MIDOPT filters

For those interested in photography “beyond the visible”, some of the filters available from Midwest Optical Technologies Inc. under the MIDOPT brand name should be very interesting. They are distributed in European countries by Stemmer AG. Both companies are specialised in the supply of machine vision equipment. What adds additional interest is that filters are supplied in very many different sizes (from M13.25 all the way to M105, mounted and unmounted, and even with mounts suitable for installation at the back of objectives with C-mount).

NIR photography

Not only long pass filters are listed in the catalogue, but also several NIR bandpass filters. Even more interesting are double and triple band pass filters. These filters on a broad-spectrum converted camera should tweak the sensitivity of the RGB Bayer-array channels in the sensors to narrow their spectral sensitivity or “move” one of the channels to the NIR. For example for B/G/NIR response the filter TB475/550/850. (The plot bellow is from the published specifications given with 10 nm steps, which results in “odd” shaped peaks.)


Also a narrow band-pass interference filter at the “red edge” of leaves’ absorption spectrum is advertised as suitable for diagnosing “plants’ health”.


UV photography

In this case the filters offered are less unusual. Mostly filters with the expected NIR leakage. On the other hand many of the VIS pass and UV and NIR blocking filters in the MIDOPT catalogue could possibly be very good for UV- and VIS-induced imaging of fluorescence in plants.

VIS photography

It is difficult to work-out from the spectra whether any of the UV/NIR cut filters would be good for improving the white balance of full-spectrum converted cameras when used in VIS light.

I have requested price quotes from Stemmer AG for the two filters whose spectra are shown above, plus for a NIR long-pass filter suitable for use with illumination from 940 nm LEDs. [Update 2018-01-23] I have received quotations some days ago, I just mention approximate prices as these may vary. Prices for 52 mm filters are in line with what one would expect for interference filters, in the range of 450 to 300 €.



Timelapse in the “dark”

Infrared (940 nm)

The night of a plant, 11 h 20 min condensed into a 10 s-long video. Only source illumination were two LEDs emitting infrared radiation at 940 nm (LED Engin LZ1-10R702). Camera: Olympus E-M1, converted to full spectrum, adapted Soligor 35 mm f:3.5 objective at f:8, Heliopan RG780 long pass filter on lens. Movement is most evident with the video at full screen resolution. Images converted to greyscale.

Ultraviolet-A (385 nm)

The night of a plant, 9 h condensed into a 8 s-long video. Only source illumination were two LEDs emitting ultraviolet-A radiation at 385 nm (LED Engin LZ1-10UA00). Camera: Olympus E-M1, converted to full spectrum, adapted Soligor 35 mm f:3.5 objective at f:8, StraightedgeU bandpass pass filter (UVROptics) on lens. Movement is most evident with the video at full screen resolution. False color, white balance set on white PTFE target.

Ultraviolet-A (365 nm)

The night of a plant, 12 h condensed into a 11 s-long video. Only source illumination were two LEDs emitting ultraviolet-A radiation at 365 nm (LED Engin LZ1-10U00), except during the last 90 min with skylight entered through a nearby window. Camera: Olympus E-M1, converted to full spectrum, adapted Soligor 35 mm f:3.5 objective at f:8, Baader U bandpass pass filter on lens. Movement is most evident with the video at full screen resolution. False color, white balance set on white PTFE target.

Plant at the end of third session. Window light, Firecrest UVIR filter.

IR images edited in Capture One, UV images edited in PhotoNinja, videos created in ImageJ (Fiji distribution).

I will describe the rig I used in a separate post. The photograph below gives a preview.

Camera “rig” as used for the three videos. Only the LEDs and filter were swapped between sessions and the objective re-focused at the start of each session. Photographs were taken on three consecutive nights, and the plant returned to a well lighted place during each daytime.

Black anodised aluminium in IR

Many cheap extension tubes, lens adapters and lens hoods are made of aluminium and have a black anodised finish: are they always better than plastic ones? As it was revealed by the macro extension tubes comparison internal reflections can be a problem in visible light and lead to glare and low contrast. These problems are caused by a relatively small amount of visible light reflected by anodised aluminium, specially if the surface is smooth. The ribbed surface frequently used avoids especular reflections, so controlling “stars” and similar artefacts but not so much the loss of contrast due diffuse reflection.

Two aluminium lens hoods which look almost identical under visible light. Both bought from eBay sellers. (The white slab on the left is a piece of PTFE used as white reference.

Black anodised aluminium looks black as it absorbs most visible light, although not all of it. At some angles of incidence some light is reflected and and the surface may appears shinny. In the near infrared region the problem is many times worse, as black anodised aluminium reflects a large fraction of the incident radiation. Lens hoods that look almost identical under normal illumination, may differ in being coated or not on the inside. The hood from RISE(UK) is not suitable for NIR imaging while the one from JJC is, and probably this second one is slightly better also for visible light photography.

The same two lens hoods look quite different when under near infrared (930 nm LED) and photographed with a digital camera converted to “full-spectrum”.

