Film scanners are very good, but are also very slow. You may do well to average 10 slides per hour overall, so thousands of slides may take many months, and it is a good bet that you may never finish. The Nikon 5000 film scanner does have its SF-210 Auto Slide Feeder accessory ($450) for overnight runs of 50 slides, if it doesn't jam. It was one thing to sit down one evening with one roll of slides, but something entirely different to be facing a few thousand old slides.
I did scan the better candidates, but I kept putting off all notions of scanning all of them. What I finally did was to use a slide copy attachment on a digital camera with macro lens. Then hundreds, even a thousand slides per day is possible and was accomplished, including most post-processing. The camera is extremely fast. Perhaps this method is a rushed job, but the job does not happen otherwise. I am speaking of a DSLR with a good macro lens, and the optical quality can be superb, but rushing through thousands can miss a little. And there are other choices too, below. Regardless, the results seem plenty good enough, assuming the slides are decently exposed, and clean and in good condition. I'd guess you will likely resample all of them smaller to computer or TV screen size anyway. If you want to be able to view all your old slides again, then a digital camera can be a very fast way to do it.
A $60 Nikon ES-1 52 mm Slide Copy Adapter is shown, on a Nikon D70S DSLR with a 60 mm f/2.8 D macro lens (which focuses to 1:1 enlargement). To work on a DX camera, the setup as shown also requires an extra 20 mm extension tube (not included). $60 seems expensive for the slide holder, but the job it does is about priceless.
Image dimensions for 6 megapixels are roughly equivalent to scanning at 2200 dpi.
A 12 megapixel camera would compare to scanning at 3000 dpi.
24 megapixels is about 4200 dpi - computed as (6000 pixels of image width / 36 mm of slide width) x 25.4 mm per inch = 4233 pixels per inch.
36 megapixels is nearly 5200 dpi (but is still 7360 pixels / 36mm = 204 lines/mm, or "only" 102 line pair per mm... realistic for a good lens, maybe less likely in a slide).
These are maximums, assuming you are able to fill the frame with the slide image, but no cropping of slide.
This article is about using the ES-1. However, the ES-1 is not necessarily required, there certainly are other easy ways without it, or for larger film. Just put the camera on a tripod and aim it at the slide, lighted from behind. If using a longer macro lens, like 90 or 105 mm, then the ES-1 is not a choice anyway - the ES-1 would need unreal extension out to the slide holder. So, you can just devise some way to position the film at the right distance out in front of the macro lens. The slide holder will be the only hard part, but it need not be elaborate (you will finish all your slides pretty early, and have no more use for it.) The camera lens ought to be 90 degrees straight onto the slide. Devise some diffused way to light the film from behind it. Bouncing a flash on a large white card background should work very well (however then focusing in dim light without flash may be difficult). Some people use the illuminated white plastic slide-sorting-trays for this, simply photographing the slide laying on it. Or cut a slot in the top of a pill bottle to stand the slide up in. Or the ES-1 is a slide holder, which with the right lens, pretty much takes care of all of this, very conveniently. There is advantage of having the slide physically attached to the lens - there is no camera shake.
This Nikon 60 mm f/2.8 D AF macro lens is about $400, and there are other similar lenses. Low end DSLR need an AF-S lens to be able to auto focus, and this is an older lens (it is Auto focus on higher models, but not AF/S). There is now a newer 60 mm AF/S lens, and a Nikon 40 mm AF/S macro lens, both of which have shorter working distance in front of the lens, so the spacers may need a new value too. The ES-1 attaches to a 52 mm filter thread, so it should fit any brand of DSLR. My 60 mm lens has a 62 mm filter thread, so it needs the Nikon BR-5 Mount Adapter Ring, which is just a 62-52 mm filter step down ring (a nice one, with large knurled diameter). There are of course other thread adapters much less expensive. The ES-1 copy attachment is basically an empty tube or spacer. Two telescoping tubes actually, with a one inch length adjustment. It telescopes to hold the slide from between 45 mm to 68 mm in front of the lens filter thread. Again, the ES-1 is an empty tube, just a slide holder, and a light diffuser, and it has no optics in it. The macro lens does all of the optical work.
