[fotoreporter1975]

Staff note: This post may contain affiliate links, which means Pentax Forums may earn a small commission if a visitor clicks through and makes a purchase. If you would like to support the forum directly, you may also make a donation here.

---

Hi,
I love film, I even develop color negatives and slides by myself (what a pain…) because nobody does a decent job in my area, not anymore.

But the biggest problem for me has ever been scanning. I use from 35mm to 13x18cm, and my options were all pretty much not cheap enough or very bad quality wise.

I read various times about the possibility to use a digital camera instead of a scanner, but the results I got were always poor because of lack of alignment and low general resolution.

Now I think I found the proverbial (well, at least here in Italy!) "Columbus egg".

What you need:
- a digital camera with live view (slr, mirrorless, whatever)
- a lens hood of the right diameter for your lens and (crucial) of the right height (I will explain shortly why); alternatively, if you decide to use an inverted lens, a set of macro tubes
- a lens that you can use in macro; I didn't say a macro lens because till now I've had the best results with an old 35mm f/2 Nikkor O (yes, the pre-Ai kind) inverted (but I'm waiting for a Pentax 50/4 Macro M)
- a slide viewer or, if you don't have one yet and you are a bit capable around electricity, a transparent adapter for an old scanner. In this case just join the cable of the adapter directly to an appropriate transformer - please do not electrocute yourself!
- a software for panoramic photography that let you use the "matrix" or "multirow" template (if you shot Canon Photostitch came for free; a really good and cheap alternative, at least on Mac, is Panorama Maker; a bit complex, but free and of extreme quality, Hugin came for all operating systems)

Then you:
- put the slide viewer flat on a table;
- tape down the film flat on the viewer surface;
- put the camera with the macro lens + hood (or with the reversed lens + tubes) directly on top of the film, so you'll get rid of alignment problems (and no, the film doesn't get scratched)
- focus with the live view; if you reproduction ratio is good enough you can focus directly on the film grain!
- set the exposure in manual mode, to ensure that all the shots are the same
- start taking multiple shots
- assemble them in the panoramic software

How many shots will depend of the reproduction ratio you (and your lens) will use. Generally I use a 1:2 enlargement ratio on medium and large format film and a 3:1 ratio on 35mm and I get more or less this results:
- 35mm = 4 / 6 shots
- 4,5x6 = 4/ 6 shots
- 6x6 / 6x7 = 6 / 8 shots
- 4x5" / 13x18cm = 20 / 30 shots

Taking enough shots for, say, a 6x6 frame takes less than 15 seconds, and this because I use a 2 seconds self timer to avoid vibrations!

On the bottom you can see a quick example of the quality you can get; it's a comparison using the same 4,5x6cm negative between the Epson 700v (film holder height calibrated) and this method (using a 1:3 ratio). And please, remember that with this digital camera method you can still go to 1:1 or more and extract more detail.

Just to be clear: this are 100% corner crops, and while the Epson has been sharpened the Canon not!

If you want to see more examples, including a comparison against a Dainippon Screen drum scanner, please take a look at this post on my blog:

http://www.addicted2light.com/2012/11/23/best-film-scanner-canon-5d-mark-ii-...-vs-epson-v700

Attached Images

   

---

Last edited by fotoreporter1975; 11-24-2012 at 02:36 AM. Reason: the post did not show the album images

[germar]

Thank you for this interesting post.

I'm afraid I got lost at the point you assemble in pano software. Are you saying that you are taking individual shots of sections of the film frame and then assembling them in software? Or are you taking multiple shots of the whole negative frame and the pano is doing something else.

I'd like to see photos of the process, as I'm struggling to see how setting the camera lens down on the negative itself wont scratch the emulsion.

Neat concept and while it might offer an edge over an Epson V700, it seems like a process meant for a single keeper frame rather than replacing your entire workflow. Or is it easier than it seems?

[fotoreporter1975]

Ok, first a couple photos of the setup.

