Andrea B. Posted September 21, 2019 Share Posted September 21, 2019 Note: Please bear with me on this topic. I'm not going to be able to get everything posted all in one sitting because we have lots of things going on here at UVP Headquarters - West. MaxMax makes 3 Infrared bandpass filters: https://maxmax.com/filters/bandpass-ir These are beautifully made filters. Each has a shiny coloured side (dichroic?) which gives them their designation as blue, green and red. The color designation has nothing to do with how the raw or false colours appear in a finished IR photo - at least as far as I can determine currently. On the linked page Dan has shown some RGB stacks made with the 3 filters. My initial experiments were not quite that colourful, but that's OK. I'll get there eventually. For my first experiments I had simply wanted to explore the basics. I always like to look at the demosaiced raw photo before any RGB multipliers are applied to set the white balance. This raw composite gives an idea of how the camera is really recording through its Bayer filter. The excellent app Raw Digger outputs a very basic raw composite with only minimal contrast/saturation curves. Gear: Nikon D600 + some 35mm lens + Sunlight Measurements of the raw colour casts were made over the white Spectralon. Then a square of the fully saturated colour was added to the raw composite. IR BP Bluemax transmission about 93%half-max width about 90 nmf/3.5 for 1/500" @ ISO-400Like other IR-pass filters which pass some high red, we get lots of false colour with the BP Blue. It has a strong Orange raw colour cast. IR BP Greenmax transmission about 90%half-max width about 65 nmThe colour cast for the BP Green is very unsaturated. When the saturation is pushed, we can see that the colour cast is Red with about a 25% contribution from Blue (so headed towards Cerise, I suppose).f/3.5 for 1/500" @ ISO-400 IR BP Red max transmission about 85%half-max width about 37 nmYou can see why the BP Red requires a longer exposure time. It's lots less wide and peaks out there in the 900s.The colour cast for the BP Red is definitely Magenta after full saturation is reached.f/3.5 for 1/40" @ ISO-400 Link to comment
Andrea B. Posted September 21, 2019 Author Share Posted September 21, 2019 Before stacking up the BP R, G and B IR-bandpass filters, I want to look at the tonalities in each layer. For each photo here, the raw composite from Raw Digger was converted to Black & White in Capture NX2 and then given a black point & white point adjustment to make it look more like a proper photograph. (You can do the endpoint adjustment manually with a curve tool or by applying auto-levels on the luminance layer.) Usually we would make some more tweaks to the file than just that (raise the midtones a bit? and definitely some detail enhancement). But that's not needed for these test photos. The important thing to note in these examples is that the tones are almost the same in the BPG and BPR photos. This is why some subjects are not going to produce a very colourful stack without some help. IR Bandpass Blue IR Bandpass Green IR Bandpass Red PS Difference between BPG and BPR.The little breezes show that these were two different files. But look at the stationary areas - no not many differences! Now think what happens when you put the BPG file into, say, the blue channel and the BPR file into, say the red channel. You are going to get magenta almost everywhere. So when that's combined with the BPB file placed into the green channel, you don't get very interesting RGB stacked false colours. But of course the subject here is almost all foliage, so what could we expect from a tri-colour infrared RGB stack? Other subjects do produce a bit more stacked colour with help. (See the MaxMax link above for a reminder that this can work.) Link to comment
Andrea B. Posted September 21, 2019 Author Share Posted September 21, 2019 Tri-colour IR Stack First, choose a stationary subject........ :lol: If the breezes had been just a little stronger, we'd have had a nice Chrono-colour stack. So, just look at the stationary areas please.This stack was made by converting the raw composites to B&W and then placing them in channels.Raw Composite RGB Stack Added IMPORTANT NOTE: I was using the raw composite black & whites for that stack. I need to add a stack made with the actual photos.Added: ...and I did, just after this next one. Out of curiosity, I difference-stacked the raw composites without placing them into channels. Remember that if you do this that the order of stacking makes a difference. This is a stack from top to bottom as BPRed, BPBlue and BPGreen.If the BPBlue is moved to the top, then its raw orange colour would dominate even more than it does here. It doesn't make a difference, in this particular case, whether BPRed or BPGreen is on top because they are almost the same very unsaturated and close in raw colour.Raw Composite Diff Stack, No Channels Original File RGB StackThis stack was made by converting the original files to B&W and then placing them in channels. It looks slightly different from the raw composite stack above (first stack) because the converter is applying the usual D600 profiles & curves to the files. (That's my theory anyway. It might change. "-) ) Link to comment
Andrea B. Posted September 21, 2019 Author Share Posted September 21, 2019 I must add a follow-up note here: my channel stacks may not be perfect because I was using PS Elements and not a real channel tool. They are, however, close enough. So that's all I have for now for tri-colour IR stacks.I'll check back later and see if there are any questions, corrections or suggestions. Link to comment
Andrea B. Posted September 21, 2019 Author Share Posted September 21, 2019 Big question. For the BPB -> blue, BPG -> green, BPR -> red assignment, MaxMax got cyan coloured grass.I did not. Now I'm confused. But I'll have to figure this out later. Link to comment
bvf Posted September 21, 2019 Share Posted September 21, 2019 Andrea, Interested to see your IR Tri-colour work - that might be my next project after I've done more on UV Tri-colour. As you say, trees and vegetation are not an ideal subject for tri-colours because they tend to move! But there is another reason why vegetation may not be a good subject for IR tri-colour: the IR reflection spectrum of typical leaves is pretty level from 750nm or so to about 1300nm, so you're not going to see a lot of colour in your results. You might get more interesting coour by having a go with non-plant subjects. Link to comment
dabateman Posted September 22, 2019 Share Posted September 22, 2019 I think its important to remember the reflectance of IR light for different stuff. I am not sure you are going to see much difference.I like this image from this website:http://www.wrotniak.net/photo/infrared/index.html Link to comment
Andy Perrin Posted September 22, 2019 Share Posted September 22, 2019 You definitely have to get into SWIR to see interesting variations for the most part. (Exception is that water peak I’ve been exploring.) Link to comment
nfoto Posted September 22, 2019 Share Posted September 22, 2019 The swapping of filters on the same camera/lens introduces a temporal aspect to the ensuing photography. Sometimes, this can deliver more interesting results than the combination of tri-colour IR bands themselves. Nikon D600 full-spectrum, 200/4 + X-Nite filters Link to comment
Cadmium Posted September 22, 2019 Share Posted September 22, 2019 Birna, Quite a beautiful shot. Link to comment
Tom Lewis Posted February 29 Share Posted February 29 Yes, that is an amazing photo, Birna. @nfoto Link to comment
Tom Lewis Posted February 29 Share Posted February 29 Thank you for doing all this, @Andrea B.. Despite the fact this was 4 years ago, it’s fascinating. Link to comment
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now