Making a scanner with a DSLR

Thank you, Peter!

Quote:

---

Originally Posted by Peter De Smidt

By "geometry and mircrogeometery" do you mean size distortions from stitching?

---

Sure. :-) I've just found a better term for that: warping. It happened that I hate warped images. And stitching often produces a lot of warping. Looks like it is not an easy task to fight them.

Quote:

---

"This headache inducing illusion is simply a static image of a green background with a series of blue dots, however when you look at it the whole thing appears to move and warp in a wave-like fashion."

---

http://www.optical-illusionist.com/i...warpeddots.jpg

Quote:

---

Originally Posted by Peter De Smidt

By "profile" do you mean a template for stitching?

---

Not really. There are no perfect lenses and the profile can correct each shot distortions, thus eliminating warping in stitch.

Why I asked about camera mount is because it's very hard to place it exactly to the same position.

Regarding lens profiles, a lens designed for 1x magnification has very little distortion at 1x. My understanding is that a symmetrical lens can perform very well there. I use a Rodagon D 75mm lens optimized for 1x. My limited understanding is that PTgui corrects for lens distortion, but it's better if there's as little as possible already there.

Regarding mounting, the lens system doesn't move when the camera is mounted. I do check focus, but it's rare for there to be an issue.

Rodagon or Magnagon? Magnagon is just great. Perfect for the task. They say that it's Rodagon-D, but I'm not sure.

My inderstanding (limited) of PTgui is that it can use lens profile, if it has that profile in the base.

"Lens system doesn't move when the camera is mounted" - that's what I did not understand at first. You are right. In that case remounting camera makes no harm.

I have both a Rodagon D and a Magnagon. With my examples and in my system, the Rodagon performs better. There are two Rodagons, an f/4 and an f/4.5. The first is optimized for 1x, whereas the second is optimized for 2x, or 1/2x reversed. The Magnagon seems more like the latter.

One my attempt in progress for a few years (now i am changing axes to RPi3 HTTP REST controlled over wifi).
http://12in.cz/fotografie/velka/40_1...izacni-stolice
(prepared for 12x20" maximum film size and 5:1 magnification, light over strobe ELC Pro HD 1000)

Quote:

---

Originally Posted by Peter De Smidt

I have both a Rodagon D and a Magnagon. With my examples and in my system, the Rodagon performs better.

---

Can I ask you when you will time to scan the resolution target with your lenses. At the moment I do not understand big thing: if the lenses outperform the sensor why is the difference there?

For there not to be a difference, the lens would have to outperform the sensor in every way, resolution, contrast, chromatic aberration.....That's not the case. I've already taken those pictures many years ago. Even if I could find them, they would be of very limited use. Comparisons are extremely difficult to do well. One's not done at the same time by the same operator in the same system are unlikely to provide much information, and even then we'd be looking at a very small sample of lenses. Would even 10 samples of a lens be representative of all production? Unlikely. What does make sense is doing your own testing in your system to see which one performs best for you. In my system, that was a Nikon 5x Measuring Microscope lens. But using that lens to photograph a large negative would take way too much time, and while the resolution would be better than any scanner I've worked with, what about errors introduced by stitching all of those frames? What about the cost of all of those shutter actuations? Would there be any more detail available in a print made with that lens compared to, say, a Rodagon D 75mm f/4? In my system the answer was "No," and that's despite giving better performance with a high resolution test target. For me, 1x magnification is the sweet spot. Going higher didn't lead to any better results, and that includes when scanning 35mm Technical Pan shot as carefully as I could manage, i.e. huge tripod, mainly 2d subject, optimum aperture, cable release, mirror lock-up....... If there wasn't a difference there, it's unlikely for there to be one with the larger format film.

Quote:

---

Originally Posted by Peter De Smidt

I've already taken those pictures many years ago.

---

I would not call 5 years as many, but it's me. I've seen those tests. I agree: they do not say much about the difference.

While using ptGui do you examine medium-maximum pixel shift? Still have no idea how the glass influences it. Have to test myself I think.

https://www.dropbox.com/s/ysiy8zg6en...ning.jpg?raw=1

Rodagon 75mm f/4, Nikon D600, 1x magnification, no sharpening.

Thanks! I will post 1x Magnagon tests in Lenses thread. Both are great lenses for sure.