Introduction: The First Color Photograph
In the "Olden Days" professional science was still in its infancy. People who trained in science and practiced science were using every ounce of creativity and imagination at their disposal to discover the nature of the world. They were scrappy and inventive, not unlike the modern day hobbyists and makers you see here on Instructables.
For this "remix" instructable I will outline a modern replication of the experiment that produced the first color photograph ever made. As you will see, the method for making the original color photograph was itself a remix! It was made in 1861 by Scottish physicist, James Clerk Maxwell working with the photographer Thomas Sutton (see a nice historical blog post here). The image was of a tartan ribbon and based on ideas he had worked out and described in a scientific paper six years earlier, in 1855 (you can get Maxwell's original paper here).
In a world driven by multi-media and digital imagery, it is easy to forget that not so long ago, photography was a brand new technology. As of right now, it is less than 200 years old! The first photograph known to be taken that is still surviving today was by Joseph Nicéphore Niépce in 1826 or 1827. It is a photograph taken through an upstairs window, looking out over the rooftops of his estate in Le Grase, Burgundy, France. The earliest photograph known to include people was by Louis Daguerre in 1838, showing the Boulevard du Temple in Paris.
These early photographs were all monochromatic -- "black and white." They were largely based on chemical reactions of silver nitrate, which darkens when exposed to sunlight. Such reactions are called "photochemical." While the novelty of capturing life exactly as it appeared at a given moment was enchanting, there was always something missing: color.
This Instructable will guide you through a modern implementation of the Maxwell-Sutton method. It is not necessary by any means, but it is enjoyable process and has its own charming elements that can be launching points into your own personal explorations of the artistry and science of photography.
Step 1: Maxwell Color Separation Method
Maxwell's method is to exploit monochromatic photography in a subtle and clever way. The basic idea is this:
- Take a black and white photograph of a scene through a colored filter. The image will be black and white, but the only things in the image are the things that can be seen through the filter.
- Take a multitude of black and white images through different filters. The classic choices are RED, GREEN, and BLUE.
- To reconstruct the color image, shine a RED light through the "red" black and white image; shine a GREEN light through the "green" image, and a BLUE light through the "blue" image.
- Align the three projected images together on the screen or wall, and the correct color image will appear!
In images where the colors are well defined and separated, it is easy to imagine how this might work. Above, I show a Hubble Space Telescope image of the Helix Nebula (NGC 7293). Hubble takes images using exactly the Maxwell method, through separate color filters, with the final image being a combination of the separate images. The three black and white images are the RED, GREEN, and BLUE breakdown of the Helix Nebula. You can compare them to the color image to see how different parts of the full color image are created.
Step 2: Materials
To complete this Instructable, you'll need:
- some cardstock
- tape
- Red, Green, Blue transparent plastic (see below)
- Smartphone (or other camera)
- Tripod
- Printable transparency sheets
- A bright light
The keys to the method are the colored filters. You can try just about anything, but the overall color balance you end up with depends on the color of the filters and how dark one filter is compared to another. For accurate color representation (as opposed to artistic -- see the last step!) you want the three filters to be roughly the same in terms of how dark the world looks through them, and slightly darker is better than slightly lighter.
Good sources of filter material might be clear gel filters used in theatre lighting, or colored PVC sheets at your local art supply store. I've used clear cellophane for gift wrapping for this Instructable.
Step 3: Color Filter Mount
To make photographing simpler, we'll build a "filter slide" that holds a colored bit of filter in front of the lens on your camera. For this step I'll assume you are using a smartphone.
Begin by cutting a 1-1/2" wide strip of cardstock from the long dimension of your sheet.
Fold the strip around your smart phone so it overlaps itself making a loop sleeve.
Note where the camera is; unfold the strip and using an excacto knife cut an opening over the location of the camera.
Tape the sleeve shut into a loop, and confirm your hole sits over the camera.
Step 4: Color Filter Slide
To make the filter slide, take another 1-1/2" wide strip of cardstock about 8 inches long, and fold it in half. Cut three large windows through both sides using your exacto knife.
In each window, pick one of your three color filters and tape a bit into the window. Be careful to not get any tape in the window, otherwise it will show up in your photographs!