DX cameras: The problem is that for today's DX digital SLR with the 1.5 lens crop factor, the 35 mm slide is half again larger than the DX sensor. So 1:1 copy size is not appropriate for DX to copy slides. The 1.5x crop sensor now needs a smaller image, more like a 0.67 reproduction size (which is 1:1.5), to fill the larger slide onto the smaller sensor. The 60 mm lens view is equivalent of 90 mm on the DX body. Which means that the DX 1.5 sensor factor needs an additional spacer in front of the lens so the ES-1 can be adjusted to hold the slide farther out in front, to appear as the smaller 0.67 size, so it will not be cropped excessively.
I had an old obsolete Nikon K extension tube set (Not for modern DSLR), and these have both a 20 mm K5 tube and a 10 mm K4 tube with 52 mm threads. So I used the K5 tube shown (only the one K5 threaded tube, and NOT the rest of the extension set), which works great for ES-1 on DX with a 60 mm D lens. The K5 tube is a simple aluminum tube, 20 mm long, with 52 mm filter threads at each end.
Finding that extra extension is the problem. You can measure this to see what length your DX setup needs - see this link (next page) for how: See more about the necessary extra tube extension for DX DSLR (not an easy problem). But that next page does show a 21mm extension (for this 60mm D lens on DX) on Ebay for $13 from Hong Kong.
See the Nikon ES-1 instruction sheet.
FX cameras: The Nikon ES-1 was designed for full frame film bodies to copy mounted slides at 1:1 with a 55 mm f/2.8 macro lens. And the ES-1 instruction sheet says the 60 mm f/2.8 D lens gives 0.96 to 1.0 reproduction with the BR-5 mounting ring on a full frame camera.
At right is a FX D800 with 60 mm D lens using the ES-1 at its maximum extension (alone, with only the BR-5). Using less extension, for a closer enlarged cropped view, won't let this lens focus that close. So this existing view is it, but it seems very usable, if you crop each one (which you likely want to do anyway, in most cases). In addition to compositional improvements, cropping allows rotation to straighten and center, and 36 or 24 megapixels will allow a lot of cropping. For those all the same, Raw software could do this as a batch operation.
I measure the ES-1 with BR-5 to have a telescoping working distance range (in front of lens) of 50 mm to 74 mm. An added 10 mm extension puts the ES-1 about at the middle of its telescoping range (60 mm D lens on FX). But which still will not focus closer, so it still requires some cropping. Here is a good list showing working distance of most macro lenses. I have not tried it, but this source says the newer 60 mm AF-S lens has a 1:1 working distance of 50 mm (vs 71 mm for 60 mm D). Normally too close is not a plus for macro, but which implies it may be better suited for the ES-1. Or, you could use DX mode on the FX body, with the 20 mm extension, but which is many fewer pixels than the more mild FX cropping.
If scanning these old slides is your only goal, and assuming you already have the DSLR, and can find an extension tube for DX, you might compare the macro lens expense with a film scanner. The lens is not a film scanner of course, and a digital camera will NOT be suitable to copy color negative film, but it works for slides. The macro lens has many other photographic uses too - it has great value in its own right. The Nikon 60 mm macro lens is excellent for any close-up work, and I'd assume the other similar lenses are great too. I predict the macro would quickly become your favorite lens.
This ES-1 setup works very well for scanning mounted slides quickly - like magic after you get the hang of it. Truly fantastic for speed. The macro lens optical quality is exceptional, but the other aspects are maybe not truly optimum (haste, mounting, framing, etc), not the same as a real film scanner. But still rather easy, and which seems more than good enough for this purpose to recapture thousands of old slides for nostalgic purposes. There are some pluses too, besides the great speed, the macro lens is awesome quality. Frankly, due to the months of work that would be required on a film scanner, this job went years without happening at all.
Above is a sample image copied from a 1990 35 mm Kodachrome slide, using the ES-1 setup with the D70S. Click the image to see the full size 6 megapixel 3008x2000 pixel 900KB image file. The image is significantly larger than your monitor screen, and to see full size, you may have to save the larger image and view with an image editor, or you could turn off Automatic Image Resizing in your browser:
Image dimensions for 6 megapixels are roughly equivalent to scanning at 2200 dpi, and a 12 megapixel camera would compare to 3000 dpi. And 24 megapixels is about 4200 dpi - which is (6000 pixels of image width / 36 mm of slide width) x 25.4 mm per inch = 4233 pixels per inch (a maximum, assuming you are able to nearly fill the frame with the slide image).
