For an experiment I will hopefully be soon conducting at Johns Hopkins I need two different lenses.

The first needs to allow all wavelengths above 500 nm to pass (thus a high pass filter) and cut off everything else.
The second needs to allow all wavelengths below 370 nm to pass (thus a low pass filter) and cut off everything else.

My knowledge of optics is middling. I know that good old glass cuts of UV light, but I was hoping for something more specific. Does anyone know of the theory necessary to "tune" materials to make such filters?

Truth be told, I'm an experimentalist, so simply giving me a retail source that has such lenses would get me to where I need to go! But learning the theory would be nice as well.

Thanks, Sam

  • $\begingroup$ Hmmmm. That is an interesting question. I'm curious to see the final answer. $\endgroup$
    – user346
    Dec 31, 2010 at 3:45
  • 1
    $\begingroup$ Edge filters - sspectra.com/ed.html $\endgroup$
    – skywaddler
    Dec 31, 2010 at 4:50
  • $\begingroup$ @skywaddler: consider making that an answer. $\endgroup$
    – David Z
    Dec 31, 2010 at 7:35
  • $\begingroup$ Can you tell us more about your application and requierments? $\endgroup$
    – nibot
    Jan 3, 2011 at 21:15

4 Answers 4


There are many ways to do this. Which option you choose depends on what degree of performance you require, and how much money you re willing to spend.

First of all, you should understand that, while you could apply a wavelength selective coating to a lens, this would much more commonly and cheaply be done with a wavelength filter separate from the lens itself.

Now, the cheapest way to do this would be with a glass filter which absorbs short or long wavelengths selectively. There is a wide range of these available from Schott Glass. I use these commonly, and I've never had trouble. Most vendors will be happy to produce custom shapes, thicknesses, etc.

The down side to filter glass is that it doesn't have a particularly sharp cutoff between the bassband and the stopband. For that, you will need a dielectric coating engineered to your specifications. For that, I would look at CVI or Newport, although there are other vendors out there. There may be something you can use in their catalogs, but custom orders are normal for the optical manufacturing industry, so don't hesitate to call up a sales engineer. In my experience, sales people in this industry are very well educated on their products, or will direct you to an engineer who can tell you exactly what they can produce for you.

Again, there are other vendors you could look at, but these are the ones I would go to first. At the very least, looking at their catalogs will give you an idea of what it is you are really looking for.

  • $\begingroup$ Thank you so much! This is exactly what I am looking for! As I branch out doing my own experiments, I find it amazing the wide variety of subjects I need to have good knowledge in that have nothing to do with what I am actually running my experiment on (in this case, tensiometry). Thanks again! $\endgroup$
    – pballjew
    Jan 2, 2011 at 2:08
  • $\begingroup$ k: I thought I did so by pushing the up button. Sorry about that. I've checked it as appropriate $\endgroup$
    – pballjew
    Jan 31, 2011 at 0:23
  • $\begingroup$ You generally need to specify, how transparent you want the 'on' part, how well blocked is the 'off' part and how sharp the transition. Having tight tolerances on more than one of these parameters gets expensive! $\endgroup$ Mar 31, 2011 at 17:39

You may consider to use a prism to separate the different frequency light (essentially you are performing a Fourier transform). Then physically blocked those light with wavelength higher than 500 nm and lower than 370 nm. Finally, you merge these light ray together. This will be the device you want and you can freely adjust the frequency range.

  • $\begingroup$ The phrase "knife edge collimator" floats through my mind when you describe this tool. $\endgroup$ Dec 31, 2010 at 4:15
  • $\begingroup$ doh. Its that simple! However, I think in practice it might be rather more difficult to implement than it sounds. Experimental optics is a very tricky subject. $\endgroup$
    – user346
    Dec 31, 2010 at 4:16
  • $\begingroup$ A diffraction grating might also do the job. $\endgroup$
    – user346
    Dec 31, 2010 at 4:18
  • $\begingroup$ True, although with a diffraction grating you get a pattern with multiple bands of each wavelength, each of a slightly lower intensity. I'd try the prism before the grating. $\endgroup$
    – David Z
    Dec 31, 2010 at 7:36
  • $\begingroup$ @space_cadet: A reason of using prism rather than grating is because it should be easier to merge the light (inverse Fourier transform) using another prism. A system consists of prism-lens-prism should do the trick elegantly, except the slightly impefection of lens. $\endgroup$
    – unsym
    Dec 31, 2010 at 8:12

What you're looking for isn't a lens so much as a filter. There are filters for whatever purpose you might want out there in the world, and they are relatively inexpensive. Doing a google search on Optics Filters, I came across a couple of sites that might help you.


If you need extremely sharp filters because your wavelengths in question are either close together or need to be sepearated by a high degree, look at filters from http://www.semrock.com (no affiliation).

Other than that a spectral device (grating or prism based) combined with knife edges as described by hwlau is best, but as he said, you will need a 2nd such device to recombine your desired wavelengths into a single beam and there will be losses (non-brewster reflections on the prism, higher diffraction orders in the grating case, optical surface quality limitations) in the process.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.