32 replies to this topic

[Bernard Foot]

Following a discussion in this topic - https://www.ultravio...various-lenses/ - I thought I’d see if I could get an image at around the 300nm mark.

I think I am able to get an image at around the 303nm mark, but before I get too excited I’d be interested in other members’ opinions as to whether I really am getting an image at that wavelength, or whether there could be some type of contamination.

The kit used was:

• Sony A6000 full-spectrum conversion
  
• Simple plano-convex lens made from UV-grade Fused Silica (at about f/3.7)
  
• 303BP10 filter from Omega Optical
  
• 4 x WS560 cheap-from-China flashguns (de-lensed)

The main problems were likely to be:

• Lack of illumination provided by the flashguns (as indicated in that other post). Stuff on the internet indicates that a Xenon tube would provide illumination to well below 300nm, but the lower wavelength limit could be severely restricted depending on the material the tube was made from.
  
• The high transmission of the filter at 800nm (see below). All I had available that might block this was a U340 + ZWB1, plus ZWB1 filters over the flashguns – but IR leakage has to remain a real threat.

The target was a piece of paper with “UV Only” written on it using an ink that was transparent to IR and “UV + IR” in an ink that would show up in both UV and IR. (Unfortunately I didn’t have an ink that would show up in IR but not UV). A “normal” IR shot using an R72 filter looks like this:



The image I got using the 303BP10 + U340 + ZWB1 was this:



Was this really a UV image? Yes – using a Hoya UV(0) filter completely destroyed the image.

Is there any visible leakage? No – as shown by the Hoya UV(0).

Was there any IR leakage? I don’t think so – the “UV only” writing should have been a lighter grey or not visible at all if there was an IR component to the image. More importantly, the Hoya UV(0) would have allowed any IR through, but the blank image from that filter indicated there was no IR present. Similarly, adding an R72 filter killed the image completely. It was also noticeable that significant re-focussing was needed between the UV the IR images shown above – indicating that the UV image did not have an IR component. (In fact the helicoid I was using did not have enough travel to get the IR image in focus.)

Was there leakage from other parts of the UV range? I don’t think so. Switching from the UVFS lens to a Cassar S lens (at f/2.8) killed the image – this is as expected, as the Cassar S transmits down to about only 315nm. Adding a 345BP20 or 380BP25 filter completely killed the image. Adding a 315BP25 filter did provide a dimmed image, but this would be expected as the transmission curves of the 303BP10 and 315BP25 filters overlap (see images below).

[Stefano]

Well, judging from what you said, I would say you got to 303 nm. You tested UVA, visibile and IR leakages and found nothing.

[colinbm]

Good on you Bernard.

[Andy Perrin]

Definitely seems plausible. Can you do a UVA and UVB comparison of something like glass that we know usually darkens, like Jonathan did with that vase awhile ago?

[Bernard Foot]

Andy Perrin, on 21 November 2020 - 17:34, said:

Can you do a UVA and UVB comparison of something like glass that we know usually darkens, like Jonathan did with that vase awhile ago?

I'll give that a go tomorrow - right now I've wound down for the day with a couple of glasses of wine - so not the best time to carry out highly precise, scientific experiments ...

[dabateman]

Looks like you have all the controls in place.
I have the exact same 303bp filter and its good when you block it as you have done. It has the most 303nm transmission. My other 300bp10 filter only lets through about 10% at most.

The advantage I have found with using the single element lenses is one of there biggest disadvantages. That being that you need to refocus for each wavelength. This is good in that my 50mm BICX needs to move at least 6 inches back to get IR in focus. So it helps to isolate the UV signature.

You can get affordable continuous 302nm source lamps. Even the ExoTerra UVB bulbs will output some 302nm.

[Bernard Foot]

dabateman, on 21 November 2020 - 18:28, said:

You can get affordable continuous 302nm source lamps. Even the ExoTerra UVB bulbs will output some 302nm.

I have an Exo Terra, and I will try using this. From what I remember, it is good at around 300 nm but it has quite a high IR footprint too.

[dabateman]

Bernard Foot, on 21 November 2020 - 18:35, said:

I have an Exo Terra, and I will try using this. From what I remember, it is good at around 300 nm but it has quite a high IR footprint too.

Yes but for the ExoTerra its 708nm if I remember correctly. The filter hits just above 780nm. The flashes I think would have much less at 303nm and much more light at 800nm+.

But be careful with them. That light is still very dangerous. But you can get it really close to the subject as its not hot.

Edited by dabateman, 21 November 2020 - 20:52.

[JMC]

Bernard, it looks like you are getting a 303nm image. If you view the RAW file in Darktable or whatever RAW file viewer you have, without giving it any white balance, what colour is it? When I have tried doing 313nm imaging with my cameras with the Bayer filter still attached (a Canon and a Sony) when using a UV lens, I get a very strong green colour cast. I guess this is because of all 3 of the Bayer filters, the green has a slightly more transmission at that wavelength and dominates the image. Be interesting to know if you get a similar effect.

[Bernard Foot]

Hi, Jonathan.

