On the Influence of Wavelength used to Depth of Focus

I have been wondering about the fact what happens to the depth of focus (DOF) in multispectral imaging, i.e. what happens if you leave all the same except the wavelength of the light used. The known formulae indicate that if used UV instead of visible light the DOF should get smaller, but there have been experts claiming the opposite. So lets see if real tests confirm the theory or not!.

In all cases the UV Rodagon 60mm served as taking lens, focus was not adjusted between shots.
A Baader UV/IR cut filter was used for the visible light shots, a baader U-filter for the UV shots and a B+W 092 IR filter for the IR shots. Light sources were a tungsten lamp and my Nichia UV LED lamp (365nm).

A precise Carl Zeiss glass ruler served as a target. Distance of focused point to CCD/Chip was 320mm, height of the camera above the target plane was 245mm i.e. the shot was done at an angle of ca 50 degrees.

I tried to keep the exposure constant for the VIS, UV and IR shots while I varied the aperture in the sequence 5.6/8/11/16. Aperture setting was used as a parameter between the shot series.

[as usual, a click on an image opens up a larger view]

1) Visual shot series @f5.6/8/11/16 using Baader UV/IR cut filter [directly from the camera, no modifications except cropped]:





2) UV shot series @f5.6/8/11/16 using Baader U-Filter [directly from the camera, no modifications except cropped]:





3) IR shot series @f5.6/8/11/16 using B+W 092 filter [directly from the camera, no modifications except cropped]:





4) VIS, UV and IR @f5.6 compared:





5) VIS, UV and IR @f8 compared:





6) VIS, UV and IR @f11 compared:





7) VIS, UV and IR @f16 compared:





8) IR focus correction @f5.6 by moving back 1mm camera/lens:





So what did we learn here:
a) using UV light instead of visible light indeed leads to a narrower DOF as compared to VIS
b) using IR light instead of visible light indeed leads to a larger DOF as compared to VIS
c) the lens used shows a neglectable focus shift (i.e. no) for UV
d) there is some IR focus shift of about 1mm or 0.3% of the distance which can be easily compensated by moving the camera/lens back by 1mm (the lens, however, was never designed for being used at IR)
e) the lens shows some contrast degradation when used for IR (the lens, however, was never designed for being used at IR)

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