Apaflo Loc: Anchorage, Alaska

It really isn't just how many megalixels, but how many line pairs per millimeter. On a full frame sensor a 36 MP image has higher resolution than a 24 MP image, but if the 24 MP image is on an APS-C cropped sensor it will have better resolution than anything less than about 56 MP on a full frame sensor.

However, if you go with a 24 MP APC-C sensor there is a real problem finding an 18mm macro lens that is designed to have a flat field. It is hard enough to come up with a 35mm flat field lens. And in fact the lens is going to make more difference than the sensor!

If you can find a suitable 18mm lens the Nikon D7100 or D7200 are the highest resolution cameras. Full frame Sony and Nikon cameras have fewer MP than the Canon camera, but their sensors (both made by Sony) are better.

My choice would be a Nikon D810 with a Componon 35 mm enlarging lens.

TriX Loc: Raleigh, NC

It actually IS about how many megapixels, and if the OP's budget would allow the 5DS/r, its 50 MP would provide higher resolution than either the 36 MP Nikon or the Sony (and certainly the 24 MP crop camera) but rather than subvert this thread with the same old argument that has been rehashed multiple times, I'll start a new thread, and we can debate it there if you like.

Chris

Apaflo Loc: Anchorage, Alaska

Chris you haven't been able to grasp it yet, so there is no point in new threads.

Resolution is measured in line pairs per some linear measure, in this case millimeters.

selmslie Loc: Fernandina Beach, FL, USA

A 24 MP DX sensor with an FX lens might be able to do marginally better than a 24 MP FX sensor because the DX sensor uses only part of the lens's image circle but that's not enough to make up the difference.

But it's the total lines per picture height (or effective MP) that matters, not the lp/mm at the sensor.

That's why a 36 MP image will always have a higher total resolution than a 24 MP DX image assuming you use the best available lens.

The sharpest 35mm lenses are between $800 and $1800 as tested on cameras between 36 and 51 MP. The sharpest primes between 18 and 24 mm were all tested on the same full frame cameras.

That pretty well narrows the field down to a used 36 MP body with the best new or used 35 mm lens available to stay within the $2000 budget.

selmslie Loc: Fernandina Beach, FL, USA

You are still assuming a perfect lens. There isn't one, at any price.

Resolution is the combined result from both the sensor and the lens.

The only one here that has not grasped this concept is you.

Apaflo Loc: Anchorage, Alaska

Here's a new twist on that too. I managed to find spec sheets for the Componon 28 and 35mm enlarging lenses. They are essentially DX lenses without enough coverage to use with a full frame sensor.

That leaves using standard 35mm full frame lenses, none of which are macro/flat field. The only answer is to stop down to at least f/16, or even further if possible, to get sharp corners and then use good sharpen techniques to remove diffraction.

To choose a lens, use MTF charts to select for low astigmatism (sagittal and meridonial lines close together), and hopefully with relatively decent contrast and without too much in the way of wavy lines that indicate curvature of field. The right most edge of the graphs indicate what the corners will look like, so they shouldn't be too distinct from what the center of the graph is like. Graphs with the right side either dipping to very low contrast or with sagittal and meridonial lines getting very far apart should be avoided.

It can be a real problem finding suitable MTF charts though. Ideally comparing charts all made at the same f/stop would be best, but many are only available showing a single f/stop. If that is wide open it means very little for this purpose...

It appears as if your Sony combination might just be the only way to fit this into the proposed budget.

selmslie Loc: Fernandina Beach, FL, USA

Any enlarging lens can be used for copying since they are designed to work with flat images on both sides of the lens, much like macro lenses.

You don't need to stop down to f/16. What does that get you? More DOF? The subject is flat.

A 1.5 crop sensor at 24 MP will be diffraction limited at f/8. A full frame sensor at 36 MP* is diffraction limited at f/11 so a full frame enlarging lens would work better. The Schneider 35mm f/4 Componon is designed for full frame.

MTF charts are not going to help much. They are usually plotted for 30 lp/mm and 50 lp/mm.

I don't think that the OP needs to complicate things by attempting some Rube Goldberg setup with an enlarging lens and a special mount.

* At 51 MP, full frame is diffraction limited at f/8.

