De-Yellowing vintage lenses

ektar

Mu-43 Top Veteran
Joined
Jun 8, 2017
Messages
588
Location
Western NC
Real Name
Don
Paraphrasing someone from the distant past, "all I know is what I read (online)," but what I read is that there is a group of vintage lenses that are prone to having an element or elements that turn yellow. Not the yellowish surface appearance from the antireflective coatings of the day, but the glass itself. It is my understanding that the culprit in this yellowing glass that was formulated to include Thorium, a radioactive element. This makes the material unstable, it yellows. (Now, you'll also read of "rare earth element" glass...my father was the chemist, not me, so I don't know that all rare earth elements are radioactive, but Thorium is.) There are various (and variously consistent) sources as to what lenses are blessed with "Thoriated" elements, but the range includes Kodak Instamatics and lenses from virtually every manufacturer.

If you own one of these lenses, and you see the yellowing, you can 1) shoot B&W only, particularly if you're still shooting film, 2) rely on auto white balance, or 3) correct for the yellowing in post. Reports vary as to whether the yellowing affects IQ. My guess is that it depends on whether the discoloration is uniform or not. OR, you can try to "de-yellow," or "bleach" the lenses. You'll read about putting the lens in the sun (not good if it gets hot), using some sort of UV light source (such as a small bug zapper, or even an industrial ultraviolet light. I'm going to leave the determination of safety to you. One approach I learned of involved an LED desk lamp from IKEA. Nobdy was able to explain how the spectrum of light from this lamp reversed the yellowing, but there was more than one testimonial, so I thought... what the heck? The original video talked about the lamp being $9.99. Well, it's more than that at retail now, but poking around, I found one (open box) for about $14, shipped. I figured that if it didn't work, I had a nice non-heat-producing desk lamp.

My test subject was a Minolta MC W Rokkor-SI 28mm f 2.5. You can see pictures of the lens here. The reviews I read spoke highly of it's imaging quality. It's uncommon, but not so scarce as to be stoopid expensive. It came in, and was moderately yellowed:

This one is looking through the lens at a white sheet of paper:

28SI_Through_Before.jpg
Subscribe to see EXIF info for this image (if available)


This is looking through it at the distance.

28SIDistant_Before.jpg
Subscribe to see EXIF info for this image (if available)


For reference, this is an un-yellowed Pentax 28mm, using custom WB for the conditions that day:

PTX28CWB_1B.jpg
Subscribe to see EXIF info for this image (if available)


This is a shot through the Rokkor SI using the same WB

28SICWB_Before_A.jpg
Subscribe to see EXIF info for this image (if available)


This is a shot of a white sheet of paper:

SI28CWBWhite_Before.jpg
Subscribe to see EXIF info for this image (if available)


But, if the camera does AWB, it looks pretty much like the unaffected Pentax:

28SI-AWB_Before_A.jpg
Subscribe to see EXIF info for this image (if available)


Bored yet?

Okay, here's the IKEA lamp:

IKEA_Lamp.jpg
Subscribe to see EXIF info for this image (if available)


And here is my bleaching setup:

DeYellowingRig.jpg
Subscribe to see EXIF info for this image (if available)


The flexible gooseneck lets you get the lamp all but touching the lens element. I put a mirror under it so that the light could conceivably cycle back and forth through the lens.

I had read online that it would take about 72 hours to clear a lens. Now, that example was an SMC Takumar; the Thoriated element in that lens is the rearmost one, so the lamp was acting directly on it. For the Rokkor, the affected element(s) are internal, as this lens is of a floating element design. After 72 hours, it looked pretty good to me...until I looked through it at the white paint on our front railing and the white sheet of paper. SO, I flipped the lens over, cut a circle of aluminum foil to put inside the rear lens cap (again for reflection) and set it up for another 72 hours.

As an aside, yeah, that's a long time. But, if you consider the low cost, and the fact that you're probably not going to be time sensitive using one of these vintage lenses, and that you don't have an ultraviolet light source running in your house, it's not that big a deal. So, did it work?

Here are the views through the lens:

R28SI_Through_1_After.jpg
Subscribe to see EXIF info for this image (if available)


R28SI_Through_2_After.jpg
Subscribe to see EXIF info for this image (if available)


A little bit of tint still visible. But what about the images produced?