If you need the filter material to be a bit darker, you can fold it over on itself (I had to do this with my blue cellophane). If you do this, make sure it is as flat as possible, or your photographs will not come out crisp and in focus.
When the filter windows are covered, tape the edges of the cardstock together so you can slide it into the filter mount as one piece without it flopping open.
Be careful to place the windows so you can cover the camera with the film, not the cardstock. I had to trim my first filter slide a bit because the cardstock boundary was a bit too thick the first time I cut it, and it blocked about half of the camera view!
Step 5: Using the Filter Slide
Insert the slide with the filters underneath the filter mount against the body of your smartphone. Carefully slide the filter back and forth until one of the colors shows through the window.
If you look at the camera view, you will see a color filtered image of the world!
When you are ready for the next filter, push the slide along underneath the mount until you reach the next filter window.
Step 6: Photography Setup
The key to successful Maxwell-Sutton image composition is insuring that each of the images is taken from exactly the same vantage point and orientation. For the method to work, you are taking three identical photographs through each of the different filters, so you don't want the camera to move between shots. Since I'm doing this with my smartphone, I use a small table-top "Gorillapod" tripod from Joby.
Set-up your camera with your still life well lit up -- the camera flash will be behind your colored filters and will not illuminate the scene; I recommend turning the flash OFF so it doesn't fire and glare off the inside of the filter.
Slide your filter into the sleeve to the first color (red for me) and take the first shot -- make sure you tap to focus! Carefully slide to the next color (green for me) and take a photo, then slide to the last color (blue for me) and take a photo.
You may experiment with different exposure settings in your camera, to see how the resulting filtered images look.
Step 7: Printing the Modern Maxwell Ribbon
One difference between what we are doing and what Maxwell did is we are taking COLOR photographs through filters, then recombining them directly. Maxwell took BLACK AND WHITE photos through filters, then shined the pictures through the correct color filters to recombine them.
Each of your filtered images will look somewhat like the still-life, but not exactly. In the first image above I show a colored ribbon I used (I did not have a Maxwell family tartan ribbon!), and the three printed images taken through each of my filters.
On a piece of transparency, print out each of the red, green, and blue images separately.
As noted above, it is not perfect as the filters themselves are not perfectly balanced against one another; experimenting with different filters can make the color return more true to the original. As we will see below, all of the color information is in the filtered images, and software processing can make up for color imbalances.
Step 8: Viewing the Modern Maxwell Ribbon
Stack the three images on top of each other, and carefully align them. You'll notice in my ribbon, I had a small paper clip holding the ribbon in its shape, which was very useful for alignments!
Hold the stack up to a bright light and you will see the recombined color image!
Using random color filter material, it is hard to get a perfectly accurate color image out. It is complex combination of the color filters you use, your printer's interpretation of those colors, and the quality of the print. But you can clearly see the color pattern and the trends in the colors. In this particular case, I think the blue needed to be even darker.
The quality of your lightsource also matters. A light-table would be ideal, as would sunlight. The day I was doing this project was cloudy, but if I get a sunny day someday soon, I'll add a picture viewing with bright sunlight.
Step 9: Using Software for Recombination
Modern printing methods use a variant of the Maxwell technique to generate color on the printed page or screen, breaking images up into color channels (like RGB or CMYK) that when mixed give the full color images you are used to seeing; this is why your color printer has different color ink cartridges. In the first image above, I've put a 20x magnifying loupe up against a white space on my monitor, and you can see the individual pixels are lit up in combination of Red, Blue, and Green!
In software like Photoshop, the color separation information is stored in "channels." We can have the software combine our color filter images into a full color image using the channels.
Which color channels you have in Photoshop depends on the image Mode you are using. In my version of Photoshop (CS6) you can find this under the menu Image > Mode, as shown in the second image above (we're using RGB for this Instructable).
Selecting under the menu Windows > Channels will bring up the separated channels. In the third image above you can see I've opened a full color RGB image. The channels shows the combined RGB result, and then separate R layer, G layer, and B layer. They act just like layers do in any document -- you can turn individual ones on and off, or change them. This is what we will exploit in the following steps.
