DSLR Scanner: Lenses

[Peter De Smidt]

DIYS (Do It Yourself Scanner—pronounced like ‘dice’)--Lenses Thread

Frank Pertronio started this project by suggesting that someone come up with an affordable and contemporary drum scanner, as there is currently huge gap in price and quality between consumer and professional scanners. Domaz suggested using APS-C sensors and using them to take samples of the film, similar to what Gigapan does with large stitched mosaic images. This lead to talk about making a copy stand scanning system using a dslr, a light source and a movable negative stage. Both horizontal and vertical prototypes have been made, or are in the process of being made.

The original thread (http://www.largeformatphotography.info/forum/showthread.php?84769-Making-a-scanner-with-a-DSLR) has become very long and unwieldy. As a result, I’m creating some new specialized threads for future project development.

The new build threads are:
Camera Supports and Positioning (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=7),
Lenses (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=7),
Negative Stages (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=7),
Light Sources (http://www.largeformatphotography.info/forum/showthread.php?87536-DSLR-Scanner-Light-Sources),
Stitching and Blending of Images (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=7),
Cameras and Camera Control Software (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=7).
Workflow (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=7).


These threads are only for positive contributions to the development in the area in question. The project may not succeed, but we’re going to find that out by trying it. But we are not unkind. As the original thread showed, some people have an overpowering urge to say negative things about the project. I’ve created a thread just for this purpose. Please post your negative comments about the project here (http://www.largeformatphotography.info/forum/newthread.php?do=newthread&f=17).

I would like to thank everyone who makes, or has made, a positive contribution to this project!

I'll be summarizing the posts from the original thread about lenses here soon.

[Peter De Smidt]

Here's a nice test of some lenses of interest: http://www.macrosmuymacros.com/index.php?option=com_content&view=article&id=27&Itemid=40&lang=en

[Peter De Smidt]

Here's the whole 6x7 negative:

http://i955.photobucket.com/albums/ae37/peterdesmidt/Light_House_2nd_Manual.jpg

Just for fun, here's the results for a 4x microscope objective.

http://i955.photobucket.com/albums/ae37/peterdesmidt/ZS_4x-1.jpg

That's the whole field captured downsized to a reasonable web size.

Next we have a screen capture of a small part of the file at 100%:

http://i955.photobucket.com/albums/ae37/peterdesmidt/sc_100.jpg

[rdenney]

What lens did you use for this one, Peter? Was this the Micro-Nikkor with extension tubes to achieve 1:1?

Rick "who hasn't yet experimented with the Canon Compact Macro with the new (to me as of Friday) life-size converter" Denney

[Peter De Smidt]

It's an AmScope PA4X, marked: Plan 4/0.10/160/0.17

The '4' is the lateral magnification. The '0.10' is the numerical aperture. The '160'mm is the tube length, i.e. the distance from the camera sensor to the back of the lens, not counting the threads, and the '0.17'mm is the expected cover glass thickness.

It was this one: http://www.ebay.com/itm/350123872902?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

It's a cheap 4x that the macro folks like.

[Nathan Potter]

Peter, it appears that the magnification factor is about 20X between the 4X5 and the second image. I assumed you've imaged the grain which is what shows in the second shot then simply magnified it more (about 4X) in the third shot.

What is interesting is no sign of vibration in the capture - a key advantage of using flash in high magnification situations. Contrast seems pretty darn good which is notable for an inexpensive microscope objective.

What was your reason for doing this test? There would be hundreds of frames needed to replicate a whole 4X5 image but clearly the result would equal the best drum scanner. So is this an existence theorem?

Nate Potter, Austin TX.

[Peter De Smidt]

Oh, I've had the lens for a little bit, and I just figured out a way to mount it today. The quality will get better when my Prostar flocking material arrives soon. (http://www.fpi-protostar.com/flock.htm). The current extension tube is shiny black on the inside. Using such a lens would be absolutely overkill for 4x5, but it might be useful with 35mm, or perhaps even medium format.

[Peter De Smidt]

I just did a comparison between the 75mm Linos mounted forward and my Nikon 55mm mounted in reverse, both at 1+1. In that configuration the Nikon performs better. I'll try mounting the Linos in reverse, but that will be a bit of a pita.

[rdenney]

I just did a comparison between the 75mm Linos mounted forward and my Nikon 55mm mounted in reverse, both at 1+1. In that configuration the Nikon performs better. I'll try mounting the Linos in reverse, but that will be a bit of a pita.

I was afraid that would be the case. When I compared it to the EL-Nikkor, it wasn't quite as good.

Rick "who had hoped for a chance to work on it today, but ended up grouting bathroom tile" Denney

[Peter De Smidt]

Preliminary tests show that at 1:1 the Linos is better reversed than forward. Whether it's better than the Nikon is hard to judge, and will require more thorough testing.

[buggz]

Here is a site w/ more info about this subject:
<<http://theonlinephotographer.typepad.com/the_online_photographer/2012/01/how-to-scan-film-2.html>>

[Peter De Smidt]

Today I tried a 50mm Componon-S reversed. The results are pretty good. It's hard to compare results, though, as everything needs to be the same, especially magnification, and that's tricky.
For the time being, I'm going to leave off lens testing and look into automation.

[Donald Miller]

I have followed this thread with interest. There is a lot of good information contained here. I have a question that I don't have the answer to. For that reason I am posing it here. I have a Canon 65mm dedicated macro lens. This is the lens that can not be used for normal pictorial use (macro only) and will render at 1X up to 5X magnification. How would this degree of magnification compare to a drum scan? A good educated guess of the comparison will determine if I go forward to design a system to use that lens in a scan capacity with either my 5D MkII or 1DS Mk III. I have a Novoflex Cross Q focusing rail for two axis (X and Z) that I would like to incorporate. My other problem, should I decide to go forward with this, would be the third (y)axis. Thanks for sharing your thoughts on this.

[Peter De Smidt]

Hi Donald,

That lens has a very good reputation. It'll allow you to use live view and focus from your computer, which should be very helpful. Earlier in this thread, I show a close up from a 4x microscope objective. Your Canon lens should be able to just as well. Actually, I expect it to do better, seeing how it's a $30 lens versus a $900 one. (The macro folks say that there's no advantage to microscope lenses until you get above about 6x magnification.) The microscope lens captured better detail than I did with my Cezanne. But at 4-5x, you'll need a lot of frames to cover a negative, especially a large format one!

Detail wise, I expect that we can get very close to the resolution of a drum scan, but that's not the same thing as saying we can get the same quality. As Peter Ramm mentioned in the main thread, the big difference between a drum scanner and what we're doing is flare. What this means, I expect, is that the drum scanner will be able to see more distinct tones than we'll be able to get with our scanner, no matter how careful we are with masking off of spill light. Is the difference enough to matter? I don't know, and it's a subjective judgement in any case.

Ideally, you should get one of your negatives drum scanned so that you can compare. You can then take a close up with a simple dlsr rig of just one exposure to see how it compares. There wouldn'd be a need to do any stitching at that point. If you don't get the quality you need with that one square, you won't be able to get what you need by combining a bunch of them together. If the quality is acceptable, than you have proof of concept that the endeavor would be worth while.

I'm not sure how you plan to use the Cross Q. If you explain a bit maybe I can make some suggestions.

[Nathan Potter]

I have followed this thread with interest. There is a lot of good information contained here. I have a question that I don't have the answer to. For that reason I am posing it here. I have a Canon 65mm dedicated macro lens. This is the lens that can not be used for normal pictorial use (macro only) and will render at 1X up to 5X magnification. How would this degree of magnification compare to a drum scan? A good educated guess of the comparison will determine if I go forward to design a system to use that lens in a scan capacity with either my 5D MkII or 1DS Mk III. I have a Novoflex Cross Q focusing rail for two axis (X and Z) that I would like to incorporate. My other problem, should I decide to go forward with this, would be the third (y)axis. Thanks for sharing your thoughts on this.

Best to wait for some results from others here who are actually building a setup and may use that very camera. But the optical and capture hardware you mention is pretty well known. The pixel pitch of the 5DMKII is about 6.8 um and I would imagine that your macro lens has resolution with reasonable contrast in the 50 to 75 lp/mm. (10 to 7 um airy disk pattern) range.

Given these figures you can assume that you should get a scan resolution of about 2500 to 3500 SPI. This is of course subject to the whims of the digital processing within the camera as well as the illumination source and light scatter reduction scheme. Mechanical issues particularly such as focus maintenance will be important.

To achieve the high resolution that is endemic with the setup you mention the focus maintenance is particularly onerous. At f/2.8 for instance and a COC of 7 um (compatible with the 5DMKII pixel pitch) the depth of focus would be about 40 um (a bit under 2 mils), and needed over 5 inches.

