• Ultraviolet Photography
  •  

Am I getting an image at 303nm?

32 replies to this topic

#21 Bernard Foot

    Bernard Foot

  • Members+G
  • 494 posts
  • Location: UK

Posted 24 November 2020 - 14:48

View PostWiSi-Testpilot, on 24 November 2020 - 10:02, said:

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

Wilhelm - looks like you were using an A6000 as well. I am surprised you got so much image at 300 nm, and very surprised that you got anything through the double glazing - even at 315nm I find that double glazing largely kills the image. I would guess that you have a significant amount of IR leakage - it would be worth doing the RAW-colour test that Jonathan suggested and which I have now done and posted my results.
Bernard Foot

#22 Stefano

    Member

  • Members(+)
  • 1,435 posts
  • Location: Italy

Posted 24 November 2020 - 16:07

Bernard, nice test. Some thoughts:

-You may try the classic glass of water test, to see if it is dark.

-Your LP1000 image appears a bit foggy. Do you have any idea why?

-Your images are quite sharp considering you used a single lens. You stopped it down considerably, and this usually makes them sharper, but still you got some nice edges. Does it have a relatively long focal length? That helps too.

-Also, how did you mount it on your camera? How do you focus it? Do you have something like a tube with the lens in and you slide the lens back and forth to focus?

#23 WiSi-Testpilot

    Member

  • Members(+)
  • 98 posts
  • Location: Germany

Posted 24 November 2020 - 17:38

View PostBernard Foot, on 24 November 2020 - 14:48, said:

Wilhelm - looks like you were using an A6000 as well. I am surprised you got so much image at 300 nm, and very surprised that you got anything through the double glazing - even at 315nm I find that double glazing largely kills the image. I would guess that you have a significant amount of IR leakage - it would be worth doing the RAW-colour test that Jonathan suggested and which I have now done and posted my results.

Bernard, the 300 / 313 nm images where made with the GSENSE2020BSI sensor, not with the A6000.
I agree with you. There is only Vis and NIR in the 300 nm image. In the 313 nm image is also a lot of leakage. The sensor GSENSE2020BSI is very sensitive in the visual and NIR range and the filters block only 4 OD. The setup is not suitable for pure UV images at 300 and 313 nm in daylight.

Edited by WiSi-Testpilot, 24 November 2020 - 17:41.


#24 Stefano

    Member

  • Members(+)
  • 1,435 posts
  • Location: Italy

Posted 24 November 2020 - 17:48

I would use at least OD 6, maybe 7 to stay safe.

I think the biggest limit is money. These filters are not cheap. The GSENSE2020BSI sensor is one of the best you can get, once you have a quartz/fused silica lens you are done with the lens, and the only remaining thing to improve (apart from your light source) is the filter(s). 303 nm is really at the edge of what sunlight can provide, that's the realistic limit, any lower and it simply becomes impossible or at least extremely difficult to get an image.

Edited by Stefano, 24 November 2020 - 18:18.


#25 Bernard Foot

    Bernard Foot

  • Members+G
  • 494 posts
  • Location: UK

Posted 24 November 2020 - 22:37

Stefano - the LP1000 image appears foggy only because the contrast is low - partly because all my IR shots have low contrast if they are taken straight from the camera, and when I process the RAW with No Profile then the contrast is also lower. This is what it looks like when I process it with Nik Silver Effex Pro:

Attached Image: LP1000 Nik LoRes.jpg

The UVFS Plano-Convex image quality is OK in the centre, but definitely drops of badly as you move away from the centre, as you can see in the above image. This was taken at about f/11. It is a 75mm focal length lens - longer focal length would probably give better quality, but the field of view would be too narrow on an APS-C sensor.

Here is an image of a Lily made at 303 nm, using flash. It's actually pretty good because the corners are out of focus anyway - and in fact the curvature of field may have helped with the concave form of the flower.

Attached Image: Lily 303.jpg

The lens diameter is 25mm. I glued it using epoxy resin to a 25-37mm step-up ring and then added a 37-49mm step up ring to provide a 49mm filter mount at the front. At the back I glued a 28-37mm step-up ring and then added a 37-49mm step up ring to give a 49mm filter mount at the back. Then I added a 49mm-M42 macro reversing ring, which gives me a 42mm lens thread at the back, which screws into a focussing helicoid. The apertures are made out of 0.5mm aluminium sheet and fit between the rear 49mm filter thread and the 49mm-M42 reversing ring.

Edited by Bernard Foot, 24 November 2020 - 22:45.

Bernard Foot

#26 Stefano

    Member

  • Members(+)
  • 1,435 posts
  • Location: Italy

Posted 25 November 2020 - 03:05

Thanks Bernard. I like your Lily at 303 nm, really shows the strong absorption at these low wavelengths. These single element lenses are quite good if used properly, they will never be comparable to an actual UV lens of course, but if you don't want to spend thousands they do the job well.

Edited by Stefano, 25 November 2020 - 04:37.