If you don't have access to Photoshop, there are also good openSource image manipulation software packages, such as GIMP (Gnu Image Manipulation Program) for all platforms.
Step 10: Software and the Modern Maxwell Ribbon
Open up all three of your filtered images in Photoshop, and note down the dimensions. Open up a new document with the same dimensions as each of the filtered images.
Go to the red image. Select the entire image ("Select All"), and copy it. Switch to the new image file, and view the channels described in the last step.
Now we're going to copy the individual color filtered images into the correct color channel:
- Select the RED channel, and paste. You'll see a greyscale version of the red image in the red channel layer.
- Repeat the process by copying the blue image file, and pasting it into the blue channel of the new image.
- Lastly, copy the green image file and paste it into the green channel of the new image.
You can see the result in the last image above -- nothing has been done to rebalance this, but it looks very good! There is an excess of green, which you can see in the white background -- this is because the green of my clear cellophane is not what Photoshop thinks the basic green filter should be. Never-the-less, the color response you see is excellent -- your screen display, on a pixel by pixel basis, is much better and showing good combined color response than overlaying the fine grains of ink or laser toner on printed sheets!
Step 11: Software Color Balancing
As we noted above, software is good at rebalancing color, and can make up for the relative strength in the color of our individual filters. Photoshop has an auto-balancing feature that does its best to rebalance how much each of the color channels influences the overall image combination. Above, I've shown a rebalanced color combination, and the original full color ribbon for comparison. With the balancing, you can see the green background is closer to the truer white it is in real life.
Step 12: Last Thoughts & Artistic Games
I'm experimenting with this technique, but would like to settle on a good set of filters and good printing process so I can print a piece to hang in my house as art. The need for illumination is making me think about a shadow box with LED strips and a light diffuser behind the printed transparency stack. It will make for good conversation!
Beyond that, there is a well developed science around color which you can learn about and exploit, but the ability to play around creatively can lead to unexpected and interesting artistic results. For instance, consider the color balanced collection of Rubik's style puzzles in the first image above, created using the process outlined in this Instructable. In the second and third images I took the red, green, and blue images and pasted them in different channels -- this makes strikingly different, and clearly not correctly colored images! In the last image, I took the red and blue channel and left them alone. For the green channel, I did a "color inversion," replacing the green with it's color "opposite" (a kind of pink palette; Photoshop can do this automatically). The result is a very alien glowy green image! Awesome!
I hope you enjoyed this remix of a photographic history, and experiment some on your own with creating your own images of the world!

Runner Up in the
Remix Contest
15 Comments
3 years ago on Step 12
Very nice! Reading about the color separation method you used, I thought I would mention something that I did - for the benefit of anyone who would like to play with separate colors easily.
I have a DSLR camera that has an infrared shutter trip feature. So, I took an Arduino Duemilanove, added a 24 LED "NeoPixel" ring, an infrared emitter LED, poured in a few lines of code, stirred and got a simple, inexpensive and fun add-on to my camera.
The NeoPixel ring slides over the lens and makes nice, shadowless illumination. Not terribly bright, but good enough for experimenting. And, don't laugh but I use a strip of tape to hold the ring on!
What the software does is (as rapidly as possible) turn on full bright red, that fire the shutter, next full green and shutter, then lastly blue. The camera itself is set to take "monochrome" pictures. I can play with color balance by varying the brightness of each color that I use.
Since the light/shutter/light/shutter... sequence is so fast, I can even get decent results photographing a person.
Note that using the NeoPixel ring gives me limitless ways to experiment with color. I can do ordinary RGB, I can emulate the old "two color" Technicolor method, I can make up my own two color combinations... almost anything is possible.
Lastly, I can take the three images and combine them into a full color image. If I want to make three separate prints (like laser printed transparencies that I then sandwich together), I of course need to use subtractive colors (magenta, cyan and yellow). I don't bother with the K (black) channel.
Have fun, and thanks for a really interesting Instructable!
Reply 3 years ago
This is great. When I started this I had contemplated taking the pictures through the filters with my camera using the "monochrome" mode, but I hadn't found a satisfactory way to print the monochrome images out with just one color (red instead of black). If I could do that, it would be just a bit more like Maxwell did it. :-) Glad you enjoyed the instructable!