But in fact this far exceeds the resolution of the image you are likely to get on 4X5 film. What you will be sampling will be the smooth tonal values from overlapping circles of confusion with in the film emulsion, which may be of considerable benefit in the replication process. Done carefully I suspect you will approach the quality of some drum scanners. Keep checking here.

Nate Potter, Austin TX.

[Donald Miller]

Thanks Peter and Nathan. Much appreciated.

[marfa boomboom tx]

It still breathes --

for those concerned about lens distortions:


adobe (2010)
http://download.macromedia.com/pub/labs/lensprofile_creator/lensprofile_creator_cameramodel.pdf

Others:


http://forums.adobe.com/message/3382077


http://developer.berlios.de/forum/forum.php?forum_id=27253

http://www.agisoft.ru/products/lens/


http://lensfun.berlios.de/lens-calibration/lens-calibration.html

http://publiclaboratory.org/wiki/mapknitter-correcting-lens-distortion

======
Just a reference note, check out LCC concept used by most high Res MFDB users --

http://help.phaseone.com/en/CO6/Editing-photos/Lens-Correction.aspx


&off

[Peter De Smidt]

Thanks for the links!

The ones that I've check into have you take a picture of a fairly large scale target. Do the lens characteristics change significantly when moving to 1:1 magnification?

[marfa boomboom tx]

Thanks for the links!

The ones that I've check into have you take a picture of a fairly large scale target. Do the lens characteristics change significantly when moving to 1:1 magnification?

What most of these discussions are over is the building and maintenance of a database of lens characteristics. The reason for the target size is to get infinity focus, since that is the more likely use of the data. A picture of something distant. It also permits the target to be less precise (no depth, just x & y).

The LCC plate (1/8 thick white plastic 3" sq) works quite well at making corrections in the land of $40,000 setups.

I submitted these links for the folks coming along in 2 years, who will have different equipment and needs. So they can laugh at our rustic methods of today.

The lens distortion database, in the above links, is being used to bring images together from a wide range of original lens images. Meaning, 3 people go out and shoot, using very different cameras, yet the map folks have to combine the work into one image.

The theory of correcting lenses is the same for them and for us. You could create a database and then shoot image after image, or (what I do) shoot an LCC (with rulings ) at the start of the scan cycle, then shoot the scan-shots.

The LCC is applied to each shot in the set(I use 15 for 4x5). Then those images are combined and written as the final reconstructed original.

&off to SF for SPE

[jb7]

Struan Gray, on page 55 of the main DSLR Scanner thread, introduced a number of new ideas, including using a Telecentric lens.
I was quick to remind him of the danger of over-complicating the process, and that I would never consider buying yet another lens of such limited usefulness- never mind the risk that it might not conform to our necessary specification, without having any way of knowing, before making the purchase.

And that I have to break the habit of buying even more pricy pieces of glass.

I've recently been told that I buy glass the way ladies buy shoes, which kind of focusses the issue...

So, with that out of the way, I went out this afternoon, and bought a 55mm f/2.8 Macro Computar Telecentric.

The specifications didn't look promising, 11mm image circle, and a c mount.
However, when I got it back, I found that it lights up a full frame with only tiny bit of darkening on the corners- and that's holding the lens up to the F mount, without allowing for the additional mount I'm going to have to come up with. It's also at it's minimum magnification, which images a field of view about 6cm wide.
It focuses from there, to beyond 1:1.

So, I had thought I'd have to use it on a crop sensor, but it seems there might be more options.

There are no reviews I could find, but there are basic specifications here- http://computarganz.com/product_view.cfm?product_id=561

The lens is advertised as having minimal distortion- and a quick look through shows a bright contrasty image, and it looks straight enough-
However, I don't think I'm going to be able to test for about a month- it'll take a little while for the mounting bits to arrive from China, and I'm off to NY in ten days time, for a few weeks.

But it looks promising-
and goes to show how open to suggestion I am, as Frank and Struan can verify...

[Peter De Smidt]

So, with that out of the way, I went out this afternoon, and bought a 55mm f/2.8 Macro Computar Telecentric.


That's hilarious! Before opening up the thread, I never would've expected it.

I look forward to seeing your results. Computar made some outstanding enlarging lenses, as well as some excellent LF optics.

[Struan Gray]

So, with that out of the way, I went out this afternoon, and bought a 55mm f/2.8 Macro Computar Telecentric.

Ha!


Still, it's cheaper than the Zeiss 100 f2 by a significant margin.

[rdenney]

A bit of lens testing:

I'm still comparing my Canon 50mm Compact Macro, with the 1:1 converter, against my previously tested EL-Nikkor 105. The Canon has a significant advantage--autofocus confirmation.

But the shorter macro lens suffers from excessive fall-off, in the range of about 15%, at all apertures. I suspect this is a combination of the shorter focal length and the response of the sensor. The Nikkor's falloff is more like 3-5%. I've written in the light-source thread that I'm using an LED-backlit computer monitor panel to provide the backlight, and it's very even in an of itself. Using that for a light source also provides the means create a correction pattern that can be displayed on that monitor. More experimentation needed there, but I'm expecting that to be fairly easy. The panel is parked well behind the film being scanned so that it's pixel pattern is completely out of focus.

One test I used to determine lens performance: The size of the resulting files. I could not tell by inspecting the files of a normal negative which aperture was the sharpest, but the JPEG files sizes were: 3361, 3757, 4402, 4064, 3498, and 3225, for f/5.6, 8, 11, 16, 22 and 32, respectively. This outcome is nicely monotonic, peaking at f/11, which is completely reasonable. The EL-Nikkor file sizes ran about 15% higher than the Canon macro sizes, which were also nicely monotonic, peaking (unhappily) at f/22. That the Canon lens needed such a small aperture was not a positive statement in its behalf--aberrations took a lot of correcting before diffraction overcame them.

Remember that my camera is a 12.5MP Canon 5D when judging these files sizes. Raw files were in the 13-14 MB range, followed the same curve and were proportional to the JPEG file sizes.

Looking at the best of these files, which was the f/11 file for the EL-Nikkor, I see that the scan is not sharp to the pixel. Dust specs on the negative, which show up as white after inversion, have edges smeared over two or three pixels. They are sharper than the image on the negative, which I take to be limited by the photograph itself, though I was able to make very sharp 16x20 prints from this negative back when I had my darkroom. The taking lens was a Schneider Super Angulon 121mm f/8 at f/32. The original image is posted here. (http://www.largeformatphotography.info/forum/showthread.php?77076-Doors-and-Doorways&p=834903&viewfull=1#post834903)

I don't think I'm doing any better than the Epson than I was with my first quickie test, to be honest. I still cannot resolve beyond a hint of grain.

This picture is the whole 24x36 scan frame:
71093

Here is a 100% detail (or it will be after you click it). The staple is about 1mm on the 4x5 negative. I've adjusted levels and curves to match my conception of the image (as linked above), but this is without any sharpening.
71094

Here it is with a bit of sharpening to overcome the effects of the anti-aliasing filter on the camera. The sharpening setting is 0.6 radius, amount 200, zero threshold. I can just see a hint of grain, maybe. The detail in the negative isn't, I don't think, sharp enough to benefit much from sharpening at this level, but this is basic corrective sharpening, not the sharpening I might do when targeting the image to a particular print size.

71095

Maybe this is just as good as my 5D can do at 1:1. Or maybe I just need a better lens.

At least there's no distortion. I photographed a clear plastic ruler, and drew a line along the edge. At the resolution of my camera and lens, the edge was straight within a pixel along its entire length.

Rick "taking baby steps" Denney

[Peter De Smidt]

You might try downloaded Zerene stacker, http://zerenesystems.com/cms/stacker, which has a 30 day trial. Get the focus at the confirmation spot. Now back off the Velmex a small bit, about a 90* rotation. Take a series of photos, moving the velmex knob about 10* with each step. Go past the confirmation point by a good number. Bring the images into Zerene and run a Dmap stack. Now compare the image to what you took with a single frame. This will not only get rid of any focusing error, it will also help if the lenses' plane of sharp focus isn't really planar, and it'll alleviate any misalignment issues.

[rdenney]

You might try downloaded Zerene stacker...This will not only get rid of any focusing error, it will also help if the lenses' plane of sharp focus isn't really planar, and it'll alleviate any misalignment issues.