Again, for reference, here is today's CWB with the Pentax:

PTX28CWB_2A.jpg
Subscribe to see EXIF info for this image (if available)


And AWB with the Rokkor SI.

R28SIAWB2_After.jpg
Subscribe to see EXIF info for this image (if available)


Pretty much the same, and no visible "warmth."

But the punch line is the custom white balance:

R28SICWB_After.jpg
Subscribe to see EXIF info for this image (if available)


To my eye, just a little "warm." In fact, on a dreary day like today, you might end up with look on purpose in post.

And the white paper?

R28SIWhite_After.jpg
Subscribe to see EXIF info for this image (if available)


Pretty near imperceptible.

So, can you return a vintage yellowed radioactive lens to near color neutrality with a sub $20, non hazardous lamp? Looks like it to me!
 

Attachments

Last edited:

barry13

Mu-43.com Editor
Joined
Mar 7, 2014
Messages
9,779
Location
Southern California
Real Name
Barry
Hi, good job and info!

FWIW, I believe it's the glue between the elements that turns yellow.

Ikea also sells a smaller, USB-powered, lamp that I've read also works (though it must be slower). I have one but I don't think I have any yellowed lenses currently.

The last one I had was a Takumar, and wrapping it in saran wrap (to keep out dust) and foil, and leaving it on a windowsill for a month worked. But a LED would certainly have been more convenient.
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,428
Well, it's more than that at retail now, but poking around, I found one (open box) for about $14, shipped.
Ikea discontinued the non-USB Jansjös and holders of remaining stock have been pricing up (though US $10 sounds more like older USB version pricing). I've never put mine on a spectrometer but if a Jansjö works I'd be surprised if anything else also using a 2700K ~70 CRI 70ish+ lumen LED didn't also work. The replacing Nävlinges are 220 lumen so, if there's nothing special about the Jansjö LED, they'd presumably de-yellow a bit quicker. (I find 70 lumen overly bright for other purposes, though photomacrographers using clusters of Jansjös would probably appreciate the extra 1.65 stops.)

Ikea also sells a smaller, USB-powered, lamp that I've read also works (though it must be slower).
Ikea doesn't currently specify lumen for the USB Jansjö but it's indicated as 10 lumen elsewhere. If there's a linear relationship between lumen and reduction in yellowing that would imply a month and a half's illumination.
 
Last edited:

MadMarco

Mu-43 Veteran
Joined
Oct 30, 2014
Messages
450
Location
Guildford, England
Another approach would be to create a custom white balance, this is what I do for my IR modified E-PL5. It works a treat to give a more natural tone to IR photos.
 

ektar

Mu-43 Top Veteran
Joined
Jun 8, 2017
Messages
588
Location
Western NC
Real Name
Don
FWIW, I believe it's the glue between the elements that turns yellow.
I've read that as an option as well. Short of finding an affected lens and sacrificing it in the name of science, I don't know how one could tell with certainty.

Another approach would be to create a custom white balance, this is what I do for my IR modified E-PL5. It works a treat to give a more natural tone to IR photos.
I do the same with my IR-modded camera. It's an option, but not if you're shooting and changing between yellowed and non-yellowed lenses and don't want to have to rebalance each time you swap. I'd probably go with AWB in that instance.

Ikea discontinued the non-USB Jansjös and holders of remaining stock have been pricing up (though US $10 sounds more like older USB version pricing). I've never put mine on a spectrometer but if a Jansjö works I'd be surprised if anything else also using a 2700K ~70 CRI 70ish+ lumen LED didn't also work. The replacing Nävlinges are 220 lumen so, if there's nothing special about the Jansjö LED, they'd presumably de-yellow a bit quicker. (I find 70 lumen overly bright for other purposes, though photomacrographers using clusters of Jansjös would probably appreciate the extra 1.65 stops.)
Didn't know it had been D/C'd, as I just did a search for one on eBay. You're right, of course, that spectrum and output are the variables.
 

MadMarco

Mu-43 Veteran
Joined
Oct 30, 2014
Messages
450
Location
Guildford, England
I do the same with my IR-modded camera. It's an option, but not if you're shooting and changing between yellowed and non-yellowed lenses and don't want to have to rebalance each time you swap. I'd probably go with AWB in that instance.
With the E-PL5 you can save the custom white balance to one of the registers, there are 2 on the E-PL5 so I can easily switch between custom, auto and daylight for instance. Obviously it's still better to fix the yellowing if you have the ability to.
 

