Duplicating slides with a DSLR
Hey all,
What's up? We're coming to the end of Month #6 of entertainment lockdown here in L.A., which means that there are pitifully few events to cover around town. So, with plenty of time on my hands, I've started a new experimental project: digitally backing up my film library. Now, normally this would be done with a dedicated film scanner. However, the number of good dedicated film scanners out there has always been small, and let's face it--the highest quality ones have been discontinued for manufacturing for years now. The newer ones just don't really cut it, and the scanning software that most of them use are years old now. The remaining scanners on the market are mostly flatbed scanners, which are fine for documents, and even medium format film, but not ideal for 35mm film, which the majority of my old library is made up of. And of course, drum scanning--the ideal form of scanning for any film--is prohibitively expensive.
So, I began looking for alternatives, and a while back, I discovered a small community of people online who have hit upon such an alternative: using slide duplicators mounted onto a macro lens on a DSLR, and shooting the duplication of the slide in RAW format. Now, historically, using a slide duplicator in a film camera--which the duplicators were originally designed for--was seen as a second-rate way to copy film. However, a DSLR is an entirely different beast, particularly today's more advanced models. There are a number of reasons for this, the two major ones being (A) DSLRs today produce file sizes that are often higher than the upper limits of many film scanners, and (B) shooting in RAW can allow for a greater dynamic range than some scanners can deliver. By working in a RAW file, you have greater initial flexibility than you do when working with a film scan, drum scans excepted.
So, if you have a camera with a nice, big sensor--I'm thinking 24 MP and up--theoretically, you could make an image file that would be quite suitable for most practical applications for photographs. It wouldn't be big enough for billboards and such, but let's face it, how many pics of yours are you realistically going to put on a billboard? Not many, I'm guessing.
But, would this actually work? Luckily, one of the better online sources for information on this subject was provided by a gentleman named Jurgen Becker, on his web site "Through The F-Mount":
http://www.throughthefmount.com/articles_tips_digitise.html
Jurgen's solution to the problem involved him using a classic Micro-Nikkor 55mm f/3.5 lens, coupled with an extension tube. And as luck may have it, I happen to possess such a lens and tube in my arsenal. There are other lenses available for Nikon that will work, such as their current Micro-Nikkor 60mm macro, or their Micro-Nikkor 55mm macro.
The setup required was actually pretty minimal: basically pointing the camera lens (with Nikon's ES-1 adapter attached) at a light source. This could either be a continuous light panel, like an LED one, or a flash with a light diffuser attached. I chose the latter, since I could put the flash quite near the end of the lens, which meant that I could use a low power setting on the flash. Between the ES-1 and the flash, the light becomes very even, as well as bright. I used a remote trigger to reduce movement as much as possible.
One of the things that's going to directly affect the quality of your image is the cleanliness of your DSLR's sensor--not a problem that one normally encounters with a film scanner, to be sure. So, make sure that you have a clean sensor at the get-go, so as to save yourself loads of post-production time. And naturally, you need to do the same with your slides.
So, after acquiring an ES-1, I adapted Jurgen's setup and began experimenting. I quickly found that a good base to start from was shooting with my Nikon D750 at ISO 160, 1/200 second, in RAW+F mode. My old Nikon SB-900 flash was used in Remote mode at 1/16 power setting, and the Nikkor was set at f/8 or f/11, depending on the slide in question. I used a combination of histogram and monitor inspection to determine which to go with. One note: since the Nikkor is a manual focus lens, and since not all slides are aligned precisely alike (particularly ones in plastic mounts), it's a very good idea to check the focus on each slide before shooting.
I found out that the film's image doesn't completely cover all of the D750's full-frame sensor, but it comes close--the cropping is pretty minimal. I expect that a shorter extension tube could address that issue. (Also, a smaller APS-C sensor in the camera would likely work quite well) After processing the cropped RAW image in Photoshop, the image usually comes out to a good 12"x18" at 300 dpi, or about 3600 x 5500 pixels, which would be a perfectly acceptable resolution for a conventional film scan after being downsized from its initial super hi-res scan. I prefer Photoshop over Lightroom for work like this, because much of my early electronic music work was all over the place in all sorts of environments, exposure wise, so most of the time I need to work on single images individually, which Photoshop is ideally suited to.
As I began editing different sorts of images, I did a couple of scans of photos that I had used in my photo book Dancefloor Thunderstorm. When I compared the new duplications with the older scans, I noticed that the resolutions were actually pretty similar. If I had gone to town on the slide duplication, I probably could have made the images nearly identical. Given the high quality of the imagery in the book, it was a pleasant surprise to realize that I could get similar results from this duplication technique.
Now, make no mistake, this is not a panacea for your image library. It's not a cure-all. However, it is a good way to produce quite workable image files that will be suitable for a wide variety of uses. The fine details are normally captured quite well, and most of the time, the noise levels in the shadow areas of the converted RAW images in my pictures were perfectly acceptable. Some required more work than others, but then, such is life in the photography world.
At this point, I thought I'd go for something a bit different, so I dug into my archives and pulled a few almost-at-random slides from a European tour I did in 2000. Most of this is material that's more conventional looking, so you'll be able to see the capabilities of this duplication technique nicely.
So to wrap things up, I would say that I'm initially quite pleased with the results I've been able to achieve thus far. I think it's fair to say that with practice, things will continue to get better. This could be the solution to a number of storage and archiving issues.
More to come later. Thanks for reading.