My next step is to check focus by running some tests, and that may entail a series of photos with small slide movements. But I don't think I need to go as far as focus stacking. Nothing I've seen suggests any misalignment--the image is uniformly sharp edge to edge. It's just hard to focus that camera that critically with a manual lens. I'm trying again with an eyepiece magnifier, and if that doesn't help, I'll make a series. With that Velmex slide, the minimum effect I can almost see is 180 degrees of the handle. The Canon auto focus narrowed that to about 45 degrees. I will also try a different adapter with Canon electronics in it, though I'm not sure that one is correctly clocked and camera ends up not quite level side to side.

Rick "back to the basement" Denney

[Nathan Potter]

Rick, I wonder about the difficulty of aligning the plane of the sensor to that of the film. Maybe it would be possible to make use of a cheap laser pointer. How about shooting the beam through the film mounting glass and reflecting it off a mirror taped to the front of the DSLR lens. Then adjust the DSLR plane to shoot the beam exactly co-linearly back through the film plane. A bit of fussing with this, but if the DSLR lens axis is precisely orthogonal to the sensor (high quality DSLR should be I suppose) then you should have a pretty good alignment. As I recall in your setup you may not be using a glass plate for film mounting - but maybe you could substitute one. The laser pointer would be temporarily substituted for the normal light source. Just thinking about alignment issues when I start fiddling with this.

Nate Potter, Austin TX.

[rdenney]

Okay, experiments have continued.

The focus was correct with the EL-Nikkor. I made a series of exposures and looked at each of them at high magnification, and I had to turn the Velmex handle a whole turn (0.025") to noticeably degrade the image at f/11 (indicated, f/22 effective) and 1:1. Everything in the range of +/- 0.015", at the very least, looked identical. So, I stopped worrying about focusing. I just focused one way until the the image, blown up on the camera's LCD, was noticeably less sharp, turned it the other way past focus until I got the same result, and then split the difference. Just like I used to do with the enlarger.

I also set the camera to lock up the mirror, pause three seconds, and then open the shutter, using a remote cable. Made no discernible difference at any shutter speed.

What was not correct with the EL-Nikkor was the EL-Nikkor. It's okay, but I doubt it's optimized for this magnification. And it's just a wee tad long to get much greater magnification with my setup. And my Canon macro lens only goes to 1:1.

So, I tried something different. I took a cheapie 80mm f/2.8 Arsat lens, put it on an Arsat reversing ring, and mounted it on my bellows arrangement reversed. The Arsat has a 62mm filter ring--not common--so that's the lens I have that will fit that reversing ring. I had bolted my bellows standards down onto an Arca plate, so I just slid the whole thing forward, focused it, and made some photos. Magnification is greater--about 1.8:1. The reversing ring provided considerable additional extension, as it turned out. So, the field of view is 13mm by 20mm. I can pretend I'm using my wife's D300, heh.

For the curious, the Arsat is a 6-element double-gauss lens, multicoated, and a good performer used normally except at apertures larger than f/5.6. It is, of course, a medium-format lens intended for 6x6, and this application is just using the sharpest part in the middle of the field. In this case, the greatest detail was recorded at f/11 according to the file size, but I could not see any improvement over f/8. f/5.6 was visibly worse, as was f/16.

And there is the grain. Once I could see grain, I decided to exaggerate the sharpening, partly to demonstrate Ben's point. So, I used a radius of 0.7, a threshold of 0, and an amount of 500. Dust spots on the negative became crisply resolved to the pixel, and I think I'm now getting everything there is to get from the sensor. Note that what made this possible was seeing grain--seeing grain allows radical sharpening, because the sharpening is increasing the apparent acutance of the grain. Given that there is no signal at that scale, it does not increase the signal/noise ratio at larger scales. Sure enough, even viewing the 13x20mm piece of the original negative as a single image does not show anything untowards with that much sharpening at the grain level.

This suggests to me that a key to this approach is providing enough magnification or resolution to resolve grain well enough to be able to really apply sharpening to it. That improves the MTF at larger scales without the grain becoming more apparent. But if you look at the image at much greater than 100% on the monitor, it will look a bit freaky.

Comparing this image with the one from the EL-Nikkor is interesting. The EL-Nikkor resolved some grain--there are larger grain clumps visible in both images--but not enough to allow that sharpening strategy without it revealing artifacts at larger scales.

So, instead of making multiple 1:1 exposures to reduce noise, my intention is to make the same number of exposures overall, but smaller tiles instead of redundant tiles.

Here's the 13x20 sample (on my monitor, the image, when you click below, is a piece of what would be a 24x30" print):
71117

And here is the detail at (if you click it) 100%--a 50x enlargement if your monitor has 100 pixels/inch:
71118

Both are post-sharpening.

Rick "making progress" Denney

[rdenney]

Rick, I wonder about the difficulty of aligning the plane of the sensor to that of the film.

I don't have an alignment problem. The film seems to be pretty exactly parallel to the sensor--there is no difference in sharpness across the frame.

Don't be confused by the apparent distortion--the door on the front of the Espada Mission does not have one single right angle on it, and the photo itself was not corrected for laterl perspective convergence. Since the subject was not parallel to the film when the negative was exposed, there is a touch of change in apparent sharpness across the face of the door. I was also trying to keep the stone door frame in acceptable focus--the shapes of those stones are the design point of that doorway. I'm evaluating sharpness mostly using dust spots and grain, as best I can.

Rick "noting that the camera axis is normal to the film holder within a thousandth or two over a little more than a foot" Denney

[Peter De Smidt]

Some lens tests: http://www.coinimaging.com/Lens_tests.html

[rdenney]

Some lens tests: http://www.coinimaging.com/Lens_tests.html

Sorta makes me want to find an old Micro-Nikkor. I've looked a few times, but I haven't found the right deal.

Rick "who owns too many lenses already" Denney

[Peter De Smidt]

Rick, I wonder about the difficulty of aligning the plane of the sensor to that of the film. Maybe it would be possible to make use of a cheap laser pointer. How about shooting the beam through the film mounting glass and reflecting it off a mirror taped to the front of the DSLR lens. Then adjust the DSLR plane to shoot the beam exactly co-linearly back through the film plane. A bit of fussing with this, but if the DSLR lens axis is precisely orthogonal to the sensor (high quality DSLR should be I suppose) then you should have a pretty good alignment. As I recall in your setup you may not be using a glass plate for film mounting - but maybe you could substitute one. The laser pointer would be temporarily substituted for the normal light source. Just thinking about alignment issues when I start fiddling with this.

Nate Potter, Austin TX.

I use a Versalab Parallel to do exactly that. It works a little better over longer distances, but it's still a big help.

[Peter De Smidt]

Rick,

I agree with your conclusions regarding grain.

Ideally, we want a lens that performs best at between f4-f8. We should be able to find that. Reversed enlarging lenses often perform best at 4-5:1, which is more magnification than we want.

[rdenney]

Okay, I just bit the bullet at KEH, and bought an 55/3.5 Micro Nikkor (I got the Ai version, though I know the previous version has some interesting qualities, because it needs to work on my wife's D300), the correct Ai extenstion tube (Pk13), and also an old pre-Ai M tube. Oh, and a reversing ring. So, if I have to keep my magnification greater than 1:1, I can reverse the lens and add the additional ring. These lenses have always attracted magical reverence--I hope this one lives up to it. $133 total for all of the above--still a fraction of the price of that Computar.

Rick "who probably should get another Nikkor adapter for the Canon" Denney

[Peter De Smidt]

That's a newer lens than mine. It should be a very good performer.

Make sure to check the various adapters for shiny surfaces, and it might help to make a baffle out of flocking material to put inside the extension tubes. If you need a small bit of flocking, let me know the size, and I'll send you a small bit. A diy lenshade with an opening only a little bigger than the image area would also help.

[rdenney]

That's a newer lens than mine. It should be a very good performer.

Make sure to check the various adapters for shiny surfaces, and it might help to make a baffle out of flocking material to put inside the extension tubes. If you need a small bit of flocking, let me know the size, and I'll send you a small bit. A diy lenshade with an opening only a little bigger than the image area would also help.

My collection of Russian and Ukrainian cameras demands that I own a good-sized stock of flocking paper, heh.

The Micro-Nikkor is a good addition to my tiny collection of classic Nikkors, bought for their classic status and excellent current price. I have the 105/2.5 and the lovely 180/2.8ED. And the Nikon Series E 70-150, which is an amazing lens for being a cheapie, even though I don't do much with it. I've wanted the Micro since about 1975 when a buddy of mine bought one and I saw the pictures he made with it.

Rick "looking forward to trying it out on this project" Denney

[rdenney]

That's a newer lens than mine. It should be a very good performer.

The Micro-Nikkor arrived today, and it's a nice old f/3.5 Micro-Nikkor of Ai vintage--late 70's. Optics were perfect. Also arrived with the PK13 extender and a BR2 reversing ring.

