Scanning with a digital camera

[AlwaysOnAuto]

I think that if you have a good macro lens (Micro-Nikkor as an example), the limiting factor becomes the image quality of what is being copied.
I could be wrong, but that is my feeling. Nothing scientific about it at all. Just a feeling.

 

[AlwaysOnAuto]

This is similar to what I use, except the bellows is mounted on a focusing rail and the rig is mounted on a macro stand.

I usually have the bellows mounted to a board. I just had it out on the coffee table for the picture.

 

[petronius]

I use this old Hama Zoom Dia Duplicator from the fleamarket. My light source is a empty word document.

Here is an example in full size:
https://www.flickr.com/photos/182210755@N02/50137054518/sizes/o/

Agfa Optima 335 sensor, slide film 100 ASA (1992)

 

[madNbad]

The latest in the ever evolving camera scan setup. I think this one is a keeper. Sony A7II, Micro-Nikkor 55 2.8 Ai-S with PK-13 extension ring, Skier Copy box II, Alzo copy stand. I leave it assembled and the only setup is to plug in the copy box.


Scanning Setup

 

[ColSebastianMoran]

Looking for thoughts and expertise on this question: What's limiting image qual in Cam-Scans?

My attention is on cam-scan from 35mm.

My updated list of factors:
- Image detail on the film? (Film, taking lens, technique)
- Copy lens
- 24MPx ... Would more pixels improve?
- Sensor capability (Bayer design, color capability)
- Film flatness
- Technique in camera-scan (focusing, choice of aperture)
- Cam-scan setup (alignment, vibration, illumination)
- Software (e.g. negative inversion)

And, for any of this, how would we know? What comparisons would give us some answers?

Stated another way, what would have to change to get better files from film exposures?

My observations:
- I'm getting better files and better prints now than back in the old days.
- 50MPx with good lens shows 50% more resolution on USAF test target, so lens seems good to 50MPx
- 50MPx shows grain only slightly more clearly, many say "no difference", and
- 50MPx shows little advantage in image detail, so 24MPx good enough for now
- Good lenses are much better than poor lenses
- Really good macro and copy lenses are best at f/4, but given film curvature, f/8 is usually better overall tradeoff of DOF vs diffraction loss. (The exquisite copy lens from CoolScan 8000 is f/3; I can't get negatives or slides flat enough to take advantage of it; with this lens, I'll do focus stacking.) With better negs/slides and better flatness, I think my best lenses could delivery more image quality.
- Good technique and setup are critical. alignment, free of vibration, good back-light
- One shot with right exposure is enough for negatives; some slides benefit from HDR
- Software has now advanced to be quite good for inverting color-negatives

My net: To get better images, I'd have to improve my negatives by better technique or moving to MF.

Your thoughts?

 

[ColSebastianMoran]

Camera scan from the 80's:

 

[madNbad]

The latest addition! Found a Ambico compendium shade on craigslist for twenty bucks, thought it might help with extraneous light.



Ambico Compendium Shade

 

[ColSebastianMoran]

Jackson Hole WY, 1992, this barn is one of the classic shots.

 

[ColSebastianMoran]

An iPhone is a digital camera, right? Here's another approach for camera-scan of color-negatives:
- For 35mm mount a good macro add-on lens to your 12MPx phone. Exolens by Zeiss Macro and Moment Macro are excellent; most others are toys. For 120, just use iPhone.
- Shoot the color-negative with Lightroom Mobile, creates a raw file (DNG)
- Sync that to Lightroom on your desktop
- Then convert with Negative Lab Pro or software of your choice.

Here's an example: iPhone 6s (12MPx), Exolens by Zeiss, Negative Lab Pro. Makes a good 12MPx image, not as good as a 24MPx or larger digital camera, but convenient. When a non-photog friend asked me to set her up to digitize her film collection, this is the gear I recommended, Zeiss macro, a light box, and a holder.

 

[maddoc]

I originally wanted to try the iPhone solution but the Zeiss Exolens was designed for the iPhone 6s. I did not find a modified version for the 11 Pro as of yet.

 

[peterm1]

I was hoping someone would be able to advise me. Bear with me while I explain.

I am in the process of planning to build a scanning outfit using a digital camera as I have quite a few slides that I have decided I would like to digitize. The outfit I am planning will have the camera mounted horizontally rather than in a vertical mount pointed downwards as I do not have a convenient way of doing that - my tripod does not really allow it.

I have a several macro lens options that I could use. An SMC Takumar 50mm f4; a Nikon 50mm f2.8 (both the first model AF one and the AIS version); a Nikon 105mm f4 AI and a Konica 55mm f3.5 macro MF. And a Tamron 90mm f2.5 which has a good reputation though even if it lives up to that enlargement ratio.

