by Rodrigo

submit your photo

Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

This question and other internet sources propose using a DSLR with a macro lens to rapidly copy negatives and slides.

Copying color negatives, however, seems to be a challenge because the color tint of the negatives is extreme enough to make correcting the white balance in post difficult.

Thus, I am wondering if filters are available (or could be jury-rigged somehow) to reduce this tint and make post-processing more feasible, either by coloring the light source itself or being placed in the camera’s light path. They wouldn’t need to fully correct the tint, just reduce it enough to make enable complete correction in post at reasonable quality.

share|improve this question

3 Answers 3

Well, first off, its highly doubtful there is a color negative "filter" that would invert the negative in such a way that it came out color positive on the sensor. Color negative film is converted to a color positive print in a multi-step process, ultimately using the film itself as a color filter when enlarging it onto photographic paper with the correct colors of light.

If you had a camera capable of multiple exposures and the proper blending modes (like the Canon 1D X or 5D III), you might be able to simulate the process by putting the color negative film on a backlight that could emit the proper colors, and multi-expose the same frame multiple times under differing backlight. I can't offer much insight as to how well that would work. You could also try the same trick, only take three separate frames and blend them during post processing to achieve roughly the same effect (although with the added benefit that you could use RAW.) You could actually color-correct each CMY color layer that way, before blending into a digital color negative, before inverting into a color positive. You would still need to use a proper color negative inversion algorithm to turn the blended result into a color positive digital image, though.

Just out of curiosity, are you intent on using a DSLR rather than a scanner with proper negative strip scanning support?

I own a relatively cheap Canon CanoScan 8800F flatbed scanner. At the time I bought it, I think I spent about $150. I believe the current replacement for it, the CanoScan 9000F, is around the same price (and its faster to boot). It came with a dual film strip 12-slot holder for negatives and transparencies and medium format 4-slot transparency holder. This relatively cheap flatbed scanner supported scanning film at a fairly incredible 4800x9600dpi. Scanning film at that resolution was relatively slow vs. scanning a book page or other single letter-sized sheet at a much lower resolution (say 300dpi), however since it can scan 12 35mm frames at once in extremely high resolution, it was relatively fast overall (and probably much faster than photographing one color negative at a time.)

Its doubtful you would be able to achieve that kind of resolution with even a high resolution DSLR. The Canon 7D, the new Nikon D3200, and even the Nikon D800 all offer lower resolution than that on the long edge (while the D800 offers slightly more along the short edge.) You would also have noise, vignetting, distortion, diffraction and other aberrations to contend with when using a DSLR, where as that won't generally be a problem with a negative flatbed scanner. Once you scan color negatives into a computer as negatives, you simply need to use a color negative inversion tool in something like Photoshop, Gimp or similar tools to convert them into color positives. At that point, you should be able to correct for color and exposure normally with curves, levels, etc. You won't have as much correction flexibility as you might with a RAW image format, but the original image quality should be higher than if you photograph them with a lens an camera (unless you have a truly stellar lens that exhibits low vignetting and low aberration, and is capable of transferring a maximum amount of resolution to a very high resolution sensor.)

(One tip I have for scanning film is to make sure you have a nice soft brush, like a camel hair art brush, on hand for dusting off the film before and after putting it into the scanner. You'll want to dust off the scanner bed itself, dust off the film and film holder (under side, that faces down into the scanner), as well as dust off the top of the film once it is placed in the scanner. At 9600dpi, the scanner picks up every single detail, and even small bits of dust, hair, etc. can be picked up, create shadows on the film, etc. You want to get rid of as much of it as possible before scanning.)

Balanced Color Negative Inversion: Simple Form

With Adobe Photoshop, or a similar tool with the same functionality, inverting a color negative to a color positive that can be properly color corrected can be a fairly strait forward process. The key tools needed for a proper color negative inversion with this method is a precise levels tool that can be used to set white and black points, an invert command, and color balance adjustment. Lets start with a simple negative (I have none of my own, so I found this one with a Bing search):

enter image description here

If we simply invert this negative without proper preparation, the color positive inversion will usually have a fairly strong cyan cast to it:

enter image description here

The significant orange color cast is due to the classic orange film substrate used in color negative film. Normally, it is part of the color filtration process that allows properly balanced colored light to be projected through the negative in an enlarger onto photographic paper to produce a proper photographic color positive print. All of that mechanical process is meticulously designed and works in a very specific way. A color negative scanned into a computer is unlikely to exactly match that process, so some additional correction needs to be done before inversion to compensate, and some post-inversion color balance adjustments are often needed to fully color correct.