My initial test was at 1:1 with the extender. F/8 was again the sharpest aperture. F/3.5 showed obvious softening at the edges, and even f/8 didn't fully correct that. I was surprised, to say the least. But I can only assume that the field is not quite as flat at 1:1 as it is at 1:10, which is probably where this lens was really optimized.

At 1:1, the Canon 50/2.5 Compact Macro and Life-Size Converter out-performed it. The center was the same, but the edges were very much better with the Canon. Again, I was surprised. I'm not sure I should have been, though, since I'm not usually subject to the Nikon Kool-Aid. The old Micro-Nikkor is a double-gauss design with five fixed elements, and the Canon lens is a 7-element double-gauss design with two additional corrector elements that are fully floating. And the life-size converter is not an extension tube but a specially designed 4-element optical converter specially made for this lens. As Kingslake would say, it has many more degrees of freedom in the design. But the Nikkor is still better in terms of falloff where the Canon lens does poorly, which is a limitation in this application.

But one cannot reverse the Canon or add extension tubes--Canon would rather you buy their specialty macro lens for greater than 1:1. Given that I'd gotten better results with a cheapie 80mm normal lens reversed on bellows, I took the Nikkor and reversed it on the extender. Reversing it moves the glass much further away, so with the extender, the magnification was about 1.5:1--not quite as much as I'd gotten with the reversed 80mm lens. It has about the same field of view as a APS-C (DX in Nikon-speak) sensor at 1:1. An important advantage to being used this way is that I get more magnification than I can get with the Canon, but still in a lens fast enough to provide an autofocus confirmation (with my electronics-equipped adapter). I found the focus confirmation to be dead-on accurate at f/3.5, and then I stopped down manually. Reversed at 1.5:1, and at f/8, the Micro-Nikkor was completely sharp to the corners and rendered grain, though a touch more softly than the reversed 80mm lens. But I was still able to sharpen radically at a sub-pixel radius and still get quite smooth images at viewable sizes. This will be my working setup, I think. It's not really any sharper than the reversed 80mm lens on bellows, but the provision of a working focus confirmation tips the scales.

The reversed Micro-Nikkor has another really nice feature. The rear of the lens has a rectangular opening as a mask for the lens when it's racked out in normal use. I can use the focuser to extend the rear of the lens out as far as I like when reversed. It makes a very efficient shade that can be nailed right down to the edges of the field of view. That's a good thing in this application.

Rick "who now thinks the 5D AA filter is the limiting factor at 1:1" Denney

[Peter De Smidt]

I took my samples for the lighthouse picture with the Nikkor reversed at 1:1 magnification, confirmed by photographing a mm scale. (I used an extension ring too, I can't remember which one.) I used F5.6. The big difference is that my camera has a DX sensor.

[rdenney]

I took my samples for the lighthouse picture with the Nikkor reversed at 1:1 magnification, confirmed by photographing a mm scale. (I used an extension ring too, I can't remember which one.) I used F5.6. The big difference is that my camera has a DX sensor.

Must be a short extension tube, not the 27.5mm tube that is intended for 1:1 use with the lens not reversed. I think it's about 1:1.5 reversed and mounted directly on the camera, and 1.5:1 on the PK13 tube.

The smaller sensor makes a big difference in a lot of ways. I now also know how you got such even illumination, heh.

In the center, there was no observable difference between f/5.6 and f/8. But the difference in the corners was definitely observable.

Just to be goofy, I'd like to try my P67 135mm Macro-Takumar. It's a dialyte design, apparently pretty symmetrical. But I don't have a way to mount it just yet, and whatever I use has to be strong--that's a big, heavy lens. I had it out today and pondered it--I need it this weekend for another project. I'd probably spend my time better by mounting that Magnagon, heh.

Rick "doing some art copy work this weekend, on digital and on film" Denney

[Peter De Smidt]

Yep, the tube was a short one, one of Nikon's old 'K' sets.

One key to even illumination is to make the light source significantly bigger than the negative. I found that out when doing some masking on my Cezanne.

[rdenney]

Yep, the tube was a short one, one of Nikon's old 'K' sets.

One key to even illumination is to make the light source significantly bigger than the negative. I found that out when doing some masking on my Cezanne.

I'm still using my "computer monitor" LED light source--much bigger than the even the negative being scanned. Yes, in a diffusion source, the light has to hit each part of the negative being scanned from all of the same angles as all other parts. The laptop display is pretty blue, but okay for testing at this point. When I get around to it, I'll task an old display with the job and see about making some correction files. Clearly, the falloff I'm seeing is related to the sensor in my camera, and influenced by the lens. Testing that has made me realize just how dirty my sensor is, too.

I'm thinking that 1:1.5 would be just about perfect, though, for digitizing old 35mm color slides on a DX camera like my wife's D300.

Rick "who can get to about 2.5:1 with this rig" Denney

[Peter De Smidt]

I am going to be very interested to hear how your Takumar performs.

[rdenney]

I am going to be very interested to hear how your Takumar performs.

Ha! Don't hold your breath. That old, all-metal 135/4 Tak is weighty, providing coverage as it does for 6x7. A reversing ring might be necessary--I've never found anyone who can ascertain at what magnification the lens was optimized for. Apparently, it was optimized for close-up--there are many reports (some of which may be repeating each other--Struan's crowd-source) of the lens doing less well at infinity. Probably, it is optimized for something like 1:10--it doesn't even make it down to 1:3 without tubes. I have the tubes, but not the reversing ring or the adapter. I've seen the adapters, but they are sort-of expensive, though they do have what is probably an inadequate shoe. I'd have to adapt a Bogen long-lens support, probably. I wouldn't want to hang that lens off my 5D unsupported, especially on tubes.

Now, if someone would buy me a Pentax 645D, then my priorities would shift immediately, heh. 1:1 on a 33x44 sensor would require fewer tiles!

Rick "accepting donations" Denney

[Peter De Smidt]

Here's another link to some lens tests: http://savazzi.freehostia.com/photography/apo_componon_60.htm

[Peter De Smidt]

And some more interesting tests: http://www.photomacrography.net/forum/viewtopic.php?t=16608&start=15

[Peter De Smidt]

Here's a link to a layered tiff that shows a 55 nikkor forward, reversed, led source, and a flash source: http://dl.dropbox.com/u/3595413/55mmMicrotests.tif The magnifications are very close to 1:1, although there is a slight difference between the forward and reverse magnifications. The image is of a 24mm long section of a 4x5" TMX negative developed in Pyrocat MC.

[ludvig friberg]

I think they look good all of them, they all clearly resolve the grain in the film. Its hard to judge color spectrum differences on the different lighsources on black and white though. What would be a good test image to shoot for color? I have a Epson printer(4900) that has very large gamut. I was thinking of printing some kind of test and shoot on Porta and scan that as best as I can on my coolscan and use as a reference. Then try to see how close I can get with different lightsources on dslr. What would be a good testchart, testing color and sharpness?

Ludvig

[Peter De Smidt]

I wouldn't judge color on any of them. The color of the stained film is very brown. Each file was white-balanced in Light Room. I doubt the my low CRI led source would be ok for color. The best test for color would be an IT8 slide. Wolf Faust sells them cheaply.

A good test chart is any flat film that was sharply focused. You can use fancy USAF resolution testings slides, or similar, but they are expensive.

[Peter De Smidt]

While this lens is way overkill for the scanner project, some of the people interested in the scanner might also be interested in general macro work. There's a very good deal on a JML 20x finite objective. It's about $45 shipped. See: http://www.ebay.com/itm/JML-Optical-20x-Objective-w-Casing-MANY-AVAILABLE-/200718941277?pt=LH_DefaultDomain_0&hash=item2ebbc7fc5d For some discussion of these lenses, see: http://www.photomacrography.net/forum/viewtopic.php?t=16171&postdays=0&postorder=asc&start=75 Note that at this magnification focus stacking would be required for any useable depth of field.

[Old-N-Feeble]

While this lens is way overkill for the scanner project, some of the people interested in the scanner might also be interested in general macro work. There's a very good deal on a JML 20x finite objective. It's about $45 shipped. See: http://www.ebay.com/itm/JML-Optical-20x-Objective-w-Casing-MANY-AVAILABLE-/200718941277?pt=LH_DefaultDomain_0&hash=item2ebbc7fc5d For some discussion of these lenses, see: http://www.photomacrography.net/forum/viewtopic.php?t=16171&postdays=0&postorder=asc&start=75 Note that at this magnification focus stacking would be required for any useable depth of field.

Damn... I won't open a PayPal account (hate hate hate), the auction states PP only, and the seller won't reply to questions.

