Introduction
It is frequently said that lens adaptors and extension tubes devoid of “glass” do not deteriorate image quality. It is also rather frequently said that there is no reason to buy expensive adaptors and extension tubes instead of very cheap ones from eBay, AliExpress or similar sites.
In three separate pages I consider three different deficiencies that affect the performance of macro extension tubes: 1) Alignment and stability, 2) lens-to-sensor distance and 3) glare from internal reflections.
Below I describes tests about image degradation caused by glare and other problems caused by reflections from the inner surfaces of optical “tubes” such as lens adaptors and macro extension tubes.
The test results described here are not for current production items. Currently produced items of the same brands may show improved or degraded performance compared to those I tested a five years ago.
It is also important to be aware that “black coatings” in optical devices are not necessarily black, non-reflective, outside the visible region of the spectrum as I have shown in the page titled Black-anodized aluminium reflects NIR. Thus, the results presented are valid for visible light only.
How reflections are controlled
The cheapest optical devices have a smooth, sometimes even shiny, black plastic inner surfaces. Black anodised aluminium with a smooth surface is also common in cheap devices. These surfaces are prone to cause specular reflections, visible as bright patterns or shapes in photographs or ghosting. Ribbed surfaces, reduce specular reflections and ghosting, but increase diffuse glare if their surface is not matt black.
Special matt black paint that absorbs about 98% of light can be very effective, and even more effective when combined with a ribbed surface. Most effective is a good black flocking, however, flocking if not of the best quality can give out some loose fluff that is prone to cause problems inside cameras. In addition flocking tends to attract dust particles.
Three extension tubes
I tested in 2017 both then newly bought and earlier acquired automatic extension tubes. The three sets of tubes consisted in two rings of 16 mm and 10 mm thick. They have contacts that connect the camera to the lens allowing automation, even though the tube are passive.
The most elaborate packaging was that of the COMIX tube set. The outer finish of these tubes is also refined. The Kenko packing was functional and simple, the tubes at first sight not very different to the cheaper ones. The PIXCO ones came is the simplest packaging, that looked cheap, matching the low price.
The interior surface of the inner walls look to the eye quite different. The surface of the Kenko and PIXCO tubes is ribbed, non-reflective in the Kenko tubes and only slightly reflective in the PIXCO ones. The surface of the Commlite/COMIX tubes is smooth and reflective.
The Kenko set costs in 2024 about 120 €, the PIXCO set costs about 40 € and the Commlite/COMIX costs about 50 €. The photographs in current advertising suggest that the design remains the same.
Tests
I tested in 2017 both then newly bought and earlier acquired devices. I expected some differences, but not as large as I found. Some of the test conditions I used were rather extreme, but they dramatically showed that although extension tubes and most lens adapters are just hollow tubes, if poorly designed or manufactured they can drastically deteriorate image quality.
Not all cheap macro extension tube sets and adaptors are extremely bad, some are not too bad taking into consideration how little one pays for them, and may be good enough for occasional use. The tests also give some idea of what to look for in advertising photographs when choosing an item to buy.
I ran the tests indoors in a darkened room, with the camera on a tripod, mounted on a focusing rail. The light sources, two white LEDs, were firmly attached to the same rail by means of two “magic arms and a magic ball”. The camera shutter was triggered remotely. I used an Olympus E-M1 digital mirror-less camera with an M.Zuiko 45 mm f/1.8 prime lens considered to be one of the best available for Micro Four Thirds. I used as light source warm white LEDs (NICHIA type NS6L183AT-H1, with CRI = 92), driven with a constant current of 700 mA. The test target was an SD card adaptor on a matt grey card with 18% reflectance (Novoflex maxi Zebra card, A4) used to minimize reflections from the table surface. I converted the test images to grey-scale (“black and white”).
For the glare/internal reflections test I positioned one LED pointing downwards towards the target and a second identical LED pointing sideways and upwards towards the rim of lens at an angle of approximately 45 degrees but in such a way that it was just outside the image frame, i.e., the LED was not visible through the camera viewfinder. First a close up of this set up.
I tested several different diaphragm settings, I show here first those obtained at f/1.8 and below those obtained at f/8.0.
At f/1.8 the Kenko extension tubes together with the M.Zuiko lens suffer from mild glare. This is a very good performance given the set-up used in the test. Contrast is decreased but the image usable. The very cheap Pixco tubes, are not as good as the Kenko ones but probably fine in many normal use cases. The mid-priced COMIX/Commlite are a bad joke (pun intended).
The differences at f/8 need some explanation: even though at f/8 the glare is better controlled with all tubes, exposure is wrong for some of them. I think what is going on is that the camera is measuring EV with the open diaphragm, but reciprocity between f-value and shutter speed fails as the glare gets better controlled at f/8. So, the camera auto exposure system gets mislead by the strong reflections!
Note: In the photos the SD card adapter accidentally moved. This does not affect the test as the LEDs were attached to the camera and the SD card adapter on a table. Although I did not take another set of images, I moved the LED away from the objective and then returned it to a similar position and checked that the difference between the Kenko and COMIX extension tubes was the same as during the first round. Also during the first round I tested the Kenko tubes twice to make sure that the LED had not moved. There is an out-of-focus fin from the LED heat sink in the image top right region.
CONCLUSION: Extension tubes are hollow tubes, but they can still differ dramatically in how much or little they degrade image quality in actual use! Reflections can also disturb auto-exposure and auto-focus performance of a camera. Take this into account when buying optical equipment, both with and without glass in them.
END-OF-THE-STORY: I never used the COMIX/Commlite macro extension tubes after buying them. I have used the Kenko ones regularly in the last five years. I did not keep the COMIX ones, but I have kept the Pixco ones as a backup. In summary the Kenko tubes are very good and expensive, the PIXCO ones even if costing 1/3 of the Kenko are functional as cost-cutting has affected aesthetics and packaging rather drastically but function only slightly. The Commlite/COMIX look good are very nicely package, but do not do their job. In this case cost-cutting drastically hinders function while retaining the appearance of high quality.