The camera macro lens seems the obvious bet for superior optical quality. I normally use flash lighting, for convenience and consistency, but this one was shot with skylight (basically holding the slide up to the window light), and for this, the RAW file was set to Cloudy white balance. ISO 200, f/8, 1/20 second in A mode (skylight auto exposure varies with day and slide). This is a typical good slide, certainly not my worst one. More of my slides are marginal that I care to admit, but if you have decent slides, you can get good images this way too. If not so decent, then it should reproduce that too. :) Results are obviously good enough. And did I mention it is very fast?
Testing extremes perhaps, but here is the same slide copied with a Canon A620 PowerShot compact camera (point&shoot) in its macro mode. No extra attachment was used - its macro mode gets this close if zoomed to wide-angle. Click the image to see the 7 megapixel 3072x2304 pixel 1MB image file (compressed more here, camera JPG file was 3.1MB). Pixel dimensions are roughly equivalent to scanning at 2500 dpi. This was a quickly kludged setup for the one image here. (My method: keep piling on stuff to solve the next immediate problem). The camera was on a tripod. The slide was literally standing up on edge on top of a light stand pole, held with a piece of tape. The slide was only about 1/2 inch from the lens, and lighted from the rear of course. This light was a 150 watt household incandescent lamp (possibly 2900K?) in a ten inch clamp-on utility reflector on a light stand (about 15 inches from slide), through a plastic Tupperware tray (yet another light stand) covered with a white bed sheet to diffuse it sufficiently (this lighted area should be a couple of feet wide, the slide at 1/2 inch is a wide angle situation). Camera mode A at f/4 was set to Incandescent white balance. The JPG was a little blue, and was adjusted here with -Blue and +Red. Auto exposure was ISO 100 and 1/80 second (time delay shutter to let camera stop shaking). This camera takes 4:3 pictures, but the slide was 3:2, so the ends are cropped. Or, a little more distance would have made the image smaller so it would all fit, and then it could have been cropped to 3:2. The top edge ended up not quite cropped enough, due to the barrel distortion shown. A straight edge held to the top railing on the right shows a similar bow, which no one would notice. Considerable vignetting (dark corners). This is a pretty extreme situation for the little compact camera lens. Not sure you would actually want to try this, but it can work. I did feel the very strong need for a convenient slide holder. But it is still good to be able to see the old slides again, and the point is that many things are possible.
Many other methods of holding and illuminating the slide are certainly possible. You just need a light and diffuser behind the slide, and a camera and macro lens in front of it. Any macro lens capable of 1:1 will do it this way. The common way places a lighted white paper background a foot or two behind the slide, with the camera and macro lens on a tripod in front. Slide holder could be a plastic pill bottle screwed to a board, with a slot cut at top to hold the slide standing up. Camera might even be mounted on the same board (with an ordinary 1/4-20 NC screw). Or, simply standing the slide on a regular lighted slide sorting tray is basically the same thing.
Inexpensive option: A friend bought this inexpensive Wolverine F2D14 14 MP 35 mm Slides and Negatives to Digital Image Converter (which basically has a 14 megapixel digital camera mechanism inside), and is pleased with it for slides - less so for color negatives, which is asking a lot from this method.
There are slide copier attachments that are designed to fit over the lens of compact and DSLR cameras. I have no experience with these, but they are a slide holder and a +10 diopter close-up lens (an extremely strong magnifying glass in front of the camera lens). These copiers rely on the range of your zoom lens to size the enlargement right. Point&shoot cameras only have zoom lenses of course, but word on the street says you must use a zoom lens on a DSLR too, with this type of attachment.
I have not seen these adapters (I only have my Nikon ES-1), but I would normally be skeptical that a close-up lens could keep the edges of the frame sharp, however the smaller digital sensor would be a strong plus here, using only the center of the field. I already had the good macro lens, so my strong bias was for the slide-holding adapter without any optics in it, assuming better quality results from the macro lens than from inexpensive optics. There is a large difference in the cost however, and all methods have fans.
Robert Cullen in Australia shared this method of modifying a slide projector for use with a 105 mm camera macro lens. In such cases, if you reverse the slides in the slide carrier (front to back), they will come out correct (right to left) in the camera image.