[Peter De Smidt]

As per their auction: "Need to contact us? Please use the eBay message system under the "contact seller" hyperlink. We receive many emails each day and strive to answer them within 24 business hours. We rarely exceed 48 business hours when responding to emails. " He's been very responsive to others on the macro list.

[Old-N-Feeble]

Peter... That's what I did but the seller has that disabled. I received an auto-reply stating "...unfortunately the seller cannot reply..." blah blah blah.:(

[Peter De Smidt]

If you go to the <About the Seller> link, it'll take you to a page that lists their email address along with a live chat option.

http://i955.photobucket.com/albums/ae37/peterdesmidt/sunnking.jpg

[Old-N-Feeble]

Thanks, Peter. I sent them an email.

[Peter De Smidt]

Make sure you read the macro forum posts about the lens. At that magnification, it's very specialized. Without the ability to do focus stacking with less than 10 micron steps, you won't get very good images. With a dlsr scanner, it would mostly be useful for film/developer tests, where the idea would be to capture one tile of the film, as opposed to covering the whole negative, which would require a huge number of tiles. That said, how an image looks when printed at it's final size is much more important than how the film looks under a microscope.

[Old-N-Feeble]

Peter... Since it's optimized for 20x I was hoping it would perform well at 4x or so but that may not be.

[Peter De Smidt]

Peter... Since it's optimized for 20x I was hoping it would perform well at 4x or so but that may not be.

I doubt that it would be any good at all at 4x. If that's the goal, there are some better options. I posted some examples from an inexpensive 4x objective and some info in posts 3 and 5 of this thread.

None the less, for general scanning use, something like a standard 60mm macro, especially one that goes to 1:1 with autofocus, would be by far the easiest to use. Ideally, you'll be able to focus from a tethered computer. If you go much above 1:1 magnification, you'll need to use focus stacking for good results, and that greatly increases the complexity and cost of the project. For large format work, it would also be overkill.

[Old-N-Feeble]

Thank you, Peter. I need to take more notes so I can keep up. :)

[Old-N-Feeble]

I saw the following lens on the 'bay but didn't bid on it because I had other plans for my money and I don't know anything about it. My guess is one like it could be very promising.

The title begins with "Kodak Printing Ektar 103mm" and it's a recently completed auction.

[SURF]

I have tried to make a picture of the USAF resolution target with a Canon D450 (12 Mp), extention rings and some lens. It worked. The size on the target was about 24 mm x 16 mm. Was possible to get at least 3600 dpi (see the 400% crop). Really not bad.80459

[Peter De Smidt]

Hi Surf, Thanks for the test.

[Dan Fromm]

Um, impossible to evaluate without knowing the magnification at which the shot was taken. On my screen, group 6 element 2 is mush, group 6 element 1 is just there. If shot at 1:10, group 6 element 1 is 64 lp/mm. Surely you can do better.

FWIW, at f/4 a 55/2.8 MicroNikkor AIS should resolve at least 120 lp/mm.

[SURF]

Magnification is close to 1:1 as the sensor is APS-C 22.2 × 14.8 mm, 12.2 effective megapixels, 4,272 × 2,848.

[Dan Fromm]

Oh, my. If at 1:1 and my mental arithmetic is correct, group 6 element 1 is 6.4 lp/mm. Oh, my. I hope I've made a mistake. I hope that something in the chain between sensor and my monitor has betrayed you.

[SURF]

Let's talk about long side of the shot. At the camera it is sensor length and it is 22 mm. I have measured the the size at the USAF target that was shot to file - it is 24 mm on the glass target. So we have 22:24 that is very close to 1:1. The crop is 400% enlarged for better viewing.

Bottom line: one can use the setup for DSLR highres scanning with resolution about 3200-3600 dpi on target (put film there instead). You see 3200 dpi, I see 3600 dpi - it's normal when two persons examine the USAF target scan. What's good about all that is that the resolution can be measured.

I have no idea yet what numbers will be be when I will take a better camera and a better lens. Perhaps two times better? Who knows.

[JPR]

i use a rodagon D 74mm (for 1:1)

i like it very much it destroyed my 90s contax zeiss 60mm makro


i would rate it 9/10 for this kind of work there is a 2:1 / 1:2 version (sweet spot with the new nikon d3300 24MP DV without AA Filter)

[drtebi]

Trying to revive an aging thread...

I have been making a few preliminary and very rudimentary tests with a Ricoh GXR camera with the 50mm Macro lens and APSC sensor, "scanning" medium format negatives and positives.

I am confused about something:
According to DP Review, they conclude that the lens has a 0.4% barrel distortion. But when I open the same film-scanned image (Minolta Dimage Multi Pro) and layer the camera-scanned image on top, resize, center, and rotate the images appropriately, then it appears that the Ricoh has a pincushion distortion?? That is assuming that the scanner does not distort anything, which I assume it shouldn't.

Is this possible? Did I miss the point of distortion somewhere?

Confused.

[rdenney]

Trying to revive an aging thread...

I have been making a few preliminary and very rudimentary tests with a Ricoh GXR camera with the 50mm Macro lens and APSC sensor, "scanning" medium format negatives and positives.

I am confused about something:
According to DP Review, they conclude that the lens has a 0.4% barrel distortion. But when I open the same film-scanned image (Minolta Dimage Multi Pro) and layer the camera-scanned image on top, resize, center, and rotate the images appropriately, then it appears that the Ricoh has a pincushion distortion?? That is assuming that the scanner does not distort anything, which I assume it shouldn't.

Is this possible? Did I miss the point of distortion somewhere?

Confused.

It may be that the lens has barrel distortion at infinity and pincushion at 1:1.

Rick "who created his own corrections in Photoshop" Denney

[SURF]

Up!

Rodenstock lenses used in Scitex Smart 340, Smart 342, Smart 340L, Smart 342L scanners:

Rogonar-S 5.6/60 for scanning A3, A4, 8x10". Optical magnification 0.111.
Scitex S-3 4.9/67 for scanning 4x5". Optical magnification 0.215.
Scitex S-3 5.0/89 for scanning 6x6cm, 6x7cm, 6x9cm. Optical magnification 0.376.
Scitex S-3 5.0/110 for scanning 35mm. Optical magnification 0.726.
Rodagon 5.6/80 for scanning 10cm stripes LW. Optical magnification 0.346.

[SURF]

Some tests of those lenses. Optical magnification 1:1.5 (36x24mm to 24x16mm 16Mp Sony Nex).

Let's start: Rodenstock Rodagon f5.6, 80mm. Does not look very APO, but sharp and with a good contrast. Center and corner 100% crop.

[SURF]

We have a very good lens in this setup: Rodenstock-Scitex S-3 89mm-f5.0. Apo without doubt. Center and corner 100% crop. BTW the CCD length in those scanners is 35mm/5000 pixels. It covers crop-camera sensor diagonally.

[SURF]

Rodenstock Scitex S-3 4.9/67. Center and corner 100% crop.

[Dan Fromm]

Surf, what am I to conclude from your tiny fuzzy images?

[SURF]

Surf, what am I to conclude from your tiny fuzzy images?
You can compare optical resolution and optical aberrations of the lenses in my setup, that is simple and common.

[Dan Fromm]

You can compare optical resolution and optical aberrations of the lenses in my setup, that is simple and common.

Not with the tiny fuzzy scans you posted.

[SURF]

It's like a brain cells - they are tiny and fuzzy. But... OK.

We will continue: Rodenstock-Magnagon-75mm-f5.6. Center and corner 100% crop.

[SURF]

And if you go stitching 1x with Magnagon, you will get something like that (camera jpg, flip-crop-save only). Center and corner 100% crop.

[SURF]

Finally: Rodenstock Scitex S-3 f5.0, 110mm. Center and corner 100% crop. I expected it to be better.

[SURF]

Scitex Smarts were made when there were no usable PhSh and nobody planned to make stitches. The main media for images was 120 film and larger (books, magazines). So we have the best performing lens 89mm and it's for 6x6, 6x9. Another good performer is 80mm - and it was made for stitching BW large film photoforms. It has good center and corner sharpness, but suffers from pink tint a bit, that is no issue in BW.

Rodenstock Magnagon is from another era: XY and stitching scanners. One scanner had only one lens, but it was very good. It will be nice to test Scitex LFOV lens, that is in EverSmart scanners. They say it's one of the best.

How would you rate those lenses? I will position them that way (first=best):

1. Magnagon 5.6/75
2. Scitex S-3 5.0/89
3. Rodagon 5.6/80
4. Scitex S-3 5.0/110
5. Scitex S-3 4.9/67

[Peter De Smidt]

I have two Magnagons. First, there is a Linos, serial number 11922735. Second is a Rodenstock 11619890.

