Monday, March 2, 2009

Multispectral UV+VIS+IR in just one shot?

After some days of mumbling and grumbling and finally designing some special filter, I got that and now had the chance to make some initial tests, to see if the idea works.

Maybe you remember some time ago I published my remapping method to compress two images (UV, VIS) or even three (UV, VIS, IR) into a multispectral one. Quite a cumbersome process esp. making sure that all the images match while screwing in and out filters etc.

The steps looked like this using an off the shelf D70, UV Nikkor 105mm, UVIR cut filter for the VIS shot, Baader U-Filter for the UV shot (these are the old pictures):

[click on image to see a larger version]

Visual (VIS) shot:


UV shot using Baader-U filter:


and finally the composite UV+VIS remapped into one image:


in detail:



Now the result of the one-shot-process, using a full spectrum mod. D70 (clear quartz glass window >270nm), Nikon UV Nikkor 105mm plus my "secret" XUVIR filter (300-1000nm) to allow UV+VIS+IR in just one shot:


and in detail:


For comparison the identical shot using a special compensating filter, which makes the modified D70 a "normal" one, just VIS (i.e. an external filter which has the same filter characteristics as the removed internal filter had):


Be careful comparing the images, since the remapped one seems to be much sharper. Due to that embossing effect when overlaying two images it seems to have more structure, but this is not really existing. Another reason for the lack of sharpness is simply the fact that the UV Nikkor is focus corrected for UV+VIS, but not for IR.Important however is the fact, that the tiles reflect IR very strongly, which is clearly shown in red and that the tin roof above that window very strongly reflects UV, which clearly shows as deep blue. 

Another comparison, a street scene, shot some time ago in individual shots and remapped:


and about the same one, now using the new XUVIR filter and just ONE shot:

And both is accomplished now in just one shot, no remapping, no filter fumbling.

If you need such a XUVIR filter, drop me a note, I had made a few more....

[Remark: This technique shown only works well, if the reflection pattern of the subject is quite different in UV and IR, otherwise there won't be a clear separation]


Stay tuned, more will follow on that fascinating subject...

More info on this very interesting field may be found on my site http://www.pbase.com/kds315/uv_photos

Is there "life" below 350nm?

A lot has been discussed a forums dealing with UV photography what the low cut off UV wavelength might be of modern DSLR cameras i.e. how deep into the UV one could reach. A common assumption was around 350nm.

Well for that purpose I found a special steep astronomy filter which transmits only the 300-350nm band. A modified Nikon D70 camera was used, which I had modified using a clear quartz glass filter built in, which itself transmits flat from 270nm onwards. A Nikon UV-Nikkor f4.5 105mm was used as taking lens, since that transmits flat from 200nm onwards.

[click on image to see a larger version]

This now is the transmission spectra of that BP350 filter taken using a Ocean Optics USB2000 spectrometer (185-850nm; ignore that ripple on the right, measurement inaccuracy only, caused by the used Deutrium lamp):



and here in detail the spectra if stacked with a Baader U-filter (310-390nm) to insure NIR >700nm is completely suppressed, since that BP350 filter starts to leak some NIR above 900nm. The resulting transmission peaks at about 333nm and stretches from 320-350nm.



Now to the really interesting part, will the mod. D70 be able to record some useful information below 350nm? The camera was set to ISO1600 and the working aperture was f8. As a UV light source I used one shot of my modified High Power UV flash set at 400Ws, with UV front filter. Image came directly from the camera, only resized, and obviously only the RED channel records some useful information in this waveband.

So yes, there is life below 350nm, actually to about 325...330nm I would say.



And just for comparison, this is the result if only the Baader U-filter was used at the same f-stop:



It is obvious that using the BP350 and the 320-350nm band causes some 2...3 stops less sensitivity, but shows different detail of that Phalaenopsis Hybride Orchid.

Stay tuned, more will follow on that fascinating subject...

More info on this very interesting field may be found on my site http://www.pbase.com/kds315/uv_photos