Tywais

Mu-43 Veteran
Joined
Jan 23, 2014
Messages
289
Location
Chiang Mai, Thailand
Real Name
Mike
Working in nuclear/particle physics we have sometimes made Lichtenberg figures in acrylic for fun. We expose it to a high energy electron beam then hit it with a grounded nail or sharp item to cause a point discharge of the built up charge in it.

To the topic, the radiation (x-rays) that goes along with the electrons causes it to turn yellow/brown and same with glass. We place it in the sun for a period of time and the UV turns it clear again. Example below of final results plus a link to a video covering the yellow lens issue.

Subscribe to see EXIF info for this image (if available)


 

Erich_H

Mu-43 Regular
Joined
Feb 8, 2020
Messages
182
I just realized I've got two boxed Jansjös which I bought cheaply when my local IKEA had a sale.

But I don't know if I'm going to use them on my Takumar 1.8/50. Have to think about it.

If it's the glue that is yellowing, won't de-yellowing it possibly damage the lens?

Thanks/Erik
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,428
Hi, I wanted to mention I've never come across a white LED emission spectra containing ultraviolet. Based on @ektar's results here, that suggests against these two assumptions about how de-yellowing works:
You'll read about putting the lens in the sun (not good if it gets hot), using some sort of UV light source (such as a small bug zapper, or even an industrial ultraviolet light.
We place it in the sun for a period of time and the UV turns it clear again.
An alternative hypothesis might be that de-yellowing occurs faster with higher energy photos, at least up to some point where they become damaging (maybe X-rays as @Tywais suggested). If so, then white LEDs with higher color temperature or 365nm UV LEDs would be associated shorter de-yellowing times. The LEDs are fairly easy to get, so it could maybe be interesting to try something besides 2700K the next time someone needs to de-yellow a lens.
 

Tywais

Mu-43 Veteran
Joined
Jan 23, 2014
Messages
289
Location
Chiang Mai, Thailand
Real Name
Mike
Here is a group discussing different light sources for de-browning lenses and successes. High power UV led would be my choice, just stay away from UVC versions. But some just used LED lamps with success. A long time ago I built a UV light box with a large array of UV leds for production of PCBs. Think I got a bag of a hundred for $20. :)

https://www.flickr.com/groups/47441458@N00/discuss/72157609218601231/
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,428
Nice find. Seems like lots of variability in results, at least some of which is probably due to different levels of attention to how much yellowing is present. But it also seems like different lenses might de-yellow at different rates even if consistent lighting and standardized colorimetry were used.
 

barry13

Mu-43.com Editor
Joined
Mar 7, 2014
Messages
9,779
Location
Southern California
Real Name
Barry
Nice find. Seems like lots of variability in results, at least some of which is probably due to different levels of attention to how much yellowing is present. But it also seems like different lenses might de-yellow at different rates even if consistent lighting and standardized colorimetry were used.
I've read that if one of the inner elements is radioactive, it takes longer as the glass in the outer elements absorbs most of the UV before it can reached the yellowed ones.
But that only makes sense if it's pure UV light being used, or only UV light works (unlikely if daylight LEDs work).
 
Last edited:

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,428
Yes. If we're to critique that thread, I suspect the UV-C folks have an understanding that's functionally not that much different from the person who confused photons with protons. I've not come across a visible light optical glass specified beyond 250 nm (UV-C is 100-280 nm) and relatively few glasses intended for 400-700 nm use are particularly transmissive in UV-B (280-315 nm, Edmunds). This is why I suggested 365 nm LEDs for UV trials upthread.

There's I think also some question of the properties of the yellowing itself. Transmissive things get yellow because they block blue light, which seems plausibly associated with also blocking UV-A. If we presume de-yellowing is associated with photons getting absorbed and that transmitted photons have little effect, then it follows de-yellowing would be performed primarily by blue and maybe UV-A photos. If this is in the right direction then, for a given luminosity, higher colour temperature white LEDs should be associated with shorter de-yellowing times since they emit relatively more blue. In practice there'd likely be an additional effect as higher colour temperature LEDs are usually brighter (lower CRIs are usually brighter as well).