[interneg]

For what it's worth, the Magnagon 75mm fixed at f8 is used in the Hasselblad X5 (and the rest of the Imacons?).

[Peter De Smidt]

https://www.dropbox.com/s/gjwjxoyz2mj4212/magnagon.JPG?raw=1

[Patrick Gauthier]

This may not flow with the progression of this thread so far, but I thought I would contribute what I can.

I have had success with a minolta md 100mm f/4 macro lens shooting at f/8 using a sony nex-7 (aps-c). I've included a screen shot of a test I did a few years back demonstrating a 400% crop. As you can see, the resolution is impressive. However, I understand this is not a standardized line test.

172959

Fortunately there is a repository for all things Minolta which provides a more comprehensive review of this lens (and others). http://www.rokkorfiles.com/100mm%20macro.htm

If this helps at all and if there are more question that may aid in your project I would be happy to contribute what I can (I can run more tests with this lens/camera arrangement if there is a need).

[erian]

And if you go stitching 1x with Magnagon, you will get something like that (camera jpg, flip-crop-save only). Center and corner 100% crop.

Could you please elaborate a bit more this statement? What did you do to get sharper image?

[Peter De Smidt]

Good alignment and focus are critical for scanning, as is avoiding vibration or stray light.

[SURF]

Could you please elaborate a bit more this statement? What did you do to get sharper image?
The scale is different for this take. It's 1:1. Sharpness is limited with a pixel size for that lens. It outresolves my humble 16Mp Sony aps-c , so going Sony 24Mp aps-c will improve things further. I have no tests on a full frame.

My Magnagon was taken from nonfunctional Agfa XY15 scanner. Serial starts with 115. The lens is a standard M39 connection. It has a UV/IR filter (looking magentish and was very hard to unscrew). I found that filter do not harm image and left it on.

If I remember right I saw different Magnagon in Topaz scanner: it was smaller and was different focus length.

BTW those Scitex lenses were taken from formally Screen scanner. There was a breaf period when Screen sold relabled Scitex Smart scanners. Maybe Scitex put dog lenses in those? ;-) Anyway when you see marks on a Scitex S-3 lens it's OK. They marked lenses for the best optical position in scanner.



Fortunately there is a repository for all things Minolta which provides a more comprehensive review of this lens (and others). http://www.rokkorfiles.com/100mm%20macro.htm

This is a good link.


Good alignment and focus are critical for scanning, as is avoiding vibration or stray light.
Yes. In my plans is to write something about it. I made a target that can be placed on a lens plus mirror trick for alignment.

Another BTW: I did't touch focus between center-corner tests. Will show you my setup later.

[Dan Fromm]

Pardon my ignorance, but Agfa made Magnogon lenses. Is the lens you pulled from that Agfa scanner one of them or is it a Rodenstock Magnagon?

[SURF]

The word "pulled" made me to remember What Was It to remove the cold welded lense from the scanner going to landfill. After some hours this Rodenstock Magnagon lens was saved.

[SURF]

Just to keep the "reference" here: Epson Perfection V800 Photo taken from http://www.filmscanner.info/en/EpsonPerfectionV800Photo.html

Those guys do not bother to test corner performance. Anyway here it is (center):

[Peter De Smidt]

Yes, it's not that hard to do better than an Epson with a Dslr system, and by that I mean about 1000 dpi more resolution and a stop more dynamic range. Whether that's a big deal or not depends on the end use.

[Two23]

OK, I'm convinced I should at least seriously try this scanning. I have a Nikon D800E & Nikon 105mm micro VR lens. Also have an older small color corrected light panel. I want to scan 35mm, 120, 4x5, and 5x7. I'm thinking I could scan 35mm & 6x6 with one shot, 6x9 with two shots & stitch, 4x5 with 4 shots & stitch, and 5x7 with 6 shots and stitch? What ebay copy stand will work for me? What gadget/mount will allow me to precisely move either camera or negatives to do the stitching? If this works as well as I hope, I could sell my Nikon Coolscan V and Epson v700.


Kent in SD

[barnacle]

Just an observation, not a criticism or anything... my biggest concern here is with stitching. As soon as you have the computer applying interpolation between two overlaid images, you're in the position of modifying the original scanned data - both deleting and recreating it, with added filtering in both linearly and possibly rotationally, depending where the software decides the overlay is and how well-designed the software algorithms are.

While the intent is to extract the maximum detail from the scanned image, is fictional data around the joins necessarily the best way to achieve this?

Neil

[DaveF]

I've had great success over the last year and a half using my Nikon D800E to scan all my slides and negs. Since I came to large format from predominantly shooting macro, I already had a pretty decent setup, which I've added to and adjusted over time.
My basic setup of accessories: Gitzo 3540LS tripod with levelling base, Arca Cube geared pan head on Gitzo levelling base, Novoflex Castel-L macro rail, light panel.
For digitizing 5x4 (5x4 crop-mode, stitch in photoshop of 4 frames) I use the Novoflex horizontally, and use my Nikkor 85mm PC-E with shift to achieve excellent results with the minimum of digital interference in the stitching process. (I just use books beneath the light panel to get the height about right, and use a bubble level on both the camera back and light panel to check levelling, and move the camera along the rail horizontally and shift the lens vertically up and down from centred on the middle of the slide to achieve the four frames to stitch). The film is held in place and flat on the light panel with a very rudimentary frame made out of some pieces of Lego.
For 35mm (a single frame) and 120 (6x7, 2 frames stitched) I use my 60mm AF-S micro, attached to the Novoflex rail set vertically to work like a copy stand.
With this setup I can get a 5x4 "scan" of slightly over 80mpx; the only limitation here really is the colour information the camera can catch (e.g. the colour separation between yellows and greens on a slide of Velvia 50). Not quite in the same ballpark as a drum scan, but I'm fairly confident that this would give an Epson scanner a very good run for the money, with almost certainly sharper results.

[Peter De Smidt]

Just an observation, not a criticism or anything... my biggest concern here is with stitching. As soon as you have the computer applying interpolation between two overlaid images, you're in the position of modifying the original scanned data - both deleting and recreating it, with added filtering in both linearly and possibly rotationally, depending where the software decides the overlay is and how well-designed the software algorithms are.

While the intent is to extract the maximum detail from the scanned image, is fictional data around the joins necessarily the best way to achieve this?

Neil

Neil, that's certainly a concern. I use a stitching template in PTgui for scans made with multiple captures. In practice I haven't seen a problem, but I haven't done extensive testing. For a small number of exposures, it's not hard to manually align the samples in Photoshop. Change the blending mode of the top layer to "Difference." Use arrow keys to nudge into place. Change blending mode to "normal"

[Daniel Moore]

Neil, for capture methods such as an X Y stage and linear movement, the amount of invented (interpolated) data is probably far less than say when stitching conventional captures which are projected onto a 2D field from a spherical one like a panorama head would produce when rotated around a nodal point. Since with X/Y capture systems great pains are taken to align the image edge with the sensor edge and the stitching program is 'told' that a very long focal length lens is in use (I use 1000mm focal length), a nearly orthographic projection results from the captured images. This means that again, very little information needs guessing at in order to stitch images without visible seams.

When I compare the results of my DSLR scanner to pano head stitches the results are night and day in terms of detail/sharpness. I'm not saying my film is resolving more, it isn't, just that the before and after results with film remain quite similar, not so with my digital captures.

I once painstakingly captured a distant landscape with a Leica 50mm Summicron at f/8 on a D800E. Individually the captures left nothing to be desired. Once stitched and projected onto a flat field with a Sinc 256 algorithm in PTGui, the micro-contrast and stunning detail that was initially present had all but turned to relative mush. Quite disheartening indeed. So while your concerns are well founded, they have yet to rear their head in my X/Y capture system.

[barnacle]

Excellent news, Daniel, and long may it continue.

I guess I've been too long an engineer; I've seen most of the horrors that can happen once algorithms get in the act, and I tend to look at the bad effects first :)

Neil

[SURF]

Rodenstock Scitex LFOV 108mm f5.6. Center and corner 100% crop. Excellent.

[mdarnton]

I've been messing with lenses for my 35mm scanning recently. I thought I'd done pretty well with a 63mm/2.8 EL-Nikkor, which has a good macro reputation, but just for fun I tried a 75/4 El-Nikkor. Not their best, and a simple 4-element lens, but I was surprised how good it was. I'm assuming this is because it's a Tessar type, and these have a great rep for sharpness in the middle, better than fancier formulas, supposedly. And here with it for 35mm to APS format, I'm only using the very center of its field--the best part. I also know that my Leica 65/3.5 Macro lens has been blasted for being only a 4-element Tessar rather than some fancy thing, but you know what, it did a great job for scanning my 35mm, too.