You can use daylight or a flash or incandescent light source, and White Balance should match that source (not the slide subject). It is quick to just point it at the top part of the window at the sky light (**NEVER** at the sun - which would damage your eyes), and use White Balance of "Shade", or sometimes "Cloudy" (extremely convenient, but slightly variable color as sky or day changes). Shooting RAW, you decide white balance later at the computer, with many possibilities which you can view and judge by eye. You can of course correct the original slide too. RAW is fantastic, its 12 bits have more range for this. The perceived advantage of skylight or incandescent over flash is that focusing is bright. Using skylight, I use auto exposure in A mode at f/8 (f/8 seems optimum sharpness to me for this macro lens). My typical auto exposure varies from 1/10 to 1/4 second (both skylight and slides vary). Flash could be a much faster shutter, but the slide in the adapter is attached to the camera, so camera shake is no issue.
For a compact camera, flash will NOT be an option. You must turn off the internal flash (film must be lighted from the rear). There is probably no way to sync an external flash, for sure not with auto exposure. But that still leaves skylight or incandescent light for compact cameras.
For a DSLR, TTL flash works great too, I think better. My favorite method is this: The SB-800 flash unit was in TTL mode on a hot shoe extension cord (Nikon SC-28, or I actually use the older SC-17 version), a couple of feet out in front of camera, pointed back directly into the frosted front panel of the ES-1. The ES-1 is well frosted, so this works great, a really good method. Because, directly aimed at the ES-1, (you can aim it by seeing the red focus assist light in the viewfinder through the ES-1), the Nikon DSLR will focus in the dark when the red focus assist light on the flash is carefully aimed at the slide (into the lens). In TTL mode, the auto exposure is determined by the system automatically varying the flash power. The cord is short and near, so the flash is using extremely low power there (batteries last a very long time). For flash units that might overexpose that close, you could reflect it off of a white foam board background for a greater overall path distance (but then seeing to focus may be more difficult).
Focus was no problem (with the red assist LED), but one issue was that the room too dark for me to see the image well in the viewfinder. So I added a small desk lamp out in front, just to help me see. Or having a daylight window out in front helps to see it too. Either way, you want the shutter speed fast (maximum shutter sync speed), to keep out the ambient light. Flash exposure is solely about aperture, so shutter speed does not matter, so long as it does not exceed the maximum sync speed of the shutter. I used f/8 and the fastest sync speed shutter in camera Manual mode M, and flash in TTL mode. Flash White Balance, or with a hot shoe connected flash (the SC-28 cord), Auto White Balance allows the Flash Color Information Communication feature to report actual color temperture for the power level used. After you are done experimenting, and are ready to start, I suggest using the same WB for all, so all can be tweaked together, in one RAW operation. The red focus assist light aimed into the lens allowed auto focus to work great, however sometimes the focus sensor was on a blank sky area in a slide, so another focus sensor had to be selected, to be on some actual edge detail. Or, using one fixed manual power level (manual mode on the flash) for all slides almost works too, since all the slides we keep are relatively the same correctness (don't laugh).
Either way, flash or incandescent or skylight, auto or manual, it may still need minor individual exposure tweaking later, and RAW provides some range to do that, easy and well. With regard to both exposure and cropping, it seems unreasonable to expect that they will all come out of the camera just right.
About light meters: I normally use center-weighted metering, but cannot point it where I want on the slide, so I switched to Matrix metering mode for this. Light meters work by assuming any scene averages out to middle gray, which is true most of the time. So the meter reads the average intensity of the scene and tries to reproduce it as averaging middle gray. This is well and good at the original scene for the original exposure. It is also well and good for a well-exposed slide, it sees average middle gray, and it comes out averaged to middle gray. But it is a little different when seeing a black slide, or one burnt to clear film. The metering still attempts to make both come out as middle gray (auto exposure is longer for dark frames, and shorter for light frames). This is just what light meters do. So considerable tweaking may still be needed if you need to recreate the black or clear. You could choose to deal with this at time of copy exposure for the most fastidious result. I am normally as fastidious as anyone, but speaking of thousands of slides seems a little different. I don't keep many bad ones, so my choice was to ignore it and go quick and dirty, hoping RAW processing might be able to deal with it. I assumed I could always go back again if necessary, but I have not felt the need yet.