A corollary testable hypothesis would be that one could just use a blue LED. However, there's maybe some fiddliness over matching up narrower emission spectra with absorption and blue (or UV) LEDs are probably less useful for other things after a lens is done than white ones.
 

barry13

Mu-43.com Editor
Joined
Mar 7, 2014
Messages
9,779
Location
Southern California
Real Name
Barry
There's I think also some question of the properties of the yellowing itself.
Hi, there appears to be some useful info in the last message at https://www.bnphoto.org/bnphoto/LostSites_MonaghanRadioactive.htm

From: Andrew Resnick andy.resnick@NOSPAM.grc.nasaDOTgov
Newsgroups: sci.optics; Subject: Re: Radiation discoloring of optical glass; Date: 10 Apr 2003
There is quite a body of literature on the effects of ionizing
radiation on optical glass. Ionizing radiation (usually high speed electrons, but
also protons, nuclei, and high-energy photons) create what are called
'color centers' in glass. Briefly, the ionizing particle strips off
electrons from some of the atoms in the glass, creating a localized
defect. I am not familiar with the 'bleaching' that can occur upon
subsequent exposure to UV radiation.
Radiation-induced changes to the
refractive index are of concern primarily to the deep-UV lithography
industry and the space industry, and a few papers I have that may be
of interest to you are:

Gusarov et. al., "Refractive-index changes caused by proton radiation
in silicate optical glasses", Applied Optics 41, 678 (2002)
Isbi et. al., "Enhanced photoinduced [chi]2 in gamma-ray-irradiated
bulk glass", Optics Letters 25, 902 (2000)
Al-Jumaily, "Effects of radiation on the optical properties of glass
materials", SPIE vol 1761, 26 (1992)
Griscom, D., "Nature of defects and defect generation in optical
glasses", SPIE vol 541, 38 (1985)
Williams, R. T., "Nature of defects and defect generation in optical
crystals", SPIE vol 541, 25 (1985)

This should get you started.
--
Andrew Resnick, Ph. D.
National Center for Microgravity Research
NASA Glenn Research Center
I don't have access to the papers, although it looks like at least the second one might be possible to access for free with student registration.
 

barry13

Mu-43.com Editor
Joined
Mar 7, 2014
Messages
9,779
Location
Southern California
Real Name
Barry
There's a free-to-read paper at https://www.intechopen.com/books/ra...cts-in-optical-materials-and-photonic-devices

It mentions:
"optical attenuation recovery under UV radiation"
"annealing and photobleaching"
"Samples from different manufacturers exhibited radiation induced attenuation (RIA) saturation starting from 102 Gy. Crystal impurities and defects are the primary source of the optical attenuation increase in the 190–250 nm and 500–600 nm spectral bands induced by gamma rays"
"Surface topography studies indicated the presence of hillock in the irradiated zone" - hillock appears to be a crystallography term in this context.
"post irradiation photobleaching under He-Ne laser radiation laser (λ = 543 nm) and UV"
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,428
Yeah, it's unclear if the yellowing is from Canada balsam aging, glass nanostructure changes from α, β, and γ emission from the thorium chain, or some combination thereof. I can't find anyone that's taken apart a yellowed lens to get a detailed enough look to tell. Kessel 2015 does look at ideas similar to what we're considering here but certainly isn't directly relevant. Same goes for photochromic glass bleaching responses. Overall, the bits I've been able to find seem consistent with the speculations above but really the best way would be get some reasonably controlled (rather than anecdotal) data.

There's a UV-based de-yellowing kit on eBay with list of lenses it works on that's kind of interesting.
 

ektar

Mu-43 Top Veteran
Joined
Jun 8, 2017
Messages
588
Location
Western NC
Real Name
Don
That's an interesting kit. My concern about using UV lamps was both heat and containment, and he addresses that (with your box...)

I can't remember where I saw it, but someone identified the affected elements in a lens, disassembled it, showed the yellowing on a white background, then "bleached" the element out on its own. If I can stumble back onto the thread, I'll post it up.

It had not occurred to me, but in addition to the color cast (as discussed, easily remedied even without clearing the lens), you can easily lose fractions of an f-stop of transmission.

So, in discussing other spectrum or intensity LEDs, if you can source the LED bulbs, what lamp do you install those into?
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2009-2019 Amin Forums, LLC
Top Bottom