The first step is to adjust the white and black points of the original color negative image, before inversion. The simplest approach can be done by adding a Levels adjustment layer in Photoshop, and simply clicking the "Auto" button:

enter image description here

This will pre-tune the image for proper inversion within photoshop itself. Once levels have been adjusted, you can now invert. Just add an Invert adjustment layer above the Levels adjustment layer. Depending on the film used or the state of the negatives, there may still be some color cast in the resulting image. In this case, there is still a slight cyan cast, however the results are far better than our original attempt at inversion:

enter image description here

The still-present slight cool color cast could be the result of a variety of lighting or development conditions, and some of it may be due to the simplistic nature of this approach to color negative to color positive inversion. Now that the extreme cyan cast of the unprepared direct inversion is gone, color correction can now be done with the standard color balance tools. There is quite a bit of cyan/blue in this photo, so I've added a Color Balance adjustment layer with the settings C/R +100, M/G -10, Y/B -20:

enter image description here

Almost perfect, if a touch on the cooler side. If you want the results to be even warmer, you can add another Color Balance adjustment and tune more to the reds/yellows. This final sample has an additional C/R +50 adjustment:

enter image description here

Looks quite perfect now. Further tuning could be done to taste (I've added a bit of magenta tint to eliminate what looks like a very slight green cast on my screen, however that may simply be a consequence of my calibration, and it may look different on other peoples systems.)

This approach is very simplistic, basically an Auto Levels and Inversion to get you into an easily color-correctable state. More advanced and precise techniques can be achieved via the Curves tool and RGB curve adjustments alone, resulting in perfect inversion with a single adjustment layer. It can be more time consuming, but the results can be tweaked and perfected quite a bit more than with the simplistic approach.

Balanced Color Negative Inversion: Curves Form

To invert a negative with curves, simply add a Curves adjustment layer over your negative. You will want to use RGB curve editing. There are four base curves in RGB editing, your R, G, and B channels as well as a "black" curve, which can be used to adjust contrast (well call it the contrast curve from now on). You use the curves to both invert as well as tune the color of a negative base. To invert, for each color channel (except the contrast curve), drag the corner black point from the bottom to the top along its edge, and the white point from the top to the bottom along its edge. This will invert the negative much like the direct untuned inversion from the simpler approach:

enter image description here

Once inverted, you'll need to tweak the results. Generally speaking, you will want to adjust the white and/or black point of each channel by dragging them horizontally towards the point where that channel's levels fall off to zero. You will frequently only need to change either the white or black point for any given channel, its rarer that you will need to adjust both white and black points. This will begin to eliminate the extreme cyan color cast. You may want to fine-tune the points for each channel by either placing it before the first blips of tones begin, or even after tones begin, or any other extreme...to artistic intent. An initial white point adjustment will cause your curves to look something like this:

enter image description here

Once white and black points have been set, you will need to tweak the bow of each RGB curve to balance color and style the photo as you see fit. This can be a time-consuming process, and depending on the range of tones for each color channel of the original photo, the results may differ. Its useful to give each curve a toe, where the tones peter off near the black point, as that helps with contrast and can help eliminate potentially unsightly artifacts (especially if you set your black point within the tonal range for a given channel.)

enter image description here

The above results look great, and from a color balance standpoint, seem to look better than the final results of the simpler method (which still has a bit of a green cast to it.) You still have the option of tuning contrast independently of each color channel at this point. A little tweaking of that black "contrast" curve can further tune an inverted negative photo:

enter image description here

share|improve this answer
I'm not expecting the filter to do the inversion; that would be in post (and thus the same problem would be present with the scanner). It's that the strong orange tint of the negative means other colors are severely underrepresented, making color correction after inversion difficult. Regarding speed, the article I read was talking up to 10 slides per minute for the raw capture. –  Reid Jun 3 '12 at 20:46
@Reid: What tool are you using to invert? I've done some color negative processing in Photoshop before, and I've never noticed any issues resulting from the orange substrate's color cast in the negative. A proper inversion algorithm will account for all of that, and all color channels should be correctly represented in a digital color positive after inversion. –  jrista Jun 3 '12 at 20:52
Lightroom 3. I haven't actually tried any of this yet; I want to make sure it's feasible before I buy a macro lens, etc. I think the problem people are having is not that it can't be done, but that with 2+ fewer stops of information in some channels, the quality is poor. –  Reid Jun 3 '12 at 21:06
After reading this answer I'm completely convinced on a project of scanning the family negative archive. I have a couple of suggestions to improve workflow: 1) Include a bit of the unexposed border of the negative, so I can see the orange part. I guess that after inverting and correcting the color, this part has to become pitch black. Isn't it? 2) Use only adjust layers, and spend a lot of time on the first image of each set. Then, for each new pic. I'll just put it under the same adjust layers. Won't parameters be the same for all exposures in one film roll? –  Jahaziel Jul 26 '12 at 17:38
@Jahaziel: Sorry, didn't see your comment till now. I would suspect parameters would be similar for a single roll of film. I guess it depends on how old it is, and the condition of each frame. –  jrista Jul 18 '13 at 19:37

I have been using this method for scanning my negatives into my computer via a DSLR. Scanning Negatives

Originally, I was doing this to scan black and white and slide film. Once I read some details about color correction, I was better able to use filters that I had to correct for the orange film of a color negative. What's been identified in comments above is that the blue channel will be weak on the order of 2 stops. By using a blue filter on the light source, the overall light balance will be better.

I have been using an 80A blue filter either as a screw on lens filter or as a gel on the surface of the light. This is a good fix for nearly all the color negatives and usually allows a custom white balance to be good enough. If I have a significantly dark negative (Fujifilm NPZ is pretty dark orange) then I'll have to get more blue and even some green into the light source. Adding some gels to the light source is pretty easy and essentially mimics what the copy stand device is doing in the Camera Scanning link.

In further work, I've switched from using an LED based light source to a box with a strobe in it. This allows me to use a consistent light that is much easier to alter for light spectra. It also does a better job of penetrating the dense color negatives without having to resort to long exposures.

share|improve this answer

Saw this old thread but could not resist sharing my experience for at least one person to maybe benefit from it.

I'm in the middle of converting thousands of color negs mostly Kodak Gold 100, 200, and 160NC.

I use a Nikon D810 Camera, 105mm f2.8D Micro Nikkor, Two D700 flashes in a plywood light mixing box with two 1/8" diffusers, a neg carrier that is 2" above the top diffuser, two blue (Full CTB) and two green filters to swallow the orange film cast. Exposure is 250th sec ISO 64, F11, flashes set to ~1/4 power each (depends on the neg density and your light box design). I use a remote shutter for the camera too for convenience but with using flashes the exposure is so fast ~1/4000 of a second that vibration does not matter. I set the lens to the right mag factor just under 1 to 1 so not to crop the neg any then move the camera in and out for focus. I do not focus using the lens focus ring as it changes the magnification. I focus using live view and a compact florescent bulb in the light box that I shut off for each exposure. I set the camera to all neutral everything on daylight WB so I can see if a channel is clipping (semi UniWB idea). I try to get the histogram in the middle which is easy to do actually since the negs are very low contrast. BE CAREFUL not to overexpose since it swallows your shadows in the negs (the areas of the negs that are near clear of dye or areas that look close to the film base density). Better to underexpose a bit than overexpose. You need good density in your digital capture of the neg I found or shadows clip and get crossed colors easy.

Then I put in lightroom and set to camera neutral and output to phohtoshop and use filter plugin program ColorNeg as in the link below.


It is a filter that you setup in photoshop. It costs about $75 USD but works very nice. Otherwise use the curves and invert but it takes more time. With the curves in photoshop just tighten up each RGB curve first for black and white points at the farthest dot of data then go back and use the film base color gray point picker each color individually. Add an invert layer. Go back to lightroom and adjust color/denisty/etc and filter out grain to taste.

Works pretty darn good and I have some negs of objects with many neutral colors and skin tones and they come out perfect. It is amazing the dynamic range a color neg is able to capture!!!

If you use the ColorNeg filter plugin you can skip using the blue and green filters on the light and it still comes out pretty good. If you use the color filters and use the curves and do not use the ColorNeg filter you have the potential to get the best natural color with the least amount of weird crossed colors especially in your highlights and shadows. The colorNeg program is fast though if you have many negs to do over many different types of film.

Happy Shooting :->



share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.