in some of the images- there is something really pure and special about the color- something I think I've seen in old 1940s and 30s three color type work... the process seems to love the off teal and faded red colors...
That's cool a cool experiment. But it only works for trichromacy and below with a 3-fold filter combination. I have (moderately functioning) tetrachromacy that's induced artificially via special glasses. For me there's literally an entire dimension of color missing in just an RGB image and it messes up colors a lot. For example, when I film a red-green (not yellow!) with my phone's camera it appears like a red-yellow on my screen, which is a different color. Also, I have yet to see a screen that can display a pure yellow without any red or green pollution. I'd need a camera (and screen) with RYGB, where the sensitivities to RYG are narrower than in trichromatic vision, to fully represent the colors that I can see. And to recreate your experiment, I'd need a red filter that doesn't let yellow light pass through, a yellowish filter that doesn't let red and green light pass through, a green filter that doesn't let any yellow light pass through and a standard blue filter. Inversely similar to how a strong dichromat would just need two filters to represent most to all of the colors they can see.
I haven't tried it, but I suspect it would create a pretty realistic depiction of the original colors, assuming any movement of the rainbow was minimal between shots.
Dude I absolutely loved this video, the editing is just so good and it's very informative as well. Well done and thank you for making this! Will definitely try this out.
Getting the "right" (primary colors) filters isn't easy. There are three specific Wratten filter numbers for tri-color imaging: Red 25, Green 58 and Blue 47B. Hard to get, even the Tiffen screw-on filters (they produced them until lately, or maybe they still do)! I did a similar experiment with some cheap plastic filters, but their colors were not quite right; and now, that I need some larger filters, these cheap filters look different in the size I need. So I searched and already have 25 and 58, but have trouble to get 47B …
The production went mega hard on this, I loved the creativity in the presentation! The editing was so snappy and your style is so upbeat and made this so entertaining! The analysis of the photos in the video were fun, but many of the examples you showed in succession at the end wound up outstanding with this method! Also, it gives a really nice quality using a sharp, but relatively coarse grained film like tri-x for this experiment! Really magnificent!!
I don't remember the specific filter values. I used the darker set of gels, which came from a Roscolux Cinegel swatchbook. I just picked the most dense gels which looked to be closest to the primary colors, as those would give me the highest degree of color filtration. I wasn't too concerned about the exact hue since I would be correcting slightly after scanning to achieve the most pleasing result.
If you wanted to achieve a similar effect using a digital camera, there are a couples ways you could go about it. Each color image you take with your digital camera can easily be divided into the three black and white images (known as color channels)in a program such as Photoshop. If you wanted to create the multiple exposure effect across channels, you would simply take three photos, then take the red channel from Photo 1, the blue channel from Photo 2, and the green channel from Photo 3, combing them to produce a color image, but with each channel coming from a slightly different moment in time. Shooting through filters isn't' really necessary with digital, because the camera is already filtering and storing the color information with each shot, whereas with black and white film, the filtration is the only way to capture that same color information. If your camera has a true black and white mode, in which it only captures luminance values, you could produce a similar effect using the method used in the video.
It wouldn't produce color by holding the negatives together and looking through, as the black and white negatives aren't actually producing the color. If you had three slide projectors, one with each negative and projected them in the same spot on the wall so they line up, then added a red, blue, and green gel in front of each corresponding projector, they would combine to form a full-color image on the wall.
I was expecting that you would reversal process the film to give positive images and project them on to a screen using three separate projectors with the appropriate colour filters.
The darker set of gels were from a swatch book of Rosco Cinegels. I don't remember the specific colors used, but I know they make primary gels that would work well for this purpose.
I load all three BW images on their own layer in Photoshop, then create a new blank layer that is solid white. I then copy the Red BW image into the Red channel of the solid layer, the Blue into the blue channel, and green into green. You then have a full-color layer, with each of the three photos you took as the corresponding channels for that layer.
It seems that red, green and blue makes up most, if not all the colors we can see. But would you mind doing this experiment again? This time using a green and magenta filter to see if 2 colors can show the visible color spectrum? I'm interested in this because of various 2 color movie processes from the 1910's - 1930's being able to show a lot of useful, but not quite all of the colors we can see. And the reason for green and magenta is that magenta is in between red and blue. The green is used as a filler that helps magenta reach blue. And magenta is hopefully close enough to red to show it how it's supposed to be. Do you want to give this a try?
Jesus!! Just watched your three colour method as well, and you are head and shoulders above all the tripe . And there is your problem. I sincerely hope you have a tonne more subs in real life because you bloody deserve it.
It absolutely is a treasure chest. I've spent my life as a still photographer. So I finally took a couple of cinematography classes at a local college. This is stuff they don't teach. It is absolutely wonderful. And for me inspirational.
This video just came up on my feed. I wish all videos on RU-vid were of this quality. Exceptionally well done and very creative. Information and entertainment values were tremendous.
I don't have any plans to do a tutorial of the Photoshop portion of the video, but I'm happy to answer any questions you may have. Essentially the workflow is as such: 1. Copy each R, G, B photograph into the corresponding color channel within a given layer. 2. Fix any alignment issues that may have occurred while shooting or scanning. 3. Use a Curves or Levels adjustment to tweak the contrast and color cast of the photo until it starts looking good!
Thanks for watching! For the best results, I've found that you'll want to adjust your exposure settings to compensate for any differences in density levels between filters. So for example, with the red filter in place, set your camera to properly expose the image. Then for your blue image, if for some reason it is now showing that it's underexposed by 1 stop, adjust your settings to match the same exposure you had while with the red filter. Repeat with green. That way they'll all be within a pretty close margin to allow you some flexibility for when you combine them in post.