So I wonder if this concept of using just the best centerfield area of a longer lens and then not having to worry at all about how the [cropped out] edges perform has any application here.

[Greg]

Since 2004 have been using my DSLR for years to make digital files from my negatives. First project was to make digital files from a collection of 1,033 glass plates in our town's museum. Started scanning the collection with an Epson 3200? flatbed scanner, but soon realized that using a flatbed scanner very inefficient. Switched to a FX DSLR with a 60mm Micro Nikkor and a lightbox. Think it took me about 3 times to standardize everything.

Presently I use a Nikon Multiphot so alignments of the film, lens, and digital back not a factor. For shooting 35mm and 120 film, the illumination is the Multiphot's condenser illumination or a small Aristo light box resting on the condenser or a double exposure 1/2 with the collated light of the condenser and 1/2 with the diffuse illumination of the lightbox. For 4x5 through 8x10 a 10.75"x10" graphiclite D5000 Standard Viewer lightbox (cost me an arm and a leg back when). For 11x14 I use a LED light panel: 12.5"x16" "LED COPY BOARD" model A3 from Honour Management Consulting Limited, CHINA... if I recall well under $100 and money well spent. It's evenness of illumination is close to perfect.

Lenses that I have used range from Macro NIKKORs (Multiphot lenses), Leitz Luminar, Companion-S, to Macro lenses for 35mm, and a bunch of other optics. Over they years have settled on the following combinations. Lenses are stopped down 1 f/stop from the maximum aperture except where noted:

35mm B&W and color: 200mm f/2.8 AF Micro Nikkor
120mm B&W: 65mm f/4.5 Macro NIKKOR (Multiphot lens)
4x5 film & 4x5 glass plates, 5x7 glass plates & film Whole Plate: 12cm Macro Nikkor (Multiphot lens) or 60mm f/2.8 AF Micro Nikkor at f/5.6 when using a Nikon copy stand on location.
8x10 film and glass plates & 11x14 film: 60mm f/2.8 AF Micro Nikkor at f/5.6. For these film sizes I prefer to scan the negatives with my V750 PRO Scanner (6 merged scans in Photoshop for 11x14 negatives), but when faced with a stack of 8x10 glass plates.... well shooting them on a light box was so much more efficient.

I personally print only up to 11x14". In comparison scans made with a Imacon scanner of 35mm film, the Imacon produced superior scans and the difference could be seen in an 11x14" print, but only barely. For film sizes of 120 and larger, you really couldn't see any differences. Reproduction of the tonalities between using the Imacon and the DSLR differ slightly, but not enough to matter for me. When going from film, to digital file, to digital negative, and finally to a Platinum/Palladium print... well you have to accept that some of the tonalities will change, BUT in my experience is not to the detriment of the final print.

Films and glass plates are always masked on the light box or held in film holders. I once did a comparison of shooting a 5x7 glass plate on my graphiclite D5000 Standard Viewer lightbox with and without masking out the areas on the lightbox surrounding the glass plate and the difference was amazing.

Very recently some of the circa 1890-1900 5x7 & 8x10 glass plates that I have encountered have a D-Max of 2.5 and more. For these I have planned on bracketing exposures and using Photomatrix Pro 6.0 (HDR application) to achieve a full tonal range final digital file. Have been shooting a step wedge to calibrate the results, but so far have seen a distorted recording of the evenly progressive densities in the step wedge.... need to do more work on this technique.

Ilford Antistatic cloth always used to wipe the film or glass plates pre scanning.

Comments most welcome....

[SURF]

My resolution tests are not too impressive, because the resolution of the camera is only 16Mp and lenses outresolve it. Shall I buy 24Mp crop camera to gain more resolution? Let's see. I took two x2 TeleConverters to watch the center closer. (TeleConverters are not the best - we have what we have).

Result: much more resolution - enough to start thinking about 50Mp crop camera. (No such animal yet.)

Optical magnification 1:1.5 (36x24mm to 24x16mm 16Mp Sony Nex). This is without TCs. TCs are used to see what's happening in the center of such setup.

Lenses: Rodenstock Magnagon 75mm f5.6 and Rodenstock Scitex LFOV 108mm f5.6. 100% crops

[Two23]

The only macro lens I have is a Nikon 105mm f2.8 VR (latest version.) This will not work for photoscanning 35mm, 120, and 4x5? I plan on stitching in Photoshop.


Kent in SD

[Peter De Smidt]

Sure, it can work. Why not?

[rdenney]

Since this popped back up, I made a discovery this week. When I tried my initial experiment with my 5D several years ago, I found my 50mm Compact Macro (with life-size converter) resolved grain, but the grain was bigger than it seemed under a loupe.

Well, duh. It was aliasing. My 12+ megapixel 5D, at 1:1, has a resolution of about 2800 spi. I remember from old film scanners, like my former Minolta Multi II, that aliasing was a problem with some film.

I found this out because I’m once again trying to make headway on digitizing my old Kodachromes much more efficiently than running them through the scanner. I came upon an old Nikon PB4 bellows, with the slide duplicating attachment. I thought I might get better results than I had gotten with my Illumitran, which could never manage contrast with its internal strobe, or overcompensated with the contrast control unit even on its minimum setting. I mounted an old 55mm Micro-Nikkor and put a Kodachrome slide in it. The result was an almost perfect histogram, but with radical graininess. That’s when the bell rang in my head.

My 5D is 13 years old so I thought maybe it’s time to upgrade, but I’m already mid-spending-spree in a Pentax 645z kit, so I cheaped out and bought an experienced 5DII, for about a fifth the price of the current model. The 21mp sensor will give me about 4000 spi at 1:1. That should eliminate aliasing.

Of course, I could also scan 4x5 with the 645z, which gives about 4800 spi at 1:1, with no AA filter. I do have the 120mm Macro, which does focus to 1:1. 16 photos would cover 4x5 with plenty of overlap for stitching. Hmmm.

(And I have other plans for that bellows: 1. It’s a more rigid design than the macro apparatus I built for scanning 4x5, 2. The tilt will help with photographing watch faces—a current interest—and 3. I found an old but late model 100mm Bellows Takumar that might find a happy home on those bellows.)

Rick “who wants to put a book together of old slides for a friend who accompanied me on those excursions, who needs a little uplift this year” Denney

[Oren Grad]

The 21mp sensor will give me about 4000 spi at 1:1. That should eliminate aliasing.

No, it won't, at least not if you want to copy B&W film, chromogenics excluded. My LS-9000, which delivers nominally 4000 spi, does nasty things to Tri-X. I don't know for sure how far one needs to go - 6000? 8000? But 4000 plainly isn't enough. Mind you, it's adequate for chromogenic films; the dye clouds degrade gracefully under less than perfect description. And it might be OK for a film like TMX, where the grain is so fine there's no prayer of capturing it other than under a microscope. But ISO 400 or old-fashioned-grain medium-speed silver films - no.

FWIW, I did some experiments with a rented 5DsR and a 75mm Apo-Rodagon-D 1:1 on a bellows a few months back with the hope that 6792 pixels in 24mm would give me a visible improvement. But nothing doing - couldn't focus it reliably enough with that camera's live view to achieve any advantage.

At some point it will work - a DSLR or medium format DSLR will combine enough pixels with good enough live view at the pixel level so one can actually achieve the theoretical resolution. But so far as I'm concerned, we're not there yet.

EDIT: Another observation: the 5DII has an AA filter. That, not the nominal resolution, may well take care of what you're perceiving as aliasing - at the cost of actual resolution. Depending on how far you want to enlarge and on your taste in image character, it may work for your purposes even so.

[Peter De Smidt]

Live view is not accurate enough in my experience. No, I don’t know why. I use a Velmex 4000 Unislide, and focus is determined by looking at full res files.

[Oren Grad]

Need to do the arithmetic to check this, but I suspect we're not yet at the point of getting 1 pixel = 1 effective dot in the LCD.

Checking the full-res files is of course the acid test. It would be practical for me if I had a dedicated setup with a precisely- and rigidly-aligned mounting stage so that I didn't have to go through tedious trial and error for every exposure. Maybe someday... or if and when live view gets to be good enough, that should solve the problem.

[SURF]

I put a 1.7x magnifying loupe on a LCD while focusing. Some can connect external monitor and it can be even better.


Well, duh. It was aliasing. My 12+ megapixel 5D, at 1:1, has a resolution of about 2800 spi. I remember from old film scanners, like my former Minolta Multi II, that aliasing was a problem with some film.

For what I like high-end scanners: calculated resolution is their optical resolution. With cameras it's more like only 70% of calculated. You can see it in my tests.