From top to bottom:

camera
lens + hood
slide viewer

You need to take individual shots of sections of each frame, then assemble the various shots in a panoramic software to obtain the whole frame again. And I scanned almost 300 negatives and slides so far, but none of them has been scratched till now. The trick is to put the negative emulsion (the opaque side) down. This way not only the protection layer (the shiny one) will be on top, in contact with the lens hood, but you will also avoid the formation of newton rings between the film and the slide viewer.

Last, the funny thing is that it's much, much more fast than scanning films on a flatbed scanner (or on a Coolscan). I need between 15 and 20 seconds to shoot an entire 6x7 frame, plus a minute or so to mount the sections in the panoramic software. In the meantime the Epson is still warming up the lamp...

[germar]

Thanks for this additional info and photos. I might have to give this a try.

[fotoreporter1975]

And here an overview of the shooting setup: on the left how you need to take the shots, on the right the results.

Please keep in mind that this was one of my fist attempts, so I overshoot quite a bit. If there aren't big skies or large areas of detail-less water you'll need only 4 shots.

[LesDMess]

To be sure that I understand, what resolution setting for the Dainippon and how many images will you need from the 5D MKII to match this?

[fotoreporter1975]

To be sure that I understand, what resolution setting for the Dainippon and how many images will you need from the 5D MKII to match this?

If I got this right I think that the resolution was set at about 2400 with the Dainippon (btw, not the same 2400 of a flatbed…).

To match this, for a 5x7" / 18x24cm sheet film, you'll need 24 shots more or less. It took me less that 2 minutes to take the shots, and a bit more that 10 minutes for the computer to crank up the final image.

To have the same image drum scanned took: ship it back and forth to the scan service; wait three weeks; pay a fourth of what I paid for my large format camera!

The Dainippon goes up to 12.000 dpi, but then you'll need very deep pockets to pay for the scan (the scans from a drum are usually sold by their megabyte "weight") and a supercomputer to edit and store the results.

Don't get me wrong: if you happens to live near a drum scan service (so you don't have to ship your negatives) and / or you can fetch good prices and a good service by all means go for it; it's still the best way in the quality / effort equation. But if you don't then with this technique you can, at the price of a little effort, obtain results indistinguishable or even better than that of the drums.

[LesDMess]

So about 24 shots to achieve detail resolution at 2400dpi.

What about latitude - if it is a high contrast scene, how many shots do you have to shoot given a drum scanner reaches into the darkness and controls bright skies?

[fotoreporter1975]

24 shots for a 18x25cm! With medium format films I use just 4 / 6 shots. More, the resolution depends by the reproduction ratio you decide to use: with a 2:1 or 3:1 ratio, for example, you extract practically every detail there is on film.

The latitude is better than what I've seen on most scanners (Coolscan etc., not only flatbed). I think it's on par with the drums, but on drums I have only scanned black and white films, that pose less troubles than slides.

As today, after several hundredth files "scanned", I found only 4 or 5 that have been required special attentions (developing or the raw file in two sets, one for the shadows one for the light, and merge of the files before the use of the panoramic software). We're talking about extremely backlit shots of the sun coming up behind the mountains, with mist in the valley.

Even so, the special procedure has been necessary only to extract more detail that was visible on film in normal observation. To be clear: those were old slides that I saw a lot of times projected, and I never suspected there was so much more detail waiting to be recovered in their dense shadows.

[LesDMess]

Do you have a full res sample available from all three scans?

[fotoreporter1975]

Do you have a full res sample available from all three scans?

Sorry, but the files are really huge… Even a jpg of the drum scan alone is about 100 Mb! Besides, once having the photos drum scanned I trashed every time the epson or coolscan versions to make space. This is why I compared it only against the multishot technique.

