Infrared with the M system

[Larry Rexley]

With my full spectrum cameras I've found a fair number of old photographic filters can work well. Practically any of the red/orange yellow filters for black & white film will be long pass filters equivalent to high color/super color etc.

Blue & green filters are much more variable not all pass IR & this can vary between filters supposedly of the same type. Some that have worked well on one occasion have been quite useless a few months later (possibly from varying IR in the sunlight)

A variable ND can be very good allowing the visible portion to be adjusted while constantly seeing ~800nm plus (stacked with appropriate coloured filters the affect is the same as the variable wavelength filters on e-bay)

I'll try any strongly coloured filter I come across (colour correction & warming/cooling types tend to be too weak)

A few types of 'technical glass' can be effective too, but many of these are hard to get in sizes above 50mm, tend to come without any mounts & don't have guaranteed optical flatness. Still my 25mm Schott BG3 & U330 filters have worked well on lenses small enough for them to cover. They were among a set from work that were no longer used

Thanks for those observations. I have discovered some of those principles such as the strongly colored filters seem to be the most useful for IR and the 'correction filters' less so, although they can be used in conjunction with stronger filters to create a particular effect.

Here's a link to a thread where folks stacked multiple filters including 80A and 80C filters to get 'candy pink' effect, it is quite interesting. There are other threads on this as well.

https://www.dpreview.com/forums/thread/4639879#forum-post-66403437

 

[Larry Rexley]

Yes, I've shot AMT 203 on the e/b Sunset Limited. The Yellow/Black livery should be interesting.

And I remember when TECO was yellow without the wraps....I'm really old !

I managed to get on the bike yesterday and do some shooting in the city. I got a few TECO trolleys and some river shots.

My goal was to see if the native EF-M prime lenses do better than the Canon zooms. I have found that the EF-M zooms have a lot of CA at the edges of the frame, which tend to blur and frizz out the detail, kind of spoiling the shots, and wondered if the primes do better.

Here are some of my images:


Full-spectrum Canon ,M200, Rokinon 12mm f2 lens, f4, BG3 dual-band filter, 15 degree hue shift


Full-spectrum Canon ,M200, Rokinon 12mm f2 lens, f4, BG3 dual-band filter, 15 degree hue shift


Full-spectrum Canon ,M200, Rokinon 12mm f2 lens, f4, Spencer 590 nm filter


(Unconverted) Canon M6ii, EF-M 18-150mm lens, f5.6


Full-spectrum Canon ,M200, Rokinon 12mm f2 lens, f4, Spencer 590 nm filter


Full-spectrum Canon ,M200, Rokinon 12mm f2 lens, f4, BG3 dual-band filter, 15 degree hue shift


Full-spectrum Canon ,M200, Sigma 16mm f1.4 lens, f1.4, BG3 dual-band filter, 15 degree hue shift


Full-spectrum Canon ,M200, Canon EF-M 22mm f2 lens, f4, BG3 dual-band filter, 15 degree hue shift


Full-spectrum Canon ,M200, Canon EF-M 32mm f1.4 lens, f1.4, BG3 dual-band filter, 15 degree hue shift


Interestingly, I found that using the BG3 dual-band filter resulted in less chromatic aberration and sharper images with the Sigma 16mm f1.4 lens at f4 than using the Spencer Camera 590 nm filter. This seems counterintuitive.... I'd expect the filter passing blue and IR to have more CA than the 590 nm longpass filter. Perhaps the 590 nm filter is allowing a lot more deep IR closer to 900 nm.

 

[petrochemist]

With my full spectrum cameras I've found a fair number of old photographic filters can work well. Practically any of the red/orange yellow filters for black & white film will be long pass filters equivalent to high color/super color etc.

Blue & green filters are much more variable not all pass IR & this can vary between filters supposedly of the same type. Some that have worked well on one occasion have been quite useless a few months later (possibly from varying IR in the sunlight)

A variable ND can be very good allowing the visible portion to be adjusted while constantly seeing ~800nm plus (stacked with appropriate coloured filters the affect is the same as the variable wavelength filters on e-bay)

I'll try any strongly coloured filter I come across (colour correction & warming/cooling types tend to be too weak)

A few types of 'technical glass' can be effective too, but many of these are hard to get in sizes above 50mm, tend to come without any mounts & don't have guaranteed optical flatness. Still my 25mm Schott BG3 & U330 filters have worked well on lenses small enough for them to cover. They were among a set from work that were no longer used

Thanks for those observations. I have discovered some of those principles such as the strongly colored filters seem to be the most useful for IR and the 'correction filters' less so, although they can be used in conjunction with stronger filters to create a particular effect.

