Jump to content
UltravioletPhotography

autofocus UV-capable 50mm lens?


Alex H

Recommended Posts

Browsing eBay I came across two ads by known here seller image-laboratory. And although I usually do not pay attention to camera conversion services (I do it all myself), one thing caught my eye.

 

Supposedly autofocus UV-capable 50mm lens sold in a bundle with Canon M2 (eBay ad # 201509430027) and Sony NEX 3N (eBay ad # 201389959265). The seller claims to be the only "service provider who can totally rebuild stock lenses to make them UV-grade" (quote), and posts a transmission chart in one of the ads showing useful UV-transmission to 320nm. The seller also has a bunch of warnings at the end of the post, about users not allowed to reverse-engineer the lens conversion.

 

Pictures in both ads show what appears to be partly or completely camouflaged Yongnuo EF 50mm F/1.8 lens. The transmission chart, on the other hand, specifies F/3.5 lens, and the ad – F/4 lens. How about that?

Link to comment

I don't even know where to begin in describing my reaction to this !!!!!

 

Later: OK, I've calmed down. There is a lot of hyperbole in the ad. But such is always the case with advertising.

The question is: Is there anything misleading?

 

The usual misleading item I see in image-laboratory's ads is that a stand-along Hoya U-340 or 360 filter will serve as a UV-only filter. This is of course blatently false. And I have written to image-laboratory about this. All he needs to do to clean up his ads is to mention that an additional BG filter must be used with any Hoya-U in order to obtain a pure UV photographs. He has not chosen to do this.

Link to comment

OK, I have returned to being absolutely incensed. I'm going to make a list.

 

If image-laboratory would simply respond to my suggestions to clean up his advertising to an ACCURATE level, then we wouldn't have to ever write about him here --- and eventually report him to Ebay for misleading advertising.

 

I am CONCERNED that naive customers or researchers who buy his conversions because they want to try shooting UV are going to see their UV+IR photos and wonder what the &%$# went wrong.

 

**********************

 

IMAGE-LABORATORY: BE ACCURATE in your ads and your business will grow. If your ads remain misleading and inaccurate, then Ebay will have something to say about it. I get a lot of complaints about you.

 

Olympus E-PL6 + 680nm filter Full Spectrum IR UV Astro Infrared converted camera

http://www.ebay.com/...verted-camera-/191617285672

 

Misleading Statement: Please see the first picture I took with UV 330nm (10nm full width) bandpass filter! NO ONE ELSE CAN CONVERT OLYMPUS CAMERA TO BE ABLE TO TAKE UV PICTURES IN THIS REGION!

Reason: All full spectrum converted cameras can record UV in the 330nm region (although it is not easy).

 

Misleading Statement: Please note, that this camera has Ultrasonic Sensor Cleaning System retained! This is alone $44 value! NO ONE ELSE CAN DO THIS! Unlike ANY OTHER conversion service providers we can retain sensor cleaning feature on most of our converted cameras!

Reason: All conversion shops can retain (certain) sensor shakers if they wanted to provide that as a service. They typically choose not to do so because it adds to the expense due to the re-soldering needed. The correct statement would be "most other converters do not do this" or "unlike most other converters we do retain sensor cleaners".

 

Misleading Statement: UV converted camera. This camera see only UV part of the spectrum. Which bandwidth to pass determined by the filters installed. Usual filters are U330, U340 and U360.

Reason: The Hoya U filters all pass significant amounts of IR. Any UV-only conversion making use of such filters must also supply some BG filtration either as an added layer internally or as an external filter on the lens.

 

Misleading 330 Reference in Photo: The dandelion photo (see below) may be a 330-only UV photo made with a narrowband 330nm filter with very strong IR supression, however the U-330 UV+IR filter mentioned everywhere else in this ad will not properly record a dandelion UV-bullseye as shown in this photo. Is the buyer made aware of this difference? Also if the U-330 is stacked with any BG filter to supress IR, then the transmission peak shifts to a longer UV wavelength and no longer peaks around 330nm.

[Edited this comment for accuracy on 27 March 2016.]

 

The fotos.

 

Amazing 330nm shot - but what filter?? Maybe image-laboratory does have an 10nm wide 330nm filter with OD4 IR supression. If so, he needs to give details in your ad. This foto will confuse buyers who think they might be able to get the same shot with a U-330 mentioned everywhere else in the ad.

oly330.jpg

 

 

Incorrectly labeled on left.

UVfilterNOT.jpg

 

 

Correct labeling. So we know that image-laboratory does know about Hoya U leakage of IR.

U330AsUV+IR.jpg

Link to comment

IMAGE-LABORATORY: BE ACCURATE in your ads and your business will grow. If your ads remain misleading and inaccurate, then Ebay will have something to say about it. I get a lot of complaints about you.

 

UV Full Spectrum converted Canon M2 + 50mm AF UV lens + 52mm Short-Pass Filter IR

http://www.ebay.com/...l0AAOSwh-1W3P8Q

 

TYPO: Sales title mentions "52mm Short-Pass Filter IR" but later on in the add a "52mm UV short pass filter!" is mentioned.

 

MISLEADING LENS SPEED: Lens is mentioned as f/4 in the ad, but the lens transmission chart is for an f/3.5 lens. Which one does the customer get? Where is the transmission chart for the F/4 lens?

 

MISSING ADAPTER?: Lens shown requires an adapter (shown) for the Canon M2. But the ad does not mention anywhere that the buyer will receive the adapter.

 

CANON + NIKON MISMATCH: After "What is included with the camera", we see this: "If you like to buy camera with lens please add $85 for Nikon 18-55 VR Lens! While supplies of lenses last."

SