 film, so it can be wet mounted in the cylinder.
If you have a camera scanner the it would be useful to check focus, so you would see: the quantitative effect of a focus shift, and the sweet aperture point, and average best setting. You may even want to stop the lens beyond the sweet point to increase DOF if you have a curled negative, I guess.
But, think anyway that it's really difficult that a LF negative has much information beyond 50 lp/mm, this is Group/Element 5.5
For rolls, some selected shots (non handheld, etc) may clearly benefit from a better scanning.
Thanks Pere and Peter for confirming what I suspected. It makes sense, of course.
As if often the case, the danger is that we create a solution looking for a problem! I already find prints up to 17" wide to be be pleasingly "big" for me. With my current setup I will be able to reliably produce digital version of my 4x5 negs at 2,666 ppi. That would allow for a print of 36"x24" at the Epson "ideal" of 360 ppi. In the unlikely even that I need more -- and the negative can give more -- the occasional drum scan would be more economical.
Really the only actual problem I have is dealing with large areas of low detail. Stitching these is currently possible in my setup only with a lot of manual intervention. Peter's computer-controlled approach seems to have solved that problem nicely, but I'm not prepared to go there yet. Should that ever become a significant barrier to completing a project, I'll probably have to spring for a V750 or V850....
I guess this was about spotting the line groups... not the individual lines.
Human eye may see a bit the individual lines in Group 1.
Group 2 have lines of around 0.05mm thickness. Whe can see a hair that is 0.07mm, but if two hairs are separated by a hair width then we see a single hair.
In Group 6 lines are around 0.005mm thick.
Last edited by Pere Casals; 19-Jan-2019 at 06:18.
Bookmarks