Most slides will autofocus fine, and autofocus is easily the best plan. Only a few won't focus (clear sky in that spot, which cannot be moved in the adapter), so then I just move the center focus zone over to where there are some edges to focus on. Or you could switch to Manual focus, and focus it by eye - which is easy, but it takes a few seconds. It seems important to remember to restore it to Auto focus for the next ones.
The ES-1 has sliding tubes which can move if pushed. If manual focus is used for all slides, focus must be rechecked often, because it can drift off, I think maybe due to pressure from loading the slide eventually shifting the tubes. But auto focus normally works great for almost all of them. However, this would seem an issue on the compact camera for the few slides that don't autofocus.
Once into the swing of things, I can shoot about 20 slides per minute (auto focus, auto exposure)... as fast as I can load them into the ES-1. This is the shortest part of the work. It did not include time to brush dust from the slide, which seemed unnecessary for my slides. Play and learn the first day, to see some stuff before starting the serious run. The ES-1 is not particularly helpful about controlling the position of the slide. You center it sideways by eye by judging position in the slot outside, which is easy to do, no big deal. I added a couple of layers of folded thick paper in the bottom of the slot to help center vertically. Initial alignment is not quite obvious, you crop it so the slide frame is not visible, but the viewfinder only shows about 95% of the frame anyway, etc, so it is hard to judge until you see results in the computer. Slight cropping will aid this effort, so all you worry about is getting it straight. Frankly, this location and cropping part is where the film scanner runs circles around us. But the camera's scan in a fraction of a second has much to be said for it too. Sufficient trial and error at first setup will learn about getting the right magnification size, and about centering it and rotating it straight, and then you can go very fast.
I shoot RAW, 2GB or 3GB at a time, and use Adobe Camera RAW (ACR) for post processing. ACR allows you to open all images at once (hundreds of them), and select all, and then do White Balance and saturation, and even maybe CTRL-U auto processing in one immediate operation. This is fast. In fact, any operation (like cropping) can apply to multiple selected images when appropriate. But if not done this way, it is more clicks on each one. They are all the same though, to match your flash (unless you need to correct the original White Balance too). I don't always sweat every little detail - remember, we are speaking of thousands of images. But I do look at each one (which is part of the reminiscing) and tweak the exposure processing on many, rotate those with portrait orientation, and maybe crop or straighten it. Which is just a few clicks on each, maybe 10 to 20 seconds (going fast), but on all those hundreds of images, this takes the vast majority of the time.
Even so, my best day was to copy and process 1000 slides (a hard days work - 6 hours or more). That would take at least several hard weeks of work if scanning... probably many months since there are other things to do, and scanning thousands is probably simply not going to happen. Admittedly, the scanning results might be better - this may not be my best work (due to haste, not due to equipment), but no reason to feel shame either. The digital camera method is more than fine, no issue with reproduction quality, but going so fast didn't help. My process did not include any dust processing - I did not even feel the need to brush the film. My film is stored well and the problem was mild. I can find some minor dust if I hunt, but it disappears when resampled to video screen size. Frankly, I think it may help that the slide did not sit out 30 minutes in the scanner slide tray. This digital camera method was rushed in comparison, but the images seem more than good enough to recover the memories in the old slides. No excuses, it sure beats the alternatives, of either many months of work, or no images to view. I am thrilled to have them digitized now for viewing.
RAW is much more than the ability to bypass JPG. FWIW, RAW really has no meaning from a scanner, which is already a RGB image, and we cannot use an image without gamma, and ordinary 16 bit TIF does all we could ever want. But cameras and their Bayer filters are a very different situation, plus it is the way to bypass 8 bit and JPG, but maybe the largest advantage is that the camera RAW tools are so powerful and convenient. The processed RAW files are my archive, and then Photoshop has its Batch processing with Actions to convert all the RAW files to smaller JPG copies for viewing, sized right for the situation. That may take an hour to resize a batch of 1000 images, but you can go watch TV. Frankly, I'm not sure how to work without it now. I do suspect this method is not for the faint of heart, because the close-up photography details may be slightly more difficult, and having the high quality macro lens seems a real benefit, and Photoshop ACR and Batch processing seem essential. Elements has ACR, but which does not include cropping and straightening (again, essential), nor the batch mode with Actions.
