• Ultraviolet Photography
  •  

Spectral scan of some C-mount lenses

UV Lens
12 replies to this topic

#1 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 09 September 2020 - 18:20

Revised Post to include larger screen shots of my spectra to see the axis and curves. I also provided images of the lenses.

I am posting this in case anyone was looking for a cheap lens for the HQ Pi camera. These seem to be good options for some UVA imagining:

This is an image of our favorite Wollensak 25mm F1.5 c-mount lens with AA battery for scale:
Attached Image: Wollensak_25mm_Lens.jpg

This lens will mount to HQ and m43rds and focus to infinity without any modification. The front filter thread is 25mm.

Full screen shot of a revised spectrum collected Sept. 9, for the Wollensak 25mm F1.5 C-mount lens:
Attached Image: Wollensak_Spectra_Full.jpg

Image of a Computar 8mm lens:
Attached Image: Computar_8mm_Lens.jpg

This lens will mount and fully cover the HQ sensor. It will mount to m43rds camera, but the image circle reaches the horizontal frame (Black corners), it may focus to infinity if the mount is modified, I haven't modified mine but have seen some whom have had success with specific Cmount adapters. The front filter thread is 40.5mm.

Full screen shot of Computar 8mm C-mount lens:
Attached Image: Computar_8mm_Spectra.jpg

Image of Computar 12.5mm f1.3 lens front:
Attached Image: Computar_125mm_Front.jpg

Image of Computar 12.5mm F1.3 lens side:
Attached Image: Computar_125mm_Side.jpg

This lens will fully cover the m43rds sensor and if you shave off a bit of the mount it will focus to infinity on m43rds camera with selected Cmount adapters. lots of talk about this lens online as cheap alternative to the Olympus lens. The front filter thread is 43mm.

Full Screen shot of Computar 12.5mm F1.3:
Attached Image: 12_5mm_Computar_full.jpg

Image of Cosmicar 16mm F1.4 C-mount lens:
Attached Image: Cosmicar_16mm_Lens.jpg

This lens will mount and fully cover the HQ sensor. It will mount to m43rds camera, but the image circle reaches the horizontal frame (Black corners), it may focus to infinity if the mount is modified, I haven't modified mine but have seen some whom have had success with specific Cmount adapters. The front filter thread is 40.5mm.

Full Screen shot of Cosmicar 16mm F1.4 C-mount lens:
Attached Image: 16mm_Cosmicar_Full.jpg

Image of the small Arducam 4-12mm F1.6 zoom lens which covers the full HQ Pi camera at 4mm. This will also mount to a m43rds camera but there is no front filter thread and it will not focus to infinity at 4mm. The IR means its has been corrected to focus in IR and does not have an IR block filter:
Attached Image: Arducam_4_12mm_Lens.jpg

Full screen shot of the Arducam 4-12mm F1.6 lens at the 12mm setting on the lens:
Attached Image: Arducam_4_12mm_At_12mm.jpg

Full screen shot of the Arducam 4-12mm F1.6 lens at the 4mm setting on the lens:
Attached Image: Arducam_4_12mm_At_4mm.jpg

Edited by dabateman, 10 September 2020 - 17:56.


#2 Andy Perrin

    Member

  • Members+G
  • 3,341 posts
  • Location: United States

Posted 10 September 2020 - 00:32

The text on those is a little too small to read? Even when I try to increase the size, it is pixelated and unreadable.

#3 Stefano

    Member

  • Members(+)
  • 972 posts
  • Location: Italy

Posted 10 September 2020 - 02:04

View PostAndy Perrin, on 10 September 2020 - 00:32, said:

The text on those is a little too small to read? Even when I try to increase the size, it is pixelated and unreadable.
I think it is in 50 nm increments, like 300, 350, 400 etc.

#4 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 10 September 2020 - 17:54

Ok revised with full screen shot images. Also added images of the lenses I am talking about.

Should have added that the Arducam 4-12mm will not cover the m43rds sensor. It fully covers the HQ sensor though.

My Computar 4mm F1.2 CS mount lens does not fully cover the HQ camera. But will focus to infinity and most likely has a larger field of view than expected for a 4mm lens. It has a 30.5mm front filter thread, but its UV performance wasn't great and its hard to gather all the light to get an idea of its true transmission.

#5 Andy Perrin

    Member

  • Members+G
  • 3,341 posts
  • Location: United States

Posted 10 September 2020 - 18:56

Interesting, I had the idea that the Wollensak had a better transmission than that. Also, I am surprised that some of those lenses have a pronounced color cast in the visible range.

#6 UlfW

    Ulf W

  • Members+G
  • 1,167 posts
  • Location: Sweden, Malmö

Posted 10 September 2020 - 19:44

I think that the measurement setup is done with collimators and not an integrating sphere, in the detection end, after the tested lenses.
Then measurement of the absolute transmission is not possible.

Also, if a collimator is used in the detecting end it is also difficult to avoid the dispersion affecting the coupling into the receiving optical fiber.
The focus distance at the fiber's end will be different at different wavelengths.