At this stage I was thinking that the best bet might be the Takumar - mainly because I already own a set of old but perfectly serviceable M42 extension rings and all of the above lenses only provide 2:1 magnification other than the Tamron which provides even less. Hence extension rings are an advantage as I should be able to reach 1:1 with them.

I plan to use my Sony NEX 7 for the scanning as it has several potential advantages - the rear screen tilts which would be useful in practice as I anticipate using it for focusing and framing rather than the viewfinder for obvious convenience reasons; the camera has a 24 megapixel sensor and that resolution could be handy (if not essential) and it has an APS sized sensor. I have a full frame Nikon but it only has a 12 megapixel sensor and the allure of a sensor with a larger pixel count seems too strong a pull to deny. So I am inclined to provisionally discount that camera - though subject to what others say.

All of the above is by way of background. Having said the above, the thought now occurs that I could, I suppose, use a micro four thirds camera - the Olympus OM D EM 5 for example which has a 16 megapixel sensor. Not as big as the Sony but not as small as the Nikon either. But the smaller physical size of its sensor may have an advantage.

Now here's my rationale for thinking along those lines. If I used a M4/3 which has a physically smaller sensor than the Sony I may be able to get by without using extension tubes - it is the need for the tubes that constrains me to use the Takumar macro lens rather than any of the others if 1:1 or something close to it is required. If I no longer need it as a 2 to 1 macro will pretty well fill the frame with the image (on a smaller sensor) then I might be able to use one of the other lenses if it provides a better optical result.

Of course an M4/3 camera also has a different aspect ratio in its sensor than does a conventional film slide which of course may create a framing issue that might negate any advantage of using a camera with a smaller sensor. So there is that to think about. I suppose I could cope with that by shooting slightly wider than the actual slide film / image thereby capturing some of the plastic slide holder - then cropping to get rid of that in post.

So I am just checking - does the above thinking about the potential advantages and disadvantages of M4/3 sound right to those who have been down the path of using a camera to scan slides? I think it does but may be missing something. I suppose I will sort it out with a trial run anyway but it would be nice to know what I am dealing with in advance.

 

[Ricoh]

Peter, for camera scanning I use an M240 on a BEOON that can reproduce negs and slides 1:1. I’ve also used an EM5 16MP on the same set up, mainly for B&W.
If 4x3 represents 16MP (4x3 being 12 units of area) then 3x2 (being 6 units of area) this will reduced the usable scan to 8MP, assuming my assumptions are correct.

 

[ColSebastianMoran]

I originally wanted to try the iPhone solution but the Zeiss Exolens was designed for the iPhone 6s. I did not find a modified version for the 11 Pro as of yet.

Gabor, there is no such version. The Moment Macro is also excellent, but the ver 2 I have will not capture the whole 35mm frame. It covers the short dimension of 35mm image, but is not wide enough. The Zeiss has the advantage of capturing the whole frame on the 12MPx 6s. I decided to get a 6s just for this use.

 

[ColSebastianMoran]

Peter -

I use Sony for camera scanning. Some comments:

- I use an A6000 (similar to your NEX 7) and a Sony FF body. I like the focusing with magnification.

- If your Nikkor is the 55 f/2.8 Micro-Nikkor, that's terrific. The Tamron 90 f/2.5 also has a very good rep. I tried the 105 f/4 AI and found it weaker than the two above. Don't know the Takumar or Konica.

- How are you going to setup your horizontal rig? See this page with a rig by Peter Krogh. He sells these, or perhaps it gives some ideas.

https://www.damuseful.com/collections/scanning-gear/products/nikon-ps-5-rail-system

I have a several macro lens options that I could use. An SMC Takumar 50mm f4; a Nikon 50mm f2.8 (both the first model AF one and the AIS version); a Nikon 105mm f4 AI and a Konica 55mm f3.5 macro MF. And a Tamron 90mm f2.5 which has a good reputation though even if it lives up to that enlargement ratio.

At this stage I was thinking that the best bet might be the Takumar - mainly because I already own a set of old but perfectly serviceable M42 extension rings and all of the above lenses only provide 2:1 magnification other than the Tamron which provides even less. Hence extension rings are an advantage as I should be able to reach 1:1 with them.

I plan to use my Sony NEX 7 for the scanning

 

[peterm1]

Peter -

I use Sony for camera scanning. Some comments:

- I use an A6000 (similar to your NEX 7) and a Sony FF body. I like the focusing with magnification.

- If your Nikkor is the 55 f/2.8 Micro-Nikkor, that's terrific. The Tamron 90 f/2.5 also has a very good rep. I tried the 105 f/4 AI and found it weaker than the two above. Don't know the Takumar or Konica.

