Jump to content
UltravioletPhotography

Recommended Posts

Andy Perrin

This is a rainbow shot with my "single shot IRG" method using a Tiffen#12 + DB850 filter. The combination transmits roughly 550-650nm + 800-900nm, without the 700-800nm region which otherwise contaminates the IRG since that part of the IR contributes unequally to the three channels (unlike the 800-900nm band) and so cannot be subtracted off by removing the blue channel without supplying unknown multipliers.

 

Here is my estimate of the spectrum based on the Tiffen#12 data and DB-850 data supplied by the manufacturers:

post-94-0-53954900-1593739417.png

 

The processing has been documented by me extensively already on this board, so I will just present the final image.

 

(Unfortunately I did not record shooting data, and in any case this is a panorama, so I can't give the exposure.) Lens was the metal EL-Nikkor 80mm/5.6.

 

post-94-0-06323600-1593739609.jpg

 

Interestingly, some inner bows can be seen.

 

Now compare to the visible photo, taken by iPhone XS Max:

post-94-0-86005000-1593739963.jpg

Link to comment

Very cool.

You have a double rainbow, you can see the other in the next tree over.

Which makes me feel that your LP500 is leaking a lot of UV, as I see a sharp UV band not in the iPhone shot that first thought could be second harmonic IR, but can't be based on second rainbow spacing.

So you have UV, little bit of blue, then big green band and little bit of IR.

Very cool multi band shot.

Link to comment
Andy Perrin
Nah. The combined blocking is extremely good in UV. You are probably seeing second harmonic IR. Seriously, both the DB850 AND the Tiffen are blocking the UV. It must be OD10+. And on top of that there is hardly any UV in sunlight so it would be drowned out by the vis and IR even with no filter at ALL.
Link to comment
Can't be based on spacing. Stick a U340 filter only or a ZWB1 filter over this combination and image something with a UV light. If no UV and no 700nm band leaks than the U340 would cause it to be black. But I think you have a lot of 360nm leakage here.
Link to comment
Andy Perrin

David. Again, there is hardly any UV in sunshine relative to vis+IR. Shining an intense UV light in the absence of vis+IR is not a good test. Your "based on spacing" is probably the logical fallacy here.

 

For example, they are apparently even seen in visible light sometimes:

https://www.atoptics...bows/supers.htm

Link to comment

David. You are being ridiculous. There is hardly any UV in sunshine. Shining a UV light is not a good test. Your "based on spacing" is probably the logical fallacy here.

 

For example, they are apparently even seen in visible light sometimes:

https://www.atoptics...bows/supers.htm

 

Yesterday.....full sunshine

 

post-31-0-06286000-1593748255.jpg

Link to comment

A rainbow is a giant diffraction grating. The light is cut into its wavelengths exactly on their spacing. UV, blue, green, yellow, orange red and IR. As you move a defined unit of space in the left or right you move into those wavelengths.

You have a second rainbow very far in the next tree over from the main rainbow.

So then you can count off the diffracted bands in order from the main rainbow.

Then compare with the visible shot to see the colours.

You have UV in that main rainbow clearly coming through as a weak band not visible in the visible shot.

If second harmonic, would see some sign of red in that visible image. But its not there and the spacing wouldn't make sense for that.

Link to comment
Andy Perrin

(There are a lot of things wrong with that, starting with the fact that my filter is missing 700-800nm, so the spacing is not what you might expect, and that the supernumerary bands apparently reverse the order of the colors.)

 

I don't accept your conclusions there. I don't think there's anything else to be said for now. We can let others weigh in.

Link to comment

You never have to accept my conclusions.

 

Just take a shoot of something under UV light with tiffen 12 + DB850 filter. Than with 2mm U340 or U360 or ZWB1 glass if you have it stacked on top of the tiffen 12 + DB850.

I think you will see something.

Link to comment
Andy Perrin

Taking a shoot under UV light without IR and vis in the image will...probably still not show anything, since the blocking is outstanding here, but it's not a good test because the WHOLE POINT is that the vis+IR drown out weak UV.

