RGB Stacking within the UV Waveband

[Cadmium]

EDITOR'S NOTE:

Posts within the topic Modelling UV camera response from Bayer filter meaurements about using narrowband UV-pass filters to make RGB stacks were moved here to their own topic in the Filters section.

---

 

All this gets me thinking of three narrow band UV filters, shot separately. Low, medium, and high UV-A bands. Say, 330BP15, 360BP15, and 380BP15, just as a rough idea.

Those bands are in no way thought out, they could be adjusted for peak and width to whatever one thinks is best, but I might tend to keep them a bit separated rather than overlapping.

This would do well to have a deep reaching UV transmitting lens to eliminate any lens transmission curve slope that would truncate or basically remove transmission from the lower range bandpass filter.

We have the sensitivity drop off of the sensor, so a monochrome UV conversion would do well for this set up also, however it could still be done with a standard full spectrum conversion.

Whether we use a monochrome UV camera or not, we still have a drop off in sensitivity to UV with the sensor.

Given the sensor sensitivity drop off, each shot would have a different exposure time to adjust for the sensitivity of the individual bandpass.

I think the separate exposures give us something unique compared to a single UV filter shot, because given the sensitivity drop off, our UV shots may be heavily weighted in the upper range of UV-A.

Assigning each band to RGB, in whatever arrangement one desires, might be interesting... other than just the laborious nature of shooting three shots.

I am sure it has been done, but I don't recall seeing such, and it would require some special and expensive filters, especially if you use something as large as 52mm filters.

I have of course seen it done with the IR range...

Just an idea.

Tri Color UV, or TrU Color? ;)

Well then, OK, skip the names, but the idea that the lower, middle, and upper bands would all have optimal exposure times for their own sensitivity should produce more presence in a composite mix

than seen with a single broadband UV filter shot, as well as great flexibility with color arrangement.

The Sparticle shows green, yellow, lavender/blue (basically), but have you ever seen the 340nm green in reflected shots?

I don't know, but lets say that the idea produces a stronger 340nm presence...

...and lets say that we apply the three color bands, green, yellow, lavender/blue, in whatever arrangement to RGB.

Just imagine the possibilities.

[Andrea B.]

Steve, I like this idea. I have not seen it done either.

I've played with this on the IR end using the MaxMax set of three IR bandpass filters. I did not get very interesting RGB stacks from them because IR tends to go monochrome very quickly past about 750nm. So the 921-958 nm and 795-860 nm filters produce pretty much the same photo.

They are beautifully made filters though and used separately make very nice IR photographs so the $$$ was not wasted.

An RGB stack with three UV narrowpass filters should be much more interesting.

[Cadmium]

Steve, I like this idea. I have not seen it done either.

I've played with this on the IR end using the MaxMax set of three IR bandpass filters. I did not get very interesting RGB stacks from them because IR tends to go monochrome very quickly past about 750nm. So the 921-958 nm and 795-860 nm filters produce pretty much the same photo.

They are beautifully made filters though and used separately make very nice IR photographs so the $$$ was not wasted.

An RGB stack with three UV narrowpass filters should be much more interesting.

 

If you only used something like three ###BP10 filters, it would probably look pretty cool.

I am wondering if the separation of the exposure time should show something quite unique when assembled.

I wonder if our lower UV isn't being extremely minimized in our full range exposure, even using a UV-Nikkor (or the like) because the upper UV takes over because the sensor sees that so much easier and stronger.

I have never thought about that aspect in the same way before.

It is like using a BG40 filter for visual, you don't see the UV, because the sensor sees the visual so much easier and stronger, even though the BG40 transmits UV,

so any UIV is just minimized to nothing.

However, if we stack the BG40 with UG1 (for example), then the sensor sees the UV range of the BG40, and nothing of the rest of the BG40 range above that.

Is that what is happening to our UV below ~350nm? Is the lower UV range just disappearing to our fancy filters and lenses that see down to 320nm and below,

simply because the sensor is weak to that range?

Similar in a way to how UG5 stacks show UV + B/G evenly because UG5 transmits UV more than visual, so therefor 'weights' the stack

for the UV to be more 'sensed' by the sensor which doesn't sense UV as much as it does visual.