Yes, I get green, as I do with the 315BP20 filter. Here is a shot with no profile and no WB set:



BTW - I also tried to get a shot in daylight. Here is what I got - straight out of the camera, set to give monochrome JPEGs. The image is very faint, and increasing the exposure just made it lighter without getting any more detail. This would be in keeping with the amount of 303 nm light in daylight being close to zero. And a 315 nm shot for comparison. In fact, if you look closely at the 303nm image you can see signs of IR contamination - light foliage, some other objects coming out light rather than dark, clouds standing out against the sky, and obvious focus shift - so there may be no UV component at all in this shot.

Here is the shot taken through the 303BP10 + U340 + ZWB1 (f/11, 30 sec., ISO 500):



Here for comparison is a shot taken just 15 nm or so away, through a 315BP20 + U340 + ZWB1 (f/11, 20 sec., ISO 400):

Edited by Bernard Foot, 23 November 2020 - 17:36.

[Stefano]

Seeing the world in UVB always fascinates me.

In your 303 nm image you have for sure no visible light leaks (the OD anywhere between 400 and say 650 nm should be crazy high), you may have a weak point in the 700-800 nm region (where U-340 and ZWB1 leak a bit) since you may be getting some of your filter spike leak in the 760-800s. You may also have some UV leakage, but this wouldn't make things much lighter than normal, especially foliage (even if UVB is absorbed more in general than UVA by most materials, but not by foliage).

To catch 303 nm light in sunlight you have to have a strong filter, blocked really well, and we all know it isn't easy, but your camera/sensor are for sure capable of that.

[Andrea B.]

Bernard, did you make the image in a dark room (or dark closet) to prevent IR contamination from ambient light?
I also don't think you are getting IR contamination, but a dark room adds another "filter" to the experiment.

[Bernard Foot]

Andrea - no, it wasn't a dark room. Perhaps I'll retry with a properly dark room. But as I was using flash, the shutter speed was 1/200 sec so I can't believe there would be enough ambient IR getting through the 303B10 + U340 + ZWB1 to record any kind of image.

[Bernard Foot]

Stefano - I'm sure the leak is in the IR where the U340 and ZWB1 have their secondary transmission peaks.

[Stefano]

Did you use every piece of U-340 and ZWB1 you have around? I think yes, but if not, you need to put as much glass as you can.

Also, are there other filters you have which can help you? If you have an Omega 330WB80, it cuts IR and also significantly reduces UV at 300 nm, but at least it should work.

A Baader U won't work. An Invisible Vision 308 nm filter should work, but I think you don't have one.

[Bernard Foot]

Stefano - yes, I'm using all my UV-Pass/IR-block glass.

It might be nice to get a 330WB80, but the Omega shop has disappeared again on ebay, and full price Omega filters are a bit expensive for me. Looks like I was lucky to get the 303BP10 during the shoirt period the Omega shop re-appeared.

[JMC]

Bernard, yes the green image you got looks good for it being predominately 303nm UV. It would be interesting to repeat the experiment of the 303nm and 315nm outdoor images you also did, and again look at the RAW files. If they are mainly green then that would suggest UV, if there is contamination from out of band light then the image should have a different colour cast (I suspect that the 315nm one will be more green than the 303nm one, however at this time of year there isn't much UV at all, especially at the shorter wavelength end).

Sorry if I missed it, but how thick is the U-340 you have?

[Bernard Foot]

Jonathan - Good idea to look at the RAW of the outdoor shot. Unfortunately I haven't retained it so I'll repeat the test. But looking at another 303 nm outdoor attempt where I do have the RAW but no 315nm comparison, it looks like the image will be brown (see below). But I'll re-do the 303-315 comparison, plus an IR shot to see what colour that gives (probably red).

Sorry, don't know the thickness of the U340 - I got it in the old days before I realised this was an important factor. I'm guessing 2-2.5 mm - it's not 1mm, and it's not 5mm!

Edited by Bernard Foot, 24 November 2020 - 09:43.

[WiSi-Testpilot]

Bernard, I am following this thread with great interest.
Here are a few of my 300 nm tests:
https://www.ultravio...__40#entry40102
Best regards,
Wilhelm

[Bernard Foot]

JMC, on 24 November 2020 - 09:12, said:

It would be interesting to repeat the experiment of the 303nm and 315nm outdoor images you also did, and again look at the RAW files. If they are mainly green then that would suggest UV, if there is contamination from out of band light then the image should have a different colour cast (I suspect that the 315nm one will be more green than the 303nm one, however at this time of year there isn't much UV at all, especially at the shorter wavelength end).

OK, here's the re-test. The brown 303nm colour is a cross between the green of the 315 nm and the orange of the BP735+R72, indicating a small amount of UV with significant contamination in the region of IR leakage for the (U340+ZWB1 sandwich). So the IR blocking is inadequate to allow an image to be formed from this very weak UV illumination.

So my conclusion is that with a UVFS lens and U340+ZWB1 blocking filters, a standard full-spectrum camera (Sony A6000) can record images at 303nm using an artificial light source such as flash or UV torch - but that daylight images (at least in winter) are beyond the reach of this set-up.

303BP10 + U340 + ZWB1:


315BP20 + U340 + ZWB1:


R72 (i.e. NIR broadband):


BP735 + R72 (IR, equivalent to a 745BP90, with T=20% at 800 nm):


BN850 (IR, equivalent to 860BP40):


LP1000 (IR, from about 950nm to end of sensor sensitivity):