What's up? We're coming to the end of Month #6 of entertainment lockdown here in L.A., which means that there are pitifully few events to cover around town. So, with plenty of time on my hands, I've started a new experimental project: digitally backing up my film library. Now, normally this would be done with a dedicated film scanner. However, the number of good dedicated film scanners out there has always been small, and let's face it--the highest quality ones have been discontinued for manufacturing for years now. The newer ones just don't really cut it, and the scanning software that most of them use are years old now. The remaining scanners on the market are mostly flatbed scanners, which are fine for documents, and even medium format film, but not ideal for 35mm film, which the majority of my old library is made up of. And of course, drum scanning--the ideal form of scanning for any film--is prohibitively expensive.
So, I began looking for alternatives, and a while back, I discovered a small community of people online who have hit upon such an alternative: using slide duplicators mounted onto a macro lens on a DSLR, and shooting the duplication of the slide in RAW format. Now, historically, using a slide duplicator in a film camera--which the duplicators were originally designed for--was seen as a second-rate way to copy film. However, a DSLR is an entirely different beast, particularly today's more advanced models. There are a number of reasons for this, the two major ones being (A) DSLRs today produce file sizes that are often higher than the upper limits of many film scanners, and (B) shooting in RAW can allow for a greater dynamic range than some scanners can deliver. By working in a RAW file, you have greater initial flexibility than you do when working with a film scan, drum scans excepted.
So, if you have a camera with a nice, big sensor--I'm thinking 24 MP and up--theoretically, you could make an image file that would be quite suitable for most practical applications for photographs. It wouldn't be big enough for billboards and such, but let's face it, how many pics of yours are you realistically going to put on a billboard? Not many, I'm guessing.
But, would this actually work? Luckily, one of the better online sources for information on this subject was provided by a gentleman named Jurgen Becker, on his web site "Through The F-Mount":
http://www.throughthefmount.com/articles_tips_digitise.html
Jurgen's solution to the problem involved him using a classic Micro-Nikkor 55mm f/3.5 lens, coupled with an extension tube. And as luck may have it, I happen to possess such a lens and tube in my arsenal. There are other lenses available for Nikon that will work, such as their current Micro-Nikkor 60mm macro, or their Micro-Nikkor 55mm macro.
The setup required was actually pretty minimal: basically pointing the camera lens (with Nikon's ES-1 adapter attached) at a light source. This could either be a continuous light panel, like an LED one, or a flash with a light diffuser attached. I chose the latter, since I could put the flash quite near the end of the lens, which meant that I could use a low power setting on the flash. Between the ES-1 and the flash, the light becomes very even, as well as bright. I used a remote trigger to reduce movement as much as possible.
One of the things that's going to directly affect the quality of your image is the cleanliness of your DSLR's sensor--not a problem that one normally encounters with a film scanner, to be sure. So, make sure that you have a clean sensor at the get-go, so as to save yourself loads of post-production time. And naturally, you need to do the same with your slides.
So, after acquiring an ES-1, I adapted Jurgen's setup and began experimenting. I quickly found that a good base to start from was shooting with my Nikon D750 at ISO 160, 1/200 second, in RAW+F mode. My old Nikon SB-900 flash was used in Remote mode at 1/16 power setting, and the Nikkor was set at f/8 or f/11, depending on the slide in question. I used a combination of histogram and monitor inspection to determine which to go with. One note: since the Nikkor is a manual focus lens, and since not all slides are aligned precisely alike (particularly ones in plastic mounts), it's a very good idea to check the focus on each slide before shooting.
I found out that the film's image doesn't completely cover all of the D750's full-frame sensor, but it comes close--the cropping is pretty minimal. I expect that a shorter extension tube could address that issue. (Also, a smaller APS-C sensor in the camera would likely work quite well) After processing the cropped RAW image in Photoshop, the image usually comes out to a good 12"x18" at 300 dpi, or about 3600 x 5500 pixels, which would be a perfectly acceptable resolution for a conventional film scan after being downsized from its initial super hi-res scan. I prefer Photoshop over Lightroom for work like this, because much of my early electronic music work was all over the place in all sorts of environments, exposure wise, so most of the time I need to work on single images individually, which Photoshop is ideally suited to.
As I began editing different sorts of images, I did a couple of scans of photos that I had used in my photo book Dancefloor Thunderstorm. When I compared the new duplications with the older scans, I noticed that the resolutions were actually pretty similar. If I had gone to town on the slide duplication, I probably could have made the images nearly identical. Given the high quality of the imagery in the book, it was a pleasant surprise to realize that I could get similar results from this duplication technique.
Now, make no mistake, this is not a panacea for your image library. It's not a cure-all. However, it is a good way to produce quite workable image files that will be suitable for a wide variety of uses. The fine details are normally captured quite well, and most of the time, the noise levels in the shadow areas of the converted RAW images in my pictures were perfectly acceptable. Some required more work than others, but then, such is life in the photography world.
At this point, I thought I'd go for something a bit different, so I dug into my archives and pulled a few almost-at-random slides from a European tour I did in 2000. Most of this is material that's more conventional looking, so you'll be able to see the capabilities of this duplication technique nicely.
A soccer crowd in Amsterdam, for the Euro2000 soccer championship. Sensia 100 |
General Montgomery's command tank, Imperial War Museum, London. Look at how crisp the rivets and chain on the right are. Sensia 200. |
Swiss cow. |
French Bastille Day celebration, Arc de Triomphe, Paris, France, 2000. |
So to wrap things up, I would say that I'm initially quite pleased with the results I've been able to achieve thus far. I think it's fair to say that with practice, things will continue to get better. This could be the solution to a number of storage and archiving issues.
More to come later. Thanks for reading.
Comments
Post a Comment