- How are you going to setup your horizontal rig? See this page with a rig by Peter Krogh. He sells these, or perhaps it gives some ideas.

https://www.damuseful.com/collections/scanning-gear/products/nikon-ps-5-rail-system

Thank you Col Moran. You know you are really so much nicer than that phony Sherlock Holmes had us believe. 🙂

The Micro Nikkor 55mm f2.8 has been a solid contender. So thanks for confirming how good it is. I have not used it for micro/macro work to date as it this type of photography is not really my forte nor interest. I bought the lens more for its stellar reputation rather than because I planned to use it for its primary intended purpose - such lenses often serve well as normal lenses too.

As to the the rig I was planning, what I have had in mind to date has been a much more Heather Robinson (UK) / Rube Goldberg (USA) (or perhaps Forrest Gump - if you prefer that analogy) rig than the one you linked too. Thought I do rather like the idea of using rails to move the camera about on its tripod. But no, what I was planning was nothing more than a kind of wooden box with a spare LED downlight left over from changing lighting in my house at one end shining though a hole cut in the side of the box, a home made slide holder in the middle (appropriately baffled in order to stop light shining anywhere except where it is needed so as to to prevent lens flare) and a camera at the other end with that camera most likely on a tripod and its lens peering through yet another hole on the side opposite the light source. I figured that if I got the geometry right it would be simple but effective.

Truth is though, today I was talking to my camera guy about the project and he disappeared momentarily and then came back out of the back room with an as new Pentax bellows outfit with slide duplicator attachment and offered it to me at a good price. Impeccably built it suits my purpose so I weakened and bought it - that way I can focus my energies on copying the slides not on building the rig. (Or so I tell myself). The whole kit mounts on a tripod and a camera mounts on the duplicator / bellows outfit itself. The camera mount is M42 and according to the camera store guy I do not use need to use an actual micro / macro lens with it given its just a more flexible equivalent of an extension tube. Apparently a 50mm approx lens is desirable so I will try it with something like a late m42 Pentax SMC Takumar 55mm f1.8 which I know and respect as an excellent lens of its type. Failing that I have a swag of other 50mm lenses in m 42 mount mostly Zeiss Jenas etc. Hopefully something will work.

I tried an experiment last night with a slide resting vertically against a light box and my camera hand held but suitably braced and pointing at it as near as I could, perpendicular to the plane of the slide. The lens was a macro one - an SMC Takumar 50mm f4 which I mentioned before. I did this mainly to get an idea of what is needed in terms of working distance etc if I built the above rig. While some some shots were OK sharp some looked horrifically blurry when enlarged from the sensor. No doubt some will sharpen up with the above kit as there is no scope for camera movement to blur the shot. But I still have a sneaking suspicion that some of my slides which looked OK sharp on the hand held slide viewer will in practice be blurry as hell when they are scanned using my digital camera. Life is full of disappointments! 🙂

 

[madNbad]

Peter, remember to add a spacer between the adapter and the bellows to compensate for the crop factor.

 

[peterm1]

Peter, remember to add a spacer between the adapter and the bellows to compensate for the crop factor.

Madnbad thanks for this but can you explain a bit more please - I am not quite with you.

 

[madNbad]

The bellows are designed for a 1 to1 reproduction with 135 film or a full frame sensor. To capture the entire negative or slide you need to reduce the effective focal length due to the crop factor of the sensor. You would think the bellows would allow for that compensation by the ability to tho both lengthen and shortening but because they are attached to the lens, instead of the 50 mm you’re starting with a 75 mm. Somewhere in this thread is a better explanation, it is something to be aware of when using a bellows with a crop sensor.

 

[peterm1]

The bellows are designed for a 1 to1 reproduction with 135 film or a full frame sensor. To capture the entire negative or slide you need to reduce the effective focal length due to the crop factor of the sensor. You would think the bellows would allow for that compensation by the ability to tho both lengthen and shortening but because they are attached to the lens, instead of the 50 mm you’re starting with a 75 mm. Somewhere in this thread is a better explanation, it is something to be aware of when using a bellows with a crop sensor.

OK thanks for this. I guess the full implications will become real to me as I start to use it. But that's a good start.

 

[ColSebastianMoran]

Here's the bellows issue for a 50mm lens for cam-scan of 35mm to APS:

This photo shows a 50 El-Nikkor on bellows focused to 1:1, perfectly setup for cam-scan to a FF body. . It's a Sony body, with an adapter that preserves infinity focus, just like a Nikon body. Notice the bellows is almost at minimum extension. Same will apply with the 55 f/2.8 Micro (with focus ring set to infinity).

Problem is you cannot with shorten the extension further to focus at 1:1.5 (instead of 1:1) to capture whole 35mm frame to APS.

In other settings, with other bellows/lens combination, can't get the film/slide holder far enough from the lens.

Shown is a Nikon PB-4. Same applies, I believe, with the other Nikon bellows.

Key point: check-out ability to focus and position film/slide as desired with your chosen lens and bellows.