But if you want I can send you in private (if it's technically possible?) or via mail a link to two full resolution, medium format images: one scanned with an Epson V700, the other with the multishot. Just don't take into consideration the colors, because the photo has been shot on color negative, and I'm struggling a bit to find the perfect curve (i'm mostly a black and white guy). The scene, however, resembled much more closely the warm rendering of the multishot variant.

[jt_cph_dk]

Hi, It is an interesting rig you got there and you certainly yield sharp copies with it. Have you considered/tried using a backlit bellows with film/slide holder or a copy (tripod) stand on the slide viewer, then mounting your Canon (with a macro lens for a corrected, flat field) on one of these? A 24mp camera should give good enough resolution (equal to about 2400ppi on a scanner) and the bellows or stand would make it possible to focus on the grain. I realize that by taking more shots (4-6 per photo) you are getting more resolution, and maybe more than you will ever need, but you also have a more unstable (lens hood balance?), complex workflow (stitching etc.). For color photos, I think, the sensor is very important (for instance to optain ‘real‘ reds). Any digital camera won´t do. And probably the expensive scanners will win on this point? I´m guessing since I don´t know much about sensors. More than sharp focus (as percieved on screen or in normal print sizes), it´s the digital noise (in the shadows), the narrow latitude and limited dynamic abilities of most digital cameras, that anoys me. And I can´t figure out, if more shots helps to solve these problems? It should, since you optain the high dynamics on the film and then copy that digitally with your Canon, right? But that goes for a scanner as well. Hmm. Thank´s for posting. Interesting to see the rig. Regards

[fotoreporter1975]

Hi, It is an interesting rig you got there and you certainly yield sharp copies with it. Have you considered/tried using a backlit bellows with film/slide holder or a copy (tripod) stand on the slide viewer, then mounting your Canon (with a macro lens for a corrected, flat field) on one of these? A 24mp camera should give good enough resolution (equal to about 2400ppi on a scanner) and the bellows or stand would make it possible to focus on the grain. I realize that by taking more shots (4-6 per photo) you are getting more resolution, and maybe more than you will ever need, but you also have a more unstable (lens hood balance?), complex workflow (stitching etc.). For color photos, I think, the sensor is very important (for instance to optain ‘real‘ reds). Any digital camera won´t do. And probably the expensive scanners will win on this point? I´m guessing since I don´t know much about sensors. More than sharp focus (as percieved on screen or in normal print sizes), it´s the digital noise (in the shadows), the narrow latitude and limited dynamic abilities of most digital cameras, that anoys me. And I can´t figure out, if more shots helps to solve these problems? It should, since you optain the high dynamics on the film and then copy that digitally with your Canon, right? But that goes for a scanner as well. Hmm. Thank´s for posting. Interesting to see the rig. Regards

Hi, thanks for your interest. Let's address a question at a time:

1) with a bellows + film holder setup, as you correctly guessed, you are limited to the camera resolution. Not a big deal if you got a hi-res camera, but with multiple shots you can extract all the details in the film even, say, with an old 6 megapixel camera; it will only take more shots, not a new camera

2) if you use, as I do, a good quality metal lens hood the assembly camera+lens+hood is tripod-stable; it's the camera's mass that works in favor of this setup, pushing down and so helping the stability, instead of acting against it

3) the digital noise - at least with a full frame, but I would guess that this is the case even with an aps-c - is invisible, lost in the grain clumps, even with really fine grained film (Adox 25 iso or Ektar 100) when you pull the curves to bring out the details in the dense shadows; the important thing, like always, is "expose to the right" of the histogram, to exploit all the latitude. More, all negative films and most slides (Velvia and few others excluded) have really little dynamic range, so don't give you any trouble.

4) I was too concerned about color reproduction, even if, like I said, I'm mostly a b/w aficionado. But then I "scanned" this way an old project that consisted of photos with almost out of gamut (adobe rgb gamut) bright reds, oranges and yellows. I scanned this before with various film scanners (some problem mostly with the yellows) and with the Epson (the reds were mostly gone); the camera nailed the tonalities. And this was the first setup that almost (it required a little pull to the blue curve) nailed the peculiar shade of Velvia green of some of my slides. The paramount thing to avoid headaches is to shot with a custom white balance tuned to the light source you use; if you do so than the files require little or none post processing color-wise.