Most expensive macro extension tubes, lens adapters and lens hoods, and occasionally also some cheap ones, are painted matt black on the inside, and this is one of the reasons why they perform better. Those who build telescopes as a hobby use, as an easy to find non reflective paint, black blackboard paint. An alternative is the use of black flock or black velour. Black flock with self stick backing is available but can collect dust.

As some of the cheap aluminium-made Chinese macro extension tubes and lens hoods are of good quality except for the NIR-reflective and partly VIS-reflective inner surfaces, and as equivalent items are not always available with a better finish, I have started some tests on how to improve the optical performance in the NIR waveband of a lens hood, hoping later to also improve extension tubes and lens adapters.

My first trial was with water-based black acrylic blackboard paint “Déco additif” peinture ardoise/boad paint (Lefranc & Bourgeois, Le Mans, France). I painted a lens hood of a length not available from JJC. The result was a significant improvement, but not as good as I would have liked. I will try with a third coat, but apparently this paint is not fully matt.

The lens hoods differing only in their length, but both with the black anodised aluminium finnish. The one on the right, painted on the inside with two coats of black blackboard paint. An improvement, but far from perfect.

I end with a comparison with an original lens hood from Olympus, frequently described as grossly overpriced for a piece of plastic. Its performance is excellent both in the visible and in the near infrared.

The same lens hood on the left and an original lens hood from Olympus on the right side. They look quite different when under near infrared (930 nm LED) and photographed with a digital camera converted to “full-spectrum”

In conclusion, when possible, it is probably best to stick to original-equipment lens hoods supplied by the manufacturers of high quality objectives (caveat: I have tested only Olympus hoods). I used the painting of the lens hood as a test, as I need screw-on lens hoods with a 52 mm thread for use together with special filters of this size, as I use them with step down rings, which prevents the use of the original lens hoods.

Update: I have now realised that Heliopan makes metal lens hoods, which I expect to be of good quality. I will special-order one if I get confirmation of its finish. For the tests with adapters and macro extension tubes, I have now ordered a special paint by Tetenal, giving less than 5% reflectance (thanks Hannu for the tip!).

Conclusion: If you use lens hoods only for visible light, you should be able to approximately gauge their reflectance by eye. If you take images beyond the visible, tests are needed, specially for NIR. I have tested these same lens hoods in ultraviolet A (365 nm LED), and they all absorb UV-A radiation quite effectively.



Depth of Field (DOF)

Narrow depth of field with a tele objective of 400 mm (800 mm FF equivalent, 50-200 mm zoom objective with 2x teleconverter) at f:7.0. The grass in front of the butterfly is barely visible and the grass in the background is blurred.

As a followup to the earlier posts on taking photographs through windows and on focus stacking I will list some resources after a brief introduction. There are good explanations on the internet and in many books. There are quite a few apps that can be used in Android and Apple phones to quickly estimate the depth of field while taking photographs. However, for understanding the details, and for macro work, the Windows DOF simulator Barnack is, in my opinion, the best tool. Before listing some of these sources of information, it is important to explain that depth-of field is subjective as it is based on what on observer may consider sharp enough. Observers differ both in eyesight acuity and preferences… and we reproduce photographs at different sizes… and expect them to be looked at from different distances… and under different types of illumination… (see Ctein 2013, The Practical Side of Depth of Field). Consequently, most observers will agree that the region in focus is deeper at f/8 than at f/1, but they will in most cases disagree on where the limit between sharp enough focus and not sharp enough focus is. There is no clear transition from focus to blur, it is just a gradient and it is subjective where we consider that sharpness stops being enough.

The concept is still very useful, and allows us to control what will be sharp or blurred. Sometimes we will want to keep everything in focus, while at other times we will want to make sure that only the subject of interest is in focus and everything else out of focus. The extreme example being taking photographs through dirty windows.

Most SLR and mirrorless cameras have a depth-of-field preview button or function that closes the lens diaphragm from its normally fully open position during farming (of course it is always closed to its set value during digital image capture or film exposure). In general is it easier to visualise depth of field with electronic viewfinders which compensate for the decrease in brightness caused by a smaller aperture, thus making it easier to to see the change in depth of field or the extent of sharp and blurred regions (as long as their pixel resolution is good enough).

Warning. A lot has been written in recent years about depth of field and digital sensor sizes. Any such considerations are equally relevant to film sizes. Some sensor sizes as well as film sizes are preferable in some situations and others in other situations. Choice is always a compromise, but cameras are tools, and the best tools are always those you are an expert at using. Some of the arguments used when defending the virtues of different sensor sizes were not based on a solid foundation of optical principles. Mike Johnston and Ctein give very good explanations in posts published in The Online Photogarpher (1, 2, 3), and the program called Barnack will let you experiment, and learn how to control depth of field with whatever camera you have, as long as it allows you to set the aperture used. Whatever camera (or phone) you have, the most important thing is to learn how to use it in a mode in which you can take control.


A tutorial at Cambridge in Colour: concise and clear.

A free simulator for Windows: Barnack, do look at the help.

Photography books, usually in the chapter about lenses.