Here's a link to a thread where folks stacked multiple filters including 80A and 80C filters to get 'candy pink' effect, it is quite interesting. There are other threads on this as well.

https://www.dpreview.com/forums/thread/4639879#forum-post-66403437

I get fairly similar 'candy pink' results quite easily using my old SD14 bodies, the Foveon sensor sees IR in the red, but sees other things fairly normally. (The red is often be more intense)


Foveon IR shot (dust trap removed, X1 filter & fluorescent WB)

I've also had similar results with a 590nm filter & hue adjustment:

https://flic.kr/p/25BAv9s

&

https://flic.kr/p/2abeHbN

I don't want to find a reason to collect CC filters too, I've got far too many filters to play with already

 

[petrochemist]

Nice shots Larry

I particularly like the first, which seems to have different colours than the other BG3 shots, despite similar hue shift.

 

[Larry Rexley]

Nice shots Larry

I particularly like the first, which seems to have different colours than the other BG3 shots, despite similar hue shift.

I spent a lot of time processing the first image, as I plan to enter it in my camera club competition this month. I masked many areas and adjusted the relative brightness using DxO control points, and I brought up the contrast and saturation of the yellows, and also the luminance.

I forgot to mention that the Spencer 590 nm shots all had a 180 degree hue adjustment, amounting to a red/blue channel swap. That particular filter with a 180-degree adjustment gets the sky close to blue and the foliage orange-y yellow.

 

[Larry Rexley]

With my full spectrum cameras I've found a fair number of old photographic filters can work well. Practically any of the red/orange yellow filters for black & white film will be long pass filters equivalent to high color/super color etc.

Blue & green filters are much more variable not all pass IR & this can vary between filters supposedly of the same type. Some that have worked well on one occasion have been quite useless a few months later (possibly from varying IR in the sunlight)

A variable ND can be very good allowing the visible portion to be adjusted while constantly seeing ~800nm plus (stacked with appropriate coloured filters the affect is the same as the variable wavelength filters on e-bay)

I'll try any strongly coloured filter I come across (colour correction & warming/cooling types tend to be too weak)

A few types of 'technical glass' can be effective too, but many of these are hard to get in sizes above 50mm, tend to come without any mounts & don't have guaranteed optical flatness. Still my 25mm Schott BG3 & U330 filters have worked well on lenses small enough for them to cover. They were among a set from work that were no longer used

Thanks for those observations. I have discovered some of those principles such as the strongly colored filters seem to be the most useful for IR and the 'correction filters' less so, although they can be used in conjunction with stronger filters to create a particular effect.

Here's a link to a thread where folks stacked multiple filters including 80A and 80C filters to get 'candy pink' effect, it is quite interesting. There are other threads on this as well.

https://www.dpreview.com/forums/thread/4639879#forum-post-66403437

I get fairly similar 'candy pink' results quite easily using my old SD14 bodies, the Foveon sensor sees IR in the red, but sees other things fairly normally. (The red is often be more intense)


Foveon IR shot (dust trap removed, X1 filter & fluorescent WB)

I've also had similar results with a 590nm filter & hue adjustment:

https://flic.kr/p/25BAv9s

&

https://flic.kr/p/2abeHbN

I don't want to find a reason to collect CC filters too, I've got far too many filters to play with already

Interesting result.

I hear you about collecting filters, I have 15 or so already. eBay is great for treasure hunting.... I'll buy any new filter that looks promising for IR for $5 - $10 on a super discount, knowing I can resell and not lose money. I'll pay $15 - $20 if the filter normally sells used for $30 - $40 as many of the larger 67 and 77mm filters do.

Just found someone selling a set of 4 identical Hoya Green X0 filters which pass much more IR than the Green X1 in 55mm size for less than $10 which includes shipping! I can keep one and resell the rest.

 

[petrochemist]

Larry Rexley wrote:

I hear you about collecting filters, I have 15 or so already. eBay is great for treasure hunting.... I'll buy any new filter that looks promising for IR for $5 - $10 on a super discount, knowing I can resell and not lose money. I'll pay $15 - $20 if the filter normally sells used for $30 - $40 as many of the larger 67 and 77mm filters do.

