Anyone who is into photography knows that the lenses are the most expensive part in the bag. The larger the aperture or f-stop of the lens, the more light is coming in which is better for dimly lit scenes. Consequently, the price of the larger glass can burn a hole in one’s pocket. [Anthony Kouttron] decided that he could use a Rodenstock TV-Heligon lens he found online and adapt it for his micro four-third’s camera.
The lens came attached to a Fischer Imaging TV camera which was supposedly part of the Fluorotron line of systems used for X-ray imaging. We find [Anthony’s] exploration of the equipment, and discovery of previous hacks by unknown owners, to be entertaining. Even before he begins machining the parts for his own purposes, this is an epic teardown he’s published.
Since the lens was originally mounted on a brass part, [Anthony Kouttron] knew that it would be rather easy to machine the custom part to fit standardized lens adapters. He describes in detail the process for cleaning out the original mount by sanding, machining and threading it. Along the way you’ll enjoy his tips on dealing with a part that, instead of being a perfect circle on the outside, had a formidable mounting tab (which he no longer needed) protruding from one side.
The video after the break shows the result of shooting with a very shallow depth of field. For those who already have a manual lens but lack the autofocus motor, a conversion hack works like a charm as well.
Edit: This article originally stated the video demonstrates a large depth of field. This is incorrect. The video demonstrates a very shallow depth of field. The error has been corrected.
Awesome writeup, nice machining gifs
Many thanks :D
a large depth of field. For certain definitions of large that is. In this case meaning “about the thickness of a piece of paper”, and more commonly called “extremely narrow depth of field”. Unfortunately compounded by horrible spherical aberration that looks like veiling glare.
A bit more detail in the comment I left on the blog, but the TLDR is that these lenses are carefully designed to focus on an object at infinity — they are remarkably crappy when used as macro lenses (unless used in reverse). They are also intended for use with image tubes (vidicons or plumbicons), which have a substantial thickness of glass in front of the image receptor surface. The optical design expects to focus through that sheet of glass. Without one at f/1.1 it’s a large source of aberration.
Any lens that that takes cool photos, is different than any other lens in the bag (or that anyone else is using), and makes people looking at the photos say “wow” is a lens worth having. It doesn’t have to be optically perfect, or have specs that match any standard or typical definition of what a “quality” lens should be. Imperfections make it different. Flaws make it interesting. The best reaction to any photo isn’t “that looks so real”, or “that lens you used has absolutely no distortion at all”. but simply “wow.”
Wow.
It’s “Art”. I get it.
It’s easy to degrade a technically good image to produce pictures you consider “cool” by using optical or digital filters, or by otherwise munging the image acquisition process. But It’s much harder to fix photos taken by a bad or inappropriately-used lens.
Me, I’m a scientist. I’d rather get the best possible image data to start with, and if I want to turn it into “Art” then I’d prefer to choose what i want it to look like, not hope the flaws of my instrument will give me something I might like.
This is the exact issue Cooke and Zeiss are having with their new prime and anamorphic lenses. Traditionally cinema lens makers want to produce the most optically sound lenses, but now some cinematographers are not as pleased by perfectly corrected, extremely sharp lenses. They want lenses with “character” and these precision optical companies now have to spend their resources figuring out what imperfections and irregularities will create this character their audience is after, instead of correcting them out. There was one interview with Les Zellan from cooke where they were discussing these issues. I think this is why a few lens makers are now allowing rear and front element replacements, so they can continue designing amazing lenses and offer intentionally bad optics for cinematographers who want that “look.”
That’s like starting out with CDs and wanting vinyl but not knowing it.
” They want lenses with “character” ”
And then the CG guys at the effects department are tearing their hair out trying to model those lenses to get seamless integration in the scenes.
Thanks for clarifying this! I just found a document last night on the Rodenstock 3801.212, which is a 42mm f0.75 X-Ray lens. It is found here:
handle.dtic.mil/100.2/ADA277469
It literally confirms everything you just said. From the diagram on page 34, there is a thick phosphor coated glass in front of a 100mm f1.5 xr-helidon, spaced 10mm from a reversed 42.mm 0.75 Tv-heligon which is then focused onto a CCD sensor.
It is very challenging to find any technical information on these old X-ray lenses and your insight and the information the document provides helps a bunch.
Great find. That’s a pretty amazing document. A wealth of detail for anybody interested in high performance camera design. Whoever wrote that could easily get a masters degree based on that document alone.
Thanks. The depth of field is very small. I hope a correction appears soon.
Thanks! I found out that the flange focal distance of the 50mm f1.1 is 3.7mm, so now I can finally machine the adapter to the right depth to get sharp focus. It turns out the image circle of the lens is 15mm, a bit smaller than the 21.64mm image circle of micro four thirds, but at least the lens will be appropriately focusing on the sensor plane.
These lenses and similar ones designed for monochrome enlargers are corrected for small range of object and image distance from the lens plus no colour correction because the phosphor screen being photographed is single colour i.e. green. Outside these constraints they will produce fuzzy pictures. They have never meant to be general purpose camera lens. Absolutely fine article and a pleasure to read.
Thanks! I’m still looking into the differences in design and element composition between X-Ray lenses and traditional photography lenses. I’ll be posting a project update after I’ve composed more research and re-machined the mount too. I’m definitely looking into the entire medical X-ray lens application so I can get a better understanding of the design parameters these lenses were intended for. You’re input is much appreciated. Thanks!
This is probably why there’s noticeable… fog-like effect? chromatic aberration? in the film taken with this lens – the lens doesn’t work well for a large range of wavelengths. Anyway, congrats for the hack!
One could try a colored filter to get a monochrome input as it were and then shoot in B&W and see how that comes out in terms of overall sharpness.