The results from a collimator-based setup are OK for detecting the cutoff wavelength, but not for seeing absolute transmission or relative transmission over a wide range of wavelengths.
At least that is what I have found when struggling with such methods for several years.

To do this more correctly a setup like the one Jonathan has showed some year(s?) ago and using is needed.
Then a much brighter light-source than the one I think David is using is needed, to get enough light from the integrating sphere.

Edited by UlfW, 10 September 2020 - 19:47.

Ulf Wilhelmson
Curious and trying to see the invisible.

#7 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 10 September 2020 - 19:57

Ulf,
Yes fully agree. I typically look for UV and IR cut offs not true transmission.
I have a culminating lens at the light source and at the end of these lenses to direct the light back into the fiber which leads to the spectrometer. Thus wide angle lenses give me a harder time gather back that light and why the 4mm end of the 4-12mm zoom has lower maximum than the 12mm end. I am using the DT1000 deuterium tungsten-hallogen light source.

Andy,
My Wollensak 25mm f1.5 lens may not be typically. I purchased it in mint condition. In original case with caps and original filter, which I cracked, too bad. But when I first used it, the UV transmission was near zero below 400nm.
I then took it apart, carefully polished both sides of each of the 4 elements with cerium oxide and carefully put it back together. The sharpness got much better and the UV transmission is what you see here. Its a very old lens. Others may behave differently.



#8 Andy Perrin

    Member

  • Members+G
  • 3,341 posts
  • Location: United States

Posted 10 September 2020 - 19:58

I see. So then the lenses probably don't have a color cast in visible light (or at least we can't tell from these scans if they do). But still, I thought the Wollensak's cutoff was much further out! This sample doesn't even make it to 350nm, when I thought it was comparable to the EL-Nikkor 80mm/5.6 metal.

ETA: got it, David.

Edited by Andy Perrin, 10 September 2020 - 19:58.


#9 OlDoinyo

    Member

  • Members(+)
  • 772 posts
  • Location: North Carolina

Posted 12 September 2020 - 15:42

Is the apparent secondary peak below 250 nm in these graphs just an artifact of the instrument?

#10 UlfW

    Ulf W

  • Members+G
  • 1,167 posts
  • Location: Sweden, Malmö

Posted 12 September 2020 - 17:34

View PostOlDoinyo, on 12 September 2020 - 15:42, said:

Is the apparent secondary peak below 250 nm in these graphs just an artifact of the instrument?
Yes it is.
It is a combination of lower intensity of the light source and lower sensitivity of the detector...
These measurement results are calculated by dividing a raw measurement of the lens with one without the lens, seeing the light source only.
When both become very small you are almost doing a 0/0 calculation resulting in a very noisy result that with some averaging can give very odd results.

Edited by UlfW, 13 September 2020 - 05:58.

Ulf Wilhelmson
Curious and trying to see the invisible.

#11 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 12 September 2020 - 22:51

View PostOlDoinyo, on 12 September 2020 - 15:42, said:

Is the apparent secondary peak below 250 nm in these graphs just an artifact of the instrument?

Ulf gave an excellent explanation. Thus why I like to look at raw data for low wavelengths signal. My detector can see the 185nm mercury line. Funny my IR detector isn't as sensitive and goes crazy above 940nm. But it can see into 1050nm. I have a bandpass filter there and its just observable.

#12 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,027 posts
  • Location: USA

Posted 14 September 2020 - 17:21

David, perhaps you could add a comment above each chart saying something about the artifacts and that the chart is best used for cutoffs only??

Of the lenses you tested, it appears that only the Wollensak is worth considering. The others seem to have such a low transmission rate.
Andrea G. Blum
Often found hanging out with flowers & bees.

#13 dabateman

    Da Bateman

  • Members+G
  • 2,042 posts
  • Location: Maryland

Posted 15 September 2020 - 05:02

View PostAndrea B., on 14 September 2020 - 17:21, said:

David, perhaps you could add a comment above each chart saying something about the artifacts and that the chart is best used for cutoffs only??

Of the lenses you tested, it appears that only the Wollensak is worth considering. The others seem to have such a low transmission rate.

I hope this post doesn't creat more harm than good.

Sorry Andrea, I am not using a cosine correction or integrated sphere to collect the light from the lens.
I am just using a culminating lens. So I am loosing lots of light. Especially for wide angle and tight high magnification objectives. The Wollensak 25mm is closer to my ideal focal length. So I am just not collecting all the light out from the wide angle lenses.

Looking roughly at 425nm you can approximate where 100% may be for each lens. Then use that to guess roughly the 50% transmission value in the UV. Also the UV cut off value will be true. So thats a real number.

Actually, all of these are quite nice for UVA photography with the pi HQ camera. The other advantage of the pi HQ, is you can place up to 4mm of glass in the module. So the UV filter is behind the 4-12mm zoom lens, which is good as it doesn't and can't have a front filter. In 4mm setting, the lens pushes forward, preventing a filter.

Edited by dabateman, 15 September 2020 - 05:04.