Under <300 nm?

[JCDowdy]

All these speculations about sensors aside, is there a readily-available filter anyone owns or has heard of in a reasonable diameter which would have the requisite bandpass/block? If the answer is no, then the whole discussion is indeed moot, at least for us ordinary mortals.

 

Such filters exist, most will be smaller diameter and certain to be dichroics.

This one might be worth a look if you seriously want to try this.........

[Andrea B.]

Here is an Ornithogalum umbellatum in UV using the Baader-U.

1/2 second exposure at ISO-200.

 

Now using a tiny below 300nm filter. Sorry for the horrible light leak, but at least you can see that there is a bit of life down there below 300nm. I did not have a good way to mount the little filter at the time.

8 second exposure at ISO-1000.

ADDED LATER:

Filter is 293BP10 about 15mm in diameter. So it is about 283-303nm more or less in half-bandwidth. Transmission not even reaching 15%. An impossible filter. I said "below 300nm", but really not by very much! Here is a link to the transmission chart, although the linked filter is bigger than the one I bought: http://www.ebay.com/...g-/310495177256

 

[OlDoinyo]

Thank you--I may file that datum away for future consideration. I have encountered this alleged film sensitivity data online:

 

http://photo.net/learn/optics/edscott/uv000020.htm

 

I note that the UV portion of these response curves is said to be "speculated," and I do not know how seriously to take it. Panchromatic films incorporate sensitizers optimized for visible wavelengths, and in my experience, sensitivity can fall off in the UV. Through a Baader filter, the sensitivity difference between Shanghai GP3 (nominal ISO 100) and Ilford Delta 3200 (nominal ISO 3200) is only about a stop and a half, not five stops, indicating to me that we are perhaps operating on the tail of the sensitizers' excitation curves. Also, at the very shortest wavelengths the emulsion itself becomes increasingly opaque, resulting in a "skin effect" that exacts a penalty in speed. But I would still say that there is a chance of some sensitivity in the UV-C band, and if I were making an attempt at it, that is the way I would go, rather than mucking about with de-bayered digital sensors, which is a more complex and troublesome-sounding approach. Subject matter for a proof of concept would also require a bit of care; perhaps one of the hand-held 254-nm mercury lamps could serve as a light source for the studio. Outdoors, one wonders about lightning, which probably throws off loads of UVC, if you can catch it; my hypothesis would be that the lightning channels would look narrower than they do in ordinary photos, because halation is efficiently suppressed.

 

Edit--Interesting photo, Andrea; you posted just before me and I did not see your post until mine cleared. I can see that focusing was a bear with whatever filter you used!

[Alex H]

Thank you--I may file that datum away for future consideration. I have encountered this alleged film sensitivity data online:

 

http://photo.net/lea...tt/uv000020.htm

 

I note that the UV portion of these response curves is said to be "speculated," and I do not know how seriously to take it.

 

I tried to find anything about it in Kodak technical data, but the sensitivity curve stops at 400nm in all graphs I have seen. How do they know it can reach so deep into UV? Delta 3200, on the other hand, has full sensitivity Data published, including the UV region, which is great.

 

As John pointed out, there are a number of filters available on eBay that transmit in 200-300nm range.

[Andrea B.]

The problem with that tiny little UV-pass filter was that its diameter was way too small and I was attempting to hand-hold it in place just to get some idea of whether its transmission could be recorded. Let's just say this method is not recommended. :) :) :D I had focused through the Baader-U and really didn't attempt to compensate when I switched to the little filter. Michael is looking up the label on that filter for me as I'm away from home currently.

 

ADDED LATER:

Filter is 293BP10 about 15mm in diameter. So it is about 283-303nm more or less in half-bandwidth. Transmission not even reaching 15%. An impossible filter. I said "below 300nm", but really not by very much! Here is a link to the transmission chart, although the linked filter is bigger than the one I bought: http://www.ebay.com/...g-/310495177256

 

 

I have been saying for about a year that it was time for me to get some narrow UV-bandpass filters "in house" to experiment with. Maybe it is time to make good on that! I've never been able to decide on a vendor.

[Andrea B.]

Andrea = Voice of reason. Once again.

 

Well, that's why they pay me the Big Bucks.

 

 

 

 

[Shane]

I have some unpublished data (not collected by myself, so I prefer not to release it without permission) showing a very modest response of a consumer-level CMOS Bayer sensor (equipped with standard window) well below 300 nm, but I would not call it usable for imaging. The practical limit for consumer-level Bayer sensors seems indeed to be around 300 nm.

 

Enrico - What is your definition of "practical"?

[enricosavazzi]

In this particular case, a 60 s exposure still produces an underexposed and noisy picture at wavelengths below 300 nm. Compared with a 4 s exposure sufficient to expose properly at 365 nm. The UV source is a commercial lab equipment monochromator based on a deuterium lamp, and is supposed to produce an approximately constant illumination intensity in this range of wavelengths.

[baffe]

Enrico I don't understand that. Did you (or your source of info) use a monochromator instead of a filter on the cam or in addition to it?

 

Excuse my stupid question please.

 

But I have never seen a deuterium lamp or a monochromator and know their function only from descriptions on the web.

[OlDoinyo]

That kind of monochromator is usually mounted on a broadband light source, I think, to provide selectable narrow-band illumination. They look like a box with a slit on each side and some sort of vernier.

[enricosavazzi]

In this particular test, the output of the monochromator was not used to illuminate a subject. Instead, it was used to directly illuminate the camera sensor. This eliminates the possible bias introduced by a test subject selectively absorbing/reflecting radiation at some wavelengths, or fluorescing. No filter was placed along the optical path between source and sensor, also to eliminate any bias. In this way, it is possible to compare the relative response of the sensor at different wavelengths (assuming that the output level and width of the emission peak of the monochromator remain reasonably constant at all used wavelengths).

[baffe]

Thank you enrico now I understand.

 

Can I presuppose that any optical lens that will allow a certain wavelength in the Uv range from 200nm to 300nm to pass is able to project a picture with light of this wavelength? (Independent the lenses focal length being different for different wavelength)

 

On the other end there are many lenses well transparent for ir and not forming good pictures.

 

Going to the extreme: Gamma will also pass an optical lens but not lead to a focussed picture behind it.

[nfoto]

The UV-Nikkor 105 is specified to have a response down to 200 nm.Or 220 nm depending on the source. So it will definitively form an image within the 200-300 nm band.

[baffe]

Ok, so 200nm still show a behaviour "like light" and I'm happy to handle 365nm :-)