 

I don't have a functional UV light source at the moment (all batteries are dead) and it's the middle of the night. I may try tomorrow in sunshine (which is the only reasonable choice here) but I don't think I'll see anything but black. I'm not sure what to use as a filter -- maybe my 330WB80 -- because using U340 would just show the IR leak of that filter.

Link to comment

Andy, Very cool! :smile:

 

As far as UV, it would seem from the graph of the DB850 that it does pass some UV, given it crosses the 400nm line at a fairly high level,

however, Tiffen #12 should be blocking everything below its 500nm range cutoff, wouldn't you think? Therefore why would there be UV expected in that stack?

 

I have both of those filters here, and a few others, if anyone can design a test for me to try.

Link to comment
Andy Perrin
The DB-850 passes 400-385nm in UV but the TIffen-12 blocks that too, so I see no reason to expect this to leak UV at all, especially when much stronger IR and visible are competing for the same pixels.
Link to comment
Andy Perrin

Cadmium, the settings (as best I recall anyhow, I didn't write them down and it was nearly a week ago) were EL-Nikkor 80mm, F/8, 1/250", ISO100, DB-850+Tiffen#12. I think you should use sunshine because you can always force an LED torch through almost anything. You yourself tell people not to test with them, and I agree. The problem is what to stack with, because I think the U-glasses all have IR leaks, don't they? Perhaps the Baader U?

 

I don't think we need to do any special processing to the images because we just want to know if we see light or not.

----

 

Here is another photo from the same day:

post-94-0-09517400-1593757502.jpg

 

Vis (a week later, under different, less dramatic light):

post-94-0-44353800-1593796565.jpg

Link to comment
Andy Perrin

The last image is very colorful. What are the "normal" colors of the building?

I added the visible above, as it appears today. Lighting has changed, alas.

Link to comment

So, just guessing:

- The bricks reflect almost red only;

- The oxidized copper corner reflects green and infrared;

- The yellow lines on the street reflect green and some red, but absorb infrared.

 

As you said, it must be noted that the conditions in which the two images were taken are different, and probably the VIS/IR ratio is different.

 

Link to comment

Tiffen 12 is a brick wall to UV. No way is any UV getting through it.

 

The fainter secondary rainbow outside the first is a normal feature of rainbows, as are the smaller bows inside the primary (called supernumerary arcs.) Note the reversed color order of the secondary. In theory, an infinite series of bows at different angles exists, but it is very rare to see a tertiary or higher bow in real life. An IRG rainbow will be slightly position-shifted with respect to its all-visible counterpart.

 

I don't often see green in an IRG image, except from red LEDs. Red surfaces that are IR-dark are quite uncommon.

Link to comment
Andy Perrin
What’s interesting here is that you can see the supernumeraries in IRG but not in visible.
Link to comment

Tiffen 12 is a brick wall to UV. No way is any UV getting through it.

 

The fainter secondary rainbow outside the first is a normal feature of rainbows, as are the smaller bows inside the primary (called supernumerary arcs.) Note the reversed color order of the secondary. In theory, an infinite series of bows at different angles exists, but it is very rare to see a tertiary or higher bow in real life. An IRG rainbow will be slightly position-shifted with respect to its all-visible counterpart.

 

I don't often see green in an IRG image, except from red LEDs. Red surfaces that are IR-dark are quite uncommon.

 

Sure I can believe that explaination, but only after a controlled test placing a U340 infront of the stack.

I like to see controls first to rule out the obvious explanations. Then will believe the harder to see things.

 

Some filters degrade over time if bleached. Also some leak if pushing enough light.

 

That thing I can't see in the visible is very strong in the full spectrum, conveniently lining up with were I expect to see the UV section of a rainbow, which in itself shouldn't be that strong either.

Link to comment

I can see the smaller inne bows in the visual picture too, even if they are fainter.

They are there and likely enhanced by the image processing, that increase the contrast.

Link to comment

David, I tried this here:

I used a Convoy S2+ Nichia 365nm UV torch, through a Tiffen #12, no UV makes it through.

The same using the DB850, there is the tinniest bit of barely detectable fluorescence, no LED image.

Stacked together there is nothing, just some fluorescing dust on the torch side of the filters.

That is with U-340 on the torch to remove any 400+nm light from the torch emission.

 

Does that help? Would you like some other test?

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...