Just found someone selling a set of 4 identical Hoya Green X0 filters which pass much more IR than the Green X1 in 55mm size for less than $10 which includes shipping! I can keep one and resell the rest.

I've been collecting filters & gels for years, and now have over 400, at a mean cost of ~£2.10.

Many of these came in bundles and have little use. Sadly my most expensive (an in body clip in hot mirror) got cracked and is now unusable. It's one of only two filters I've managed to damage over the last 30 years.

Fortunately many of my lenses take small filters.

 

[Larry Rexley]

After testing a lot of infrared lens and filter combinations, my favorite combo has become shooting my full-spectrum-converted M200 with the kit 15-45 IS STM lens and a Tiffen #47 (cobalt blue) filter. This is a dual-band filter which passes blue light but blocks visible green and red, passing IR below about 700 nm.

For some reason, the EF-15-45 kit lens is very well suited for infrared photography. It appears to be sharper in IR light than visible light, which is the opposite of the other native wides I've tried with IR (EF-M 22mm f2, EF-M 18-55, EF-M 18-150, Rokinon 12mm f2, Sigma 16mm f1.4) which are less sharp in IR with blurring and more CA in the corners than in visible light.

I also particularly like the color rendering of the Tiffen Blue 47 filter on the m200's full spectrum sensor. it's easy to compose false-color IR right in the camera without color swapping, and the results need a lot less color correction than other lenses and filters I've tried. All I do is white balance the camera on something gray before shooting.

Here's a shot from this morning with the Tiffen #47 and the EF-M 15-45, preceded by a similar shot with the Sigma 16mm f1.4 lens with a ZB2/BG3 blue filter (a dual-band filter like a Tiffen #47 but more violet than blue, needs a slight hue shift) and then a similar image taken with an unmodified m200 with the 15-45 lens. The IR #47 15-45 image I really like as the foliage appears to be 'luminous' -- it doesn't have the same impact in the other images.

The Sigma lens has an IR hot spot so I shot at f4 for some depth of field, but still had to do a lot of extra processing to minimize the effect of the hot spot. All processed in DxO PL5 and downsampled to 4k size.


Full-spectrum M200, Sigma 16mm f1.4, ZB2/BG3 dual-band blue + IR filter, f4, 1/200s, ISO 160


Full-spectrum M200, EF-M 15-45mm IS STM, Tiffen dual-band 47 filter Blue + IR, 15mm, f8, 1/60s, ISO 100


Stock M200< EF-M 15-45mm IS STM, 16mm, f5, 1/200s, ISO 100

I like the Tiffen #47 as it renders those beautiful royal blues and bright yellows almost right out of the camera, with nice separation and contrast. The blues in visible light imaging need a lot of processing and/or a polarizer to come out like that. Also the #47 shows those deep blues in water and other objects' reflections in a way that other IR filters and visible light shots don't.

 

[Larry Rexley]

Over the past few months I've tried quite a few filters with my full-spectrum Canon M200, enumerated in detail in prior posts in this thread.

I've been experimenting with combining filters, to find interesting and useful combinations to create nice images - inspired in part by the 'candy pink' effect described in these 3 threads:

https://www.dpreview.com/forums/post/66063630

https://www.dpreview.com/forums/thread/4585259

https://www.dpreview.com/forums/thread/4619006

I tried the 'original' candy pink (Yellow-Green filter plus a blue KB20 filter) and the 'improved' candy pink (yellow-green filter, blue KB20 filter, and 80A filter), getting nice results that are fairly consistent with what others posted in those threads.

I then tried many other combinations to see what other effects I could get, experimenting to find 'optimal' sets of filters for each effect, that would require the least amount of post-processing.

The only 'trick' I used was to Hue-Shift the 'cyan' channel in the Candy Pink images roughly 30 degrees in DxO Photolab's 'HSL' (Hue-Saturation-Luminance) control's hue wheel to change the cyan skies to a more natural sky-blue.

Otherwise I adjusted exposure levels, shadow levels, overall saturation (and vibrance to the same degree), and sometimes the saturation and luminance of one or two color channels to bring out a certain color, all techniques that are fairly common in a standard workflow.