But for what I like cameras plus high-end lens is that the image is calm and undistorted.

[Peter De Smidt]

I can only report my experience with my D600. Shooting tethered to a big screen and focusing with live view was not optimal.

[mdarnton]

On my D7200, copying 35mm BW negs using a lens on a bellows, live view focus is dodgy. Part of it is that just a breath on the focus knob moves focus too quickly. I do better on the ground glass.

[rdenney]

What about using an eyepiece magnifier?

Rick “thinking we need a grain focuser” Denney

[Oren Grad]

Can't have a grain focuser until you have enough resolution in the imaging sensor to capture the grain cleanly. And then you need a live view that will let you actually see it, at 1:1, as you are setting up the capture. Note that AA filters, sensors that record images through a Bayer (or other) matrix and live view LCDs that build the displayed image out of separate RGB dots are not our friends in this task.

Everything else we're talking about is just fiddling with different ways of looking at an inherently mushy capture in the hope that we can somehow make it look slightly less mushy.

EDIT: To be clear: current technology accessible to consumers is not up to the task of an exact reproduction of the original down to the grain level, so we all have no choice but to compromise at some point, and everyone is entitled to decide for themselves how good is good enough. And we shouldn't forget that copying film on film is also a very lossy process - no illusions here about a lost golden age when everything was perfect.

[Oren Grad]

What about using an eyepiece magnifier?

In general, DSLR focusing screens are not designed for manual focus acuity, so magnifying the screen may not help much. In a 5DII you could swap the stock screen for an Eg-S, but if you intend to use a macro lens with a smallish maximum aperture, the extra snap may be lost. Critical focus through the finder also depends on precise alignment of the mirror and focusing screen with the sensor, and that can't be assumed. Despite all that, it might still help you get better results in your particular situation - your camera, your eyes, the particular originals you're trying to copy. The only way find out is to try it.

[rdenney]

The answer then is to have enough resolution to use a cluster of pixels to represent each grain. Yes, focus will still be an issue, but at least you might see it well enough with live view.

I might try it with the 645z when it comes. At 1:1, the sensor resolution is 4800--at 2:1 it would be 9600, and grains might be 2x2 or 3x3 pixel clusters, visible in live view. The macro lens, reversed, might get that job done--most macro lenses are optimized for 1:2ish and reversed would be optimized for 2:1. Also, no AA filter. A small enough aperture to account for minor misalignments might have the same effect as the AA filter, though.

Ultimately, though, I suspect a perfect representation of grain is going too far, unless we are making billboards that will be viewed up close. I'm pretty sure my enlarger lenses weren't able to pull that off, either, although I know there some now that will. And right now I'm scanning my 4x5's in an Epson flatbed--maybe 2000 spi usable. With only four images from the Pentax, I could nearly double that. As you say, we set our own requirements based on our own needs, and for me I don't print bigger than 16x20.

For making my Kodachrome dupes, though, I do not have to have it grain sharp. My lenses weren't that good back in the day, and I won't be printing these large. Projected, they'll be fine. But I don't want obvious grain aliasing, either.

Rick "needing orthographic projection here, too" Denney

[Ted Baker]

I would like use my Nikon DSLR to test some scanning software I am working on.

I don't own any macros lens I was going to get 60mm 2.8 Autofocus of ebay, hopping that I autofocus might actually be useful for roll film. Sort of like pakon for medium format. Am I just kidding myself with the autofocus? i.e. should I just get much cheaper old 55mm micro.

[Peter De Smidt]

Hi Ted, I only use manual lenses. The 55 micro is a very good lens.

[mdarnton]

I have an old pre-AI 55/3.5, an AI one, and the 60/2.8 AF-D. They are all equal for the task. All three are better than any mainstream enlarging lens I tested, except, strangely enough, the EL-Nikkor 75/4, which I had read was a stand-out for macro work, and the 63mm/2.8, which is a fine lens. Autofocus does work well for me for larger film. It's a bit tricky for copying 35mm.

[MAubrey]

In general, DSLR focusing screens are not designed for manual focus acuity, so magnifying the screen may not help much. In a 5DII you could swap the stock screen for an Eg-S, but if you intend to use a macro lens with a smallish maximum aperture, the extra snap may be lost. Critical focus through the finder also depends on precise alignment of the mirror and focusing screen with the sensor, and that can't be assumed. Despite all that, it might still help you get better results in your particular situation - your camera, your eyes, the particular originals you're trying to copy. The only way find out is to try it.

This is actually something that makes mirrorless bodies better, pixel level zooming for critical focus can't be beat by any SLR focusing screen.

[mdarnton]

I use live view focus on my Nikon D7200, and it's adequate.

[Ted Baker]

I have an old pre-AI 55/3.5, an AI one, and the 60/2.8 AF-D. They are all equal for the task. All three are better than any mainstream enlarging lens I tested, except, strangely enough, the EL-Nikkor 75/4, which I had read was a stand-out for macro work, and the 63mm/2.8, which is a fine lens. Autofocus does work well for me for larger film. It's a bit tricky for copying 35mm.

Thanks that's good to know. I am going to bid for a 55mm 3.5 Nikkor to start with. I can't go wrong with that. I mainly want to improve my software first.

I will probably get one of 60mm af lenses at some point to see if I can make it work for faster pakon/fuji/noritsu style scanning of roll and 35 film.

[Megapixel]

While looking for macrophotography lenses I came across these nicely done tests of optics (https://www.closeuphotography.com/lenses/) on the Close-Up Photography website (https://www.closeuphotography.com/). Some of these optics may also be suitable for a DSLR/mirrorless film scanning setup. He uses an integrated circuit silicon wafer for a high resolution target and compares film scanner objectives, microscope objectives such as Mitutoyo's, macro lenses such as Canon's MP-E 65mm and Voightlander's APO-Lanthar 125mm, enlarging lenses, stacked lenses, industrial lenses, close-up lenses, etc.

Well worth a look if you are considering these types of optics.

[Megapixel]

If you are considering reverse-mounting a Canon EOS-type lens for scanning, the Vello / Mieke Macrofier Reverse Mount Adapter and Extension Tube (http://www.vellogear.com/product/6489/Vello-RM_CEF-Macrofier-Reverse-Mount-Adapter-and-Extension-Tube-for-Canon-EF*EF_S-Lenses) may be of interest.

Its unique feature is the ability to reverse-mount a Canon EOS (Electro-Optical System) type lens and maintain electronic control of the aperture (EXIF data also, and maybe autofocus but safest to focus manually). B&H (https://www.bhphotovideo.com/c/product/1050997-REG/vello_rm_cef_macrofier_reverse_mount_adapter.html) and Amazon (https://www.amazon.com/Vello-Macrofier-Reverse-Adapter-Extension/dp/B00W4DNTV8) sell it for $100, but you can also purchase it more directly online for around $60 on eBay or the Mieke Online Store (http://www.meikestore.com/product/meike-mk-c-up-auto-macro-extension-tube-af-reverse-adapter-for-canon-camera/2827.html). You should consider what sizes of adapter rings you will need. B&H includes 6 sizes versus the Mieke Online Store includes only 4 sizes.

For some macrophotography examples of this in use with the inexpensive Canon EF 40mm f/2.8 STM lens, see the Google translation of makrofokus nonaC 40mm f/2.8 (https://translate.google.com/translate?sl=sv&tl=en&js=y&prev=_t&hl=sv&ie=UTF-8&u=http%3A%2F%2Fmakrofokus.se%2Fblogg%2F2015%2F9%2F24%2Fnonac-40mm-f28.html&edit-text=&act=url) and on YouTube, How I Built My Own Super Macro Rig for $250 (https://www.youtube.com/watch?v=PW-44IKD8l0).

[Peter De Smidt]

If interested, there are a bunch of long threads on camera scanners in the DIY forum. In fact those threads were the reason that the DIY forums were made. There's a whole thread on lenses. In short, use a lens optimized for the magnification you're scanning at. 1x is very different than 1/4x. The best value is an older macro (micro in Nikon terminology lens.) These fit on your camera without an adapter. They have fine focusing. The performance is pretty good. To get better, you have to get more complicated and expensive. For example, a 75mm Rodagon D is excellent at 1x. There's also a Rodagon optimized for 2x (or 1/2x). The highest resolving lens I tested was a Nikon 5x measuring microscope lens.....but the number of frames needed to cover even small film is excessive. In my own system, I found that going above 1x didn't help, even with 35mm techpan shot under controlled circumstances. Shooting a test target did show increases in resolution, but I don't make prints from resolution targets. Lenses also tested: Mitutoyo 2x APO, a whole bunch of enlarging lenses, both forward and reversed, Magnagons, Componon-Ms........