#27 Bernard Foot

    Bernard Foot

  • Members+G
  • 494 posts
  • Location: UK

Posted 05 December 2020 - 11:20

I've tried again to get a daylight image using the 303BP10 filter, this time using a second U340 filter to add more IR blocking. This time did get an image - of sorts! Results are below. I'm confident this is a UV image - adding an R72 to the filter stack or simply closing the window killed the image.

Lighting was UK mid-winter sunlight at about 10.30 a.m.

So you can get an image at 303nm with a modified A6000 camera and a UVFS lens - but image quality is like Fox Talbot's images in Lacock Abbey. A high sun in the summer would of course reduce the exposure time, and the reduced UV scattering might might provide some more detail in the image. But it probably won't allow for prize-winning landscapes.

First, here's a comparison image through a 315BP20 filter + single U340. ISO 400, 30 secs, ISO 400:

Attached Image: DSC00033 LoRes f11 ISO400 30 sec 315BP20  U340.jpg

Here's the best I could get through the 303BP10, with 2 x U340 (2mm I think) and a ZWB1 (again 2mm I think). Focus and framing are out as there was no liveview image to work with. Image straight out of the camera. f/11, 30 secs., ISO 3200:

Attached Image: DSC00042 LoRes f11 ISO3200 30 sec 2xU340 ZWB1.jpg

And here's the image with some post-processing:

Attached Image: DSC00042 HiCon LoRes.jpg
Bernard Foot

#28 colinbm

    Member

  • Members+G
  • 2,117 posts
  • Location: Australia

Posted 05 December 2020 - 11:35

I do like this one Bernard..."First, here's a comparison image through a 315BP20 filter + single U340. ISO 400, 30 secs, ISO 400 "

#29 Stefano

    Member

  • Members(+)
  • 1,435 posts
  • Location: Italy

Posted 05 December 2020 - 13:52

Bernard, I think you can't get much more than this. 302 nm under sunlight is really the limit.

#30 Bernard Foot

    Bernard Foot

  • Members+G
  • 494 posts
  • Location: UK

Posted 05 December 2020 - 14:31

View PostStefano, on 05 December 2020 - 13:52, said:

Bernard, I think you can't get much more than this. 302 nm under sunlight is really the limit.

I'm sure you're right. Of course, with a 303BP10 filter there is still a reasonable amount of transmission up to about 308nm, so this will have been contributing to the image.

But with flash I could get good images with the 303BP10, so this answered the original question, triggered by a comment from Cadmium: can the sensor on an off-the-shelf consumer camera, adapted to full spectrum, reach down to 300nm, with the limiting factor normally being the lens?.
Bernard Foot

#31 JMC

    Member

  • Members+G
  • 1,201 posts
  • Location: London, UK

Posted 05 December 2020 - 15:56

Bernard, if you've not seen it here's some of my work on looking at sensitivity that far down - https://www.ultravio...20%20monochrome

On my Canon the sensor coverglass itself was blocking all the short wavelength UVB even when the filter stack was removed. Removing that resulted in it being bale to see down to 308nm with my filter setup (and I suspect even lower if I had suitable filters and light sources).
Jonathan M. Crowther

http://jmcscientificconsulting.com

#32 Bernard Foot

    Bernard Foot

  • Members+G
  • 494 posts
  • Location: UK

Posted 05 December 2020 - 22:29

Jonathan - if I understand that link correctly, with your multispectral 5D SR (which I assume means a standard camera with the internal UV/IR cut filter removed) you're effectively getting nothing at 308 nm.

Then you say you could get lower (at least to 308nm, perhaps deeper) by removing the cover glass. Not sure what you mean by the "cover glass" - a clear glass to replace the UV/IR cut filter, or the sensor's factory-sealed coverglass?

Alan Burch, who did my full-spectrum conversion, does not replace the UV/IR cut filter with anything. So if by "cover glass" you mean a UV/IR cut replacement, the full-spectrum Sony A6000 would be equivalent to your Canon with the cover glass removed and we're getting compatible results.

On the other hand, if by "Cover Glass" you mean the factory-sealed coverglass, then we're getting different results. The Canon needs to have this coverglass removed to get down to the low 300's whereas the Sony A6000 is getting down to 303-ish with that coverglass still in place.

Edited by Bernard Foot, 05 December 2020 - 22:30.

Bernard Foot

#33 JMC

    Member

  • Members+G
  • 1,201 posts
  • Location: London, UK

Posted 05 December 2020 - 23:19

Bernard, the cover glass I'm referring to is the one that's directly attached to the front of the sensor.

On some cameras this seems to be glass, on others something else. On my Eos 5dsr it started blocking UV just below 400nm, and by 300nm was essentially opaque. It was only when I had this replaced with a different cover glass (Schott WG280, and I've also had quartz on a couple of cameras) that I was able to see anything down that far.
Jonathan M. Crowther

http://jmcscientificconsulting.com