Another technique I used was a global hue-shift for images taken with an infrared-cutoff filter of 520 nm and below. Such a hue shift 'restores' the natural order of the colors our human eyes expect, so that the shortest wavelengths hitting the sensor appear blue and the longer wavelengths appear red and yellow (as described in a prior post in this thread).

The techniques and results shown below are only possible with a camera modified for full-spectrum photography. They are not possible with an unmodified camera, or with a camera modified for reduced bandwidth photography such as a conversion to 550 nm, 590 nm, 720 nm, visible+hydrogen-alpha for astrophotography, etc.

For all images below, I white-balanced each image in post processing on the color-neutral concrete on the bridge in the middle of the image, to calibrate the white balance before making any adjustments.

Original Candy Pink photography

Here's an image with a Hoya X0 (yellow-green) filter stacked with a B+W KB20 (blue) filter. This is the 'original candy pink' photography combination described in the thread in the first link referenced above in this post:


Original Candy Pink: Hoya X0 & B+W KB20 filters, no hue shift

Here's the same image with the cyan channel in DxO Photolab shifted about 30 degrees, which takes the cyan parts of the image to a more natural sky blue. This cyan hue shift was used for the next 3 'candy pink' images:


Original Candy Pink: Hoya X0 & B+W KB20 filters, 30 degree cyan shift to the blue

Improved Candy Pink

Improved candy pink, described in the second thread linked above in this post, adds a blue 80A (or 80C) filter to the two filters used in the original candy pink style, resulting in a more vibrant image. This image also uses a 30-degree cyan channel shift for more natural sky color. I also increased the red channel saturation and reduced red luminance slightly to bring out the candy pink:


Improved Candy Pink: Hoya X0 & B+W KB20 & Calumet 80A filters, 30 degree cyan shift to the blue, red channel saturation brought up and luminance reduced

Improved Candy Pink using a Green (X1) filter instead of Yellow-Green (X0)

I also tried Improved Candy pink but used a green Hoya X1 filter instead of the yellow-green Hoya X0 filter, so this image was taken with an X1 filter, KB20, and 80A. It has the 30 degree cyan shift to make the skies blue, plus a slight bump to the red channel saturation to bring out the pink. The result is a different shade of pink that's more reddish:


Improved Candy Pink: Hoya X1 & B&W KB20 & Calumet 80A filters, 30 degree cyan shift to the blue, red channel saturation brought up

Candy Pink with a single filter: Yellow Y15 and global hue shift

I found that a Tiffen Deep Yellow Y15 filter alone, with a global hue shift of about 130 degrees on DxO's HSL 'hue wheel' plus adjustments to increase saturation of the sky and foliage also results in very nice 'candy pink' images. So it appears to be possible to do candy pink photography with single commonly used yellow filter instead of 3 filters stacked (plus the KB20 filter is not easy to find!):


Yellow filter Y15 Candy Pink: Tiffen Y15 filter (only), global hue shift of 130 degrees, green channel saturation increased and luminance decreased to bring out the hue-shifted sky, blue channel saturation increased and luminance decreased to bring out the hue-shifted foliage

Autumn Yellow photography

Personally, I prefer the foliage in infrared images to be more of an autumn-yellow color. Perhaps it's because I grew up in New England with those gorgeous Fall colors, plus yellow and a blue sky in an image are nice complementary colors.

This effect can be achieved with a Tiffen #47 blue filter alone, with saturation and luminance adjustments to the yellow, orange, and blue channels:. The Tiffen #47 filter acts as a dual band filter, passing violet and blue in the visible spectrum, plus infrared below about 700 nm, blocking visible green and red:


Autumn Yellow: Tiffen #47 filter, orange, yellow, and blue channel saturation and luminance adjusted

I found that adding an FL-D or FL-W filter (usually used for fluorescent light conversion) increases the overall vibrance of the image, yielding an even better result. I liked the result of combining the Tiffen #47 with the stronger FL-W filter best. Also, with this combination, less saturation and luminance adjustment was needed in post. The result out-of-camera was more pleasing to me:


Autumn Yellow: Tiffen #47 & Hoya FL-W filter, orange, yellow, and blue channel saturation and luminance adjusted

I also found that I could get a similar Autumn-Yellow effect by using TWO KB20 filters stacked. The KB20 filter, unlike the Tiffen #47, allows a small amount of green and red light to pass... but stacking two of them has an effect approaching that of one Tiffen #47 filter.