[hinman]

I thank you for your posting. I am amateur in films and scanning. I watch this with great interest as I enjoy 120 films while I struggle with scanning at home with a Canon 8800F flatbed scanner. And I do wish to have simple scanning alternative with both 120 and 135 films.

Can I ask if this light-box or two slide viewers I found in ebay can serve the purpose of the flat table top slide viewer that you have mentioned? I may mis-understand your slide viewer as I do see a large projected image on your slide viewer and I wonder if a normal film or transparency viewer can do the job.

1. Gagne Porta-Trace Light Boxes with Stainless Steel Frames - BLICK art materials
2. Slide Viewer for Fuji Kodak 135 120 4x5 film #J016 | eBay
3. MEDA Light Slide Viewer for Fuji Kodak 135 120 film SV-9 J018 | eBay

I don't know enough on the use of light-box but the first option seems to be flexible enough to have a larger surface to work with multiple film strips not requiring the cutting of film strip while laying flat on the top surface. And I wonder use of anti-newton glass will help in the case of the 35mm film strip as I am not sure if the curling of the smaller films will make it harder to take multiple shots.

Your inputs and clarification can help me greatly,

Regards,
Hin

[fotoreporter1975]

I thank you for your posting. I am amateur in films and scanning. I watch this with great interest as I enjoy 120 films while I struggle with scanning at home with a Canon 8800F flatbed scanner. And I do wish to have simple scanning alternative with both 120 and 135 films.

Can I ask if this light-box or two slide viewers I found in ebay can serve the purpose of the flat table top slide viewer that you have mentioned? I may mis-understand your slide viewer as I do see a large projected image on your slide viewer and I wonder if a normal film or transparency viewer can do the job.

1. Gagne Porta-Trace Light Boxes with Stainless Steel Frames - BLICK art materials
2. Slide Viewer for Fuji Kodak 135 120 4x5 film #J016 | eBay
3. MEDA Light Slide Viewer for Fuji Kodak 135 120 film SV-9 J018 | eBay

I don't know enough on the use of light-box but the first option seems to be flexible enough to have a larger surface to work with multiple film strips not requiring the cutting of film strip while laying flat on the top surface. And I wonder use of anti-newton glass will help in the case of the 35mm film strip as I am not sure if the curling of the smaller films will make it harder to take multiple shots.

Your inputs and clarification can help me greatly,

Regards,
Hin

Hi hinman, thanks to you for your interest.

First of all, to avoid confusion: on my picture there is not a "large projected image"; it's simply a large - 10x12cm / 4x5" - negative on a small - 18x24 / 5x7" - slide viewer!

Of the various slide viewers you've listed I'd avoid only the last one, the MEDA, because it seems kinda small. When you move the camera around on the film frame to shot the various sections you don't want to worry that it may fell off the viewer.

And I'd avoid to use anti-newton glass, or any glass at all. Any additional layer between the lens and the film makes me nervous, because it can cause reflections, attract dust, affect the resolving power etc.

I simply use two strips of paper masking tape, the kind that painters use to fix the protective plastic sheets in place and that leaves no residues, to hold the films flat. Works like a charm!

Now a bit of unrequested advice

If you want to try the technique without spending big money you can use the transparency adapter of a scanner. If you look closely enough you will find that this adapters (at least the Epson and the HP ones) have a sort of hive-like texture. This texture spreads any light coming from above pretty evenly; so you can use a transparency adapter (even a broken one) with the negatives taped down and just a couple of flashes or lamps pointed at the film frame sides. Actually this was the setup with which I conducted my first tests.

If you got any more doubts, please, don't hesitate to ask.