B&W negatives are easy with the camera, just invert them and you're done (at a menu like PS Image - Adjustments - Invert). Except contrast does improve B&W photos significantly, so after simple inversion, in the Levels tool, move the histogram White Point down to near where the image data ends, and move the Histogram Black Point up to near where the image data starts (very much like shown below, except B&W only has the one grayscale channel). This increases contrast (blacker blacks, whiter whites) by filling the full range with actual image data. Ansel Adams did not use digital histograms of course, but one of his first rules was that something needs to be solid black, and something needs to be pure white... which is contrast. Deep contrast is less appealing in color prints, but speaking of black and white photo prints, he said:
A print should exhibit a complete range of tones, as well as the extremes - white and black - in order to have strength and conviction. The amounts of pure white and solid black maybe be very small, but in almost every case, either one of both of these extremes should appear in order to "key" the values. Ansel Adams, The Print, page 7.
But regarding inversion, color negatives have the orange mask all over. This becomes deep blue when inverted to positive. It is a lot to deal with. Regular color printing removes it with an orange filter (Magenta and Yellow) on the enlarger light (analog). Film scanners remove it (in Color Negative Mode) by varying the exposure (time duration) of the three RGB channels (analog). The blue channel is exposed perhaps 4x longer (than red), and the green channel is exposed perhaps 2x longer (than red). This acts as if an analog glass filter were on the lens, and the longer exposure boosts the blue and green, leaving the orange complement behind. The important point is, both of these conventional methods are done with analog light, which has no limits, no clipping at digital 255. But after the scanner or camera has digitized it, then the 255 end is a hard limit. We cannot shift the data much then, the data just falls off of the 255 end, and disappears (remains stuck at 255).
With the camera, you could add similar color printing filters on the flash head, and then do simple invert. Not exact same filters, stronger, because enlarger was incandescent light (already orange), but the flash is daylight. Also different brands of film vary the orange mask a little. Or if you might have access to an old enlarger color head (like Omega) - a light source with the dichroic filters, and three CMY knobs to set the color of the light output - using this as the background light source could be a proper analog correction, and would be similar to a scanners corrections, but could still allow camera speed rate. Digital can handle slight shifts, but this major operation is best done while still analog.
This is not to say you cannot play with it in the photo editor. It might come out nearly acceptable or usable, but the range is less, and really, it's just not the same as a film scanner... and certainly not as easy.
For example, the Photoshop Curve tool even has a Color Negative preset:
You could duplicate this in any curve tool. Inversion with Curve tools: Just grab the lower left end of the curve and pull it to the top, then pull the upper right end down to the bottom. Then the RGB slopes run downward at 45 degrees, which inverts the image. Input at zero (black) becomes output at 100% (white), and ideally, input at 100% (white) becomes output at zero (black). This is the standard inversion technique. Inverts colors to their complement too, yellow becomes blue. Numerically, all RGB components become 255-previous (255-255=0, 255-0=255), which is inverted. See more about the Curve Tool.
The negative image becomes positive, but now the orange cast becomes blue. Then the channel high ends are trimmed drastically to balance the colors (reduce the inverted blue). However, there is probably no data there anyway. It is a good try, maybe dramatic, but tough job digitally in 8 bits. This loss is why film scanners instead boost the analog blue and green color first (reduces the orange color, without loss). Yes, there is this option to do this, but you may not like results. It could be good enough for some things, but you really want to use a film scanner for color negatives. Analog correction is superior this time (has full range).
Here are some samples, some options. Three methods, which are all just dumb automation. Results will of course depend on individual images. None are necessarily correct color, but just tries at a more reasonable probable situation for the color channels.
Another method is to invert the color negative, and then use a White Balance tool to click in the unexposed border around the frames (was orange, but blue after inversion) to neutralize that color. Which is about the actual color cast present, and it ought to work best, but for me, only ballpark, not real close. Clicking something white in the scene may work better. But of course, if it is already digitized, then clipping at 255 probably will be an issue.
Color negatives are a special problem, and a real film scanner will be best. But slides are quite easy with the digital camera and a good macro lens.
This was surely too chatty, sorry, but hopefully it is of help to someone. There are many details involved in all of this, and several choices, but a digital camera is in fact a good way to "scan" thousands of old slides quickly.