My 'second' KB20 filter is really a 'QB2' filter ordered from China on eBay, however its spectral response curve and performance on my M200 is identical to my B+W KB20, so I believe it can be considered a KB20 filter. With this combo only the Yellow channel saturation had to be increased, the blue channel actually had to have its saturation decreased. The Tiffen #47 must pass somewhat more IR than the KB20 filters:


Autumn Yellow: KB20 & QB2 (equivalent to a KB20) filters, yellow channel saturation increased, blue channel saturation decreased

As with the Tiffen #47 autumn yellow image, adding the FL-W filter to 2 KB20's seemed to improve the vibrance and overall appearance of the image to my taste:


Autumn Yellow: Kb20 & Qb2 (equivalent to a KB20) & Hoya FL-W filters, yellow channel saturation increased, blue saturation decreased

Autumn Orange photography

The other interesting effect I found was a filter combo that caused the foliage to have a nice Autumn-like orange color.

This was achieved using the Hoya Green X1 filter and a Hoya R25A red filter, and doing a 180-degree global hue shift (similar to doing a red-blue channel swap which is common in IR photography). For this image, the overall saturation was increased, then the blue channel saturation was increased and its luminance decreased to bring out the 'orange' color in the foliage


Autumn Orange: Hoya X1 & Hoya R25A filters, Global hue shift of 180 degrees (like an R-B channel swap), Overall saturation increased, blue channel saturation increased and luminance decreased to bring out the hue-shifted foliage color

I hope this is helpful to some folks, and perhaps encourages experimentation for folks who have a full-spectrum camera!

 

[JeanPierre Martel]

Your experiments are very interesting.

Thanks for sharing.

 

[Satyaa]

Are these filters that you mount on the front of the lens via the threads, via sensor modification, or a combination of both?

Thanks.

 

[Larry Rexley]

Are these filters that you mount on the front of the lens via the threads, via sensor modification, or a combination of both?

Thanks.

As it's a full-spectrum modified Canon M200, the stock filters were removed from the sensor. All my images were taken with filters threaded on the front of the lens. Some of my filters are 55mm, some 67mm, and some 77mm thread size and I used step-up rings as needed.

 

[Fedia Le Grill]

Wow if this graph is true I want a canon 350D immediatly :

https://www.researchgate.net/figure/Spectral-sensitivity-of-unfiltered-Canon-350D-sensor_fig1_229046541

the peaks in red first, then green, then blue seem distinct enough to produce a real three-color image only with the 700 to 1000nm IR spectrum !

The thing is I have a canon 1000D full spectrum (wich is only 3 years more recent than the 350D, it has a different sensor though...) and I never noticed that they were any color variation in the tones it produced when I use it with a 720nm filter....

Does it mean that the that objects reflecting IR refelect all IR wavelengths the same way ? That it is impossible to extract Three primaries like for exemple 700nm, 800nm and 900nm since all those wavelengths are reflected the same way ?

I did notice though that when I use a 720nm filter on the camera and I add a GRB3 filter that cut the longer IR wavenlegths the image color slide towards Red. I suppose that If I added a 850nm filter that cut the shorter IR wavelength the image color would slide toward Blue or deep violet. I dont have this filter just now, but I ordered it.

 

[Fedia Le Grill]

I find the images number 7 and number 8 particulary intresting. the dark purple color of the trees and the natural tones of the rest is charming.

Now I dont really get why when you add the violet FL-W filter on top of the green X1, the image actually goes more purple instead of going less purple since the FL-W is antagonistic to the X1. To me it should reduce its effect (though not cancel it since the X1 is more powerfull than the FL-W).

It seems wrong to me... Do you have an explaination ?

 

[Satyaa]

Thanks.

 

[barbara j]

Are these filters that you mount on the front of the lens via the threads, via sensor modification, or a combination of both?

Thanks.

My camera is a full spectrum modification. I can shoot IR without adding any filters. I am experimenting with the simulated filters built into the camera's OS. I do have an over the lens filter that replaces the one removed from over the sensor to allow me to shoot the camera "normally".

The examples in this thread are shot using a full spectrum mod with filters added over the lens. There are many filters, especially older ones, that will give interesting effects.