Apaflo Loc: Anchorage, Alaska

Yes enlarging lenses are supposed to be flat field. Try one of the Belsar 50mm f/3.5 lenses that came with a Beseler 23C enlarger, or an El Omegar that came with Omega enlargers. Then compare to a 50mm f/2.8 El Nikkor lens. The corners are not sharp with the Belsar because, like virtually all inexpensive enlarging lenses, it isn't all that flat! Same with the Omegar. The El Nikkor will have sharp corners (and they originally cost significantly higher too).

Then try a 35mm Schneider Componon, and you'll discover it won't cover a full frame sensor...

Not all enlarging lenses are the same! You can't use just any enlarging lens.

The subject is flat, but the plane of focus is not.

How can "a full frame enlarging lens would work better" in terms of diffraction? The amount of coverage, full frame or otherwise, does not change the diffraction.

The Schneider Componon 35mm f/4 lens is specified with an "Image circle" of 32.5mm. Obviously a 36x24mm full frame sensor will suffer severe vignetting.


Actually most MTF charts do not plot 50 l/mm, though in recent years that has become more common. 10 and 30 l/mm plots are the norm, and are used for example by both Canon and Nikon. 10 l/mm basically shows contrast, and the higher 30 l/mm plots show resolution.

There are many really good sites that explain the value of MTF charts. There is no question that your claim they won't help is not true. Read up on what the charts indicate and you'll learn why everyone interested in camera optics, from Nikon to Roger Cicala at LensRentals.com pays a lot of attention to them. Here are two of the better tutorials, but others are good too.

http://learn.usa.canon.com/resources/articles/2013/reading_MTF_charts.shtml
https://photographylife.com/how-to-read-mtf-charts


He needs valid advice from people with experience. Your comments are intended to be disruptive.

RWR Loc: La Mesa, CA

The OP says, “Most of our documents are in the 17" x 24" size range.” At less than 1:16 (full frame), how necessary is a flat-field lens?

selmslie Loc: Fernandina Beach, FL, USA

I don't have to try anything. I have used the 50mm 2.8 EL for years. Totally adequate for the highest resolution 35 mm I ever used, Kodak Tech Pan (well over 150 lp/mm) printed at 40x60 inches. I have also used my 80 mm Schneider for MF and the 105 mm Nikon for 4x5. But all of these are longer than 35 mm.

Of course it's flat! That's how an enlarger lens works - a flat film and a flat piece of paper. It works both ways. That's what makes it just like a macro lens. You must have never used an enlarger.

Take a look at LENS DIFFRACTION & PHOTOGRAPHY. Page down to "CALCULATING THE DIFFRACTION LIMIT" and click on "show advanced". You will see how format size and megapixels affect diffraction limits.

I guess that rules you out. You have neither the experience nor the understanding.

But all of this BS is moot. The OP does not need the maximum possible resolution. He is not taking images of microfilm or targets. He just needs enough to optimize the image for OCR.

PS: Vignette for the 35 mm goes away by f/8 or f/11.

Apaflo Loc: Anchorage, Alaska

Please read what was stated and respond to the point.

The subject (a document) is flat. The sensor is flat. The lens' plane of focus is not flat

As mentioned, the flatness of field for an enlarger lens is quite variable, with the less expensive models not being particularly good. The result is blurred corners and edges. The only remedy available with a not so flat "flat field" lens is to use small apertures to gain enough Depth of Field to mask the effects of curvature of field.

Your other equally non-sequitur commentary is not worth correcting...

Apaflo Loc: Anchorage, Alaska

As long as a relatively wide angle lens is used and the subject fills the frame, flattness of field is pretty important.

selmslie Loc: Fernandina Beach, FL, USA

Maybe the lenses you recommended (and never used) are not flat. The ones I use are.

I never use them wide open because the exposure time would be too short. Stopping down helps with resolution, vignette and flatness.

My final point is that you are hijacking this thread for your own amusement. That's not helpful to anyone.

Apaflo Loc: Anchorage, Alaska

We can all see who is hijacking and who is helpful...

RWR Loc: La Mesa, CA

I'm thinking that at that distance, a regular 35mm prime should give satisfactory results. I won’t be home til Thursday, else I’d stick a newspaper on the wall and test it. Hopefully someone else can do that before then. Interesting article (once I discount the nonsense!).