I think that any 3 narrow BP filters in the UV-A range should make for a pretty interesting test. Space them apart, and don't worry too much about the details as long as they don't overlap,

I think it might looks interesting.

Basically creating a visual spectrum out of the UV-A range. Of course, this has absolutely nothing to do with bees, but life and UV is not all about bees... :)

 

Yeah, with IR you basically need to have one of the filters be a (1) red range filter, (2) an RG9 or such, and (3) a higher filter like 850 or 1000.

For the red range filter you could even create a stack that blocks most of the IR, say RG590 + BG40 or BG38... that will give you a very high red BP, and should fade into the RG9 range pretty well.

Bjorn has some such pic posted on here somewhere... from Sierra Nevada, as I remember.

[Jim Lloyd]

Oh and as regard Cadmium's TrU colour - great idea! - surely that is how we would make a UV camera to do 1. above "a human in UV land" - it's funny that UV image tend to look like they do because we are dealing with the accidental crumbs left behind in the design of a camera for the visual range ...

[JCDowdy]

All this gets me thinking of three narrow band UV filters, shot separately. Low, medium, and high UV-A bands......

........

Tri Color UV, or TrU Color? ;)

 

Yep!

 

That is the classic way of doing scientific multispectral imaging with a monochrome camera back and filters. The renowned lens designer Brian Caldwell has designed a UV-Vis-NIR super lens for this system. The big difference of course is that it is not an out of band CFA transmittance but a well defined spectral band. Astronomy filters from Baader abound in many colors for this very purpose. With a suitable set of filters, non existent as far as I know, in the UV one could recombine UV bands into a pseudocolor UV palette but then you have come full circle and must decide what color to put where! ;)

 

[added back links somehow lost]

[Cadmium]

John, Have you seen any examples of UV-A composites?

[JCDowdy]

John, Have you seen any examples of UV-A composites?

 

Nope, But I have been searching specs on potentially suitable filters for a while. What I was imagining was a filter set in the CIE UV-B, UV-A2, UV-A1 bands.

 

I think the Straight Edge U might be a possible choice for the longwavelength UV-A1, perhaps an Omega 335WB70 for the midrange UV-A2 and some kind of ~300 to 320nm BP~30. Problem is that most of the candidates are dichroic and begin transmitting again in the Vis or NIR range of the camera. Another problem is that the MAX T% is intolerably low for the shorter filters where the sensitivity also falls off.

 

Added:

I think some overlap, as with standard Bayer CFA, would work better.

[Cadmium]

John, I am thinking to trying it with just these:

340PB10, 360BP15, and 380BP10.

Those are around I think for a fair price (I think), size?

Depending on price, might be an interesting test.

I have all those in 12.5mm size, so I guess if I wanted to use those I could possibly.

[Cadmium]

Quick and dirty Tri-Color UV-A separation and post assembled composite test of black light bulb.

Using Baader U stacked with the following, 380BP10 = blue channel, 365BP15 = green channel, 340BP10 = red channel.

Each shot is monochrome.

I don't expect anyone to get too excited about this, I cut every corner in the book to shoot this fast, and it is only a light source, very little reflected target, other than the little bit of aluminum reflector.

But I see some color differentiation going on, which may be more interesting with a completely reflective target.

High ISO, rather dirty 12.5mm filters (need cleaning) in front of Badder U, hand held, AP mode, not focused either.

 

Tri-Color UV composite

 

Baader U alone

[Cadmium]

Here is a graph example using the 'high transmission' version Asahi 10nm wide bandpass filters.

https://www.asahi-spectra.com/opticalfilters/bandpass_filters.html

They have three sizes, 25mm dia, 50m dia, and 50x50mm, but only 25mm dia and 50x50mm for all three of these BP's.

The high transmission versions cost a lot, about $400 (25mm dia) to $800 (50x50mm) each.

These would need to be stacked with a Baader U or some such to suppress the 760nm+ IR range leaks of these specific filters (not all BP filters have such leaks).

[colinbm]

Steve, the Baader U covers these filters better then S8612 ?

Col

[Cadmium]

If you are using a UV-Nikkor, then I think maybe yes, if you are using anything like a Kuribayashi 35mm or less, then no.

[colinbm]

OK thanks Steve.

Col