Reply 3 years ago
To convert your three monochrome images into R, G and B, simply make a color layer and paste it on top of the monochrome image as a "multiply".
For example, if you have a 1024 x 768 monochrome image that represents the red channel, simply make another 1024 x 768 "image" and fill it with #FF0000 (full red). Then copy it and paste into the monochrome image using "multiply".
Do this for all 3 monochrome channel images, then finally paste each layer together using "addition".
If this doesn't make sense, let me know and I'll try to explain it better.
3 years ago
Ah ha, now I get it! I’d heard somewhere that images from Hubble and other space pix were transmitted back to Earth in black-and-white, then ‘recolored’ when processed, which made me wonder who got to stare through a ‘scope to figure out what colors to use. Using filters makes way more sense, and will still give you the true and accurate colors of what’s actually out there. Thanks for solving the mystery for me, and provoking my curiosity for my own photos.
3 years ago
I have seen the Maxwell Ribbon with my own eyes, and marvelled.
But I've no idea how he did it without access to colour-sensitive emulsions.
But even if it's a Fascinating Fake, Maxwell's great insight - that three colour separations can almost wholly satisfy our three-colour eyes - is still a massive achievement.
David (ex Kodak)
Reply 3 years ago
That's awesome you've seen the original! Something historically I find interesting and don't know about is why it took several years between when Maxwell had the three color deduction and was able to try and do it experimentally?
3 years ago
Hi there
Wonderful Instructable and history lesson...
Before digital photography, the printing industry used this method to produce color images...
Since todays photorgraphers don't see the need for filters (but that's another story) you will find proper color filters on Ebay for cheap.
#25 - red, #47 - Blue, #58 (also X1 or a #67) for green will work wonderfully for shooting & printing...
It's amazing what you can do with them... Thanks again...
Reply 3 years ago
Thanks! That is a great hint for finding some filters to work with! I will definitely see what I can find. I'm glad you enjoyed the instructable!
3 years ago
Thanks everyone who commented on the IR part of this story -- you are correct! In fact, modern solid-state sensors, like the one in your phone and DSLR, are also still IR sensitive. You don't get a lot of IR response because there is a filter that blocks the IR and only lets the visible light through. You can pull that filter out, or more to the point, replace it with one that only passes IR if you like (there is a an entire IR hobby community that does this). I've done this with some students with an off the shelf camera from a department store, and we flew two cameras side by side, one IR modified and one not, on a high altitude balloon. I include the images here side by side for you to see the differences. :-) Thanks for looking at the instructable, everyone!
3 years ago
Excellent! Well done technique, and an A+ on your historical synopsis.
Reply 3 years ago
Thanks, glad you liked it!
3 years ago
Yes, it is true that early black and white photographic emulsions were not red sensitive ! That's why most early photo's were somewhat fuzzy, not mention chromatic aberrations of the early lenses. But in some lens, the aberrations helped, due to the infrared light focus was further away from the blue and green objects ! But as newer lens technology increased and new pan-chromatic film emulsions were developed, everything is tack sharp !! PS: If you want to see how far the IR shift is, just look on most high quality 35mm lenses.... You will see a red line on the barrel. When you get the object in manual focus for normal viewing, you then mark that spot and move it to the red line. Then your object will be in IR focus !
3 years ago
While it was mentioned that the original experiment used black and white film, it was left un said that the film of the day was not sensitive to RED light, so the original experiment would have failed to produce a RED transparency. We all know that a red filter is used in the darkroom when processing B&W film. So how did it work? Was it a fraud? It turns out that the film WAS sensitive to infared radiation and the RED areas of the picture ALSO radiated in the infared range so it worked. Had the experimenter been carefull in his research he would have realised that the B&W film wasnt sensitive to RED and he would not have tried the experiment and it might have been many more years before we had color photography....LOL
3 years ago
This is a great image processing tutorial! This could be a fun science project too!
Reply 3 years ago
Thanks! I'm glad you liked it. I'm a physics professor, and I always thought this would be a good project for my students to learn about astronomical imaging. Now I have it written up! :-)