--

@luredbylight

 

[JeanPierre Martel]

Are these filters that you mount on the front of the lens via the threads, via sensor modification, or a combination of both?

Sorry for my late reply.

The filters are stacked in front on my lenses, as shown in the picture below.

The main problem is vignetting on wide-angle lenses. On theses lenses, you need to use larger filters and, consequently, step-up rings.

The photo hereunder shows my previous setting. These days, on my main camera, the zoom accepts 62mm filters but the ones that I use are 15mm wider.

 

[Larry Rexley]

Wow if this graph is true I want a canon 350D immediatly :

https://www.researchgate.net/figure...f-unfiltered-Canon-350D-sensor_fig1_229046541

the peaks in red first, then green, then blue seem distinct enough to produce a real three-color image only with the 700 to 1000nm IR spectrum !

The thing is I have a canon 1000D full spectrum (wich is only 3 years more recent than the 350D, it has a different sensor though...) and I never noticed that they were any color variation in the tones it produced when I use it with a 720nm filter....

Does it mean that the that objects reflecting IR refelect all IR wavelengths the same way ? That it is impossible to extract Three primaries like for exemple 700nm, 800nm and 900nm since all those wavelengths are reflected the same way ?

I did notice though that when I use a 720nm filter on the camera and I add a GRB3 filter that cut the longer IR wavenlegths the image color slide towards Red. I suppose that If I added a 850nm filter that cut the shorter IR wavelength the image color would slide toward Blue or deep violet. I dont have this filter just now, but I ordered it.

It looks like the 3 color curves were 'normalized' for that 350D spectral chart so that their peaks are at the same height in the chart, and doesn't reflect their 'true' responses relative to each other.

Otherwise the chart looks similar to other Canon cameras and those from other brands.

The distinct 'red' peak is around 600 nm, which is why a 590 nm 'orange' filter gives distinct (swapped) colors, and a 550 nm filter or other yellow filter like a Tiffen deep yellow #15 gives even stronger colors.

 

[petrochemist]

Wow if this graph is true I want a canon 350D immediatly :

https://www.researchgate.net/figure...f-unfiltered-Canon-350D-sensor_fig1_229046541

the peaks in red first, then green, then blue seem distinct enough to produce a real three-color image only with the 700 to 1000nm IR spectrum !

The thing is I have a canon 1000D full spectrum (wich is only 3 years more recent than the 350D, it has a different sensor though...) and I never noticed that they were any color variation in the tones it produced when I use it with a 720nm filter....

Does it mean that the that objects reflecting IR refelect all IR wavelengths the same way ? That it is impossible to extract Three primaries like for exemple 700nm, 800nm and 900nm since all those wavelengths are reflected the same way ?

I did notice though that when I use a 720nm filter on the camera and I add a GRB3 filter that cut the longer IR wavenlegths the image color slide towards Red. I suppose that If I added a 850nm filter that cut the shorter IR wavelength the image color would slide toward Blue or deep violet. I dont have this filter just now, but I ordered it.

It certainly doesn't look right to me.

I'd expect an unfiltered sensor to show some sensitivity below 400nm & quite a bit more in the 900-1000nm region...

The IR peak in the blue channel also looks out of proportion. I have seen subtle increases in blue for some wavelengths above 700nm (often showing up dyed hair) but I've not seen other spectra showing much more than 10% variation between the channels above 800nm.

Usually sensors have a distinct red bias between 700 & 800nm with relatively similar transmission in all three channels above that.

With the right WB there can be a reasonably amount of colour variation with 720nm shots, (not something I've seen with 860nm & 960nm filters). Some objects such as foliage tend to reflect all wavelengths of near infra red, but this is certainly not true of everything.

 

[Fedia Le Grill]

Interesting ! I made a little experiment.

I took these pictures of an halogene ceiling lamp using my 720nm filter.

I made a white balance so it looks like this :

Then I added the GRB3. the tint slide towards Yellow.





Then I took the Lee swatchbook and searched for the gels that cut the deep reds and very near infrared. Most gels look transparent but some appear blueish. I picked "primary green" and used it a s a filter :

So yes color variation is perceptible within the IR Spectrum. But unfortunately it seems two-dimensional. It goes on a straight line from yellow to blue. Not exactly a rich color palette... I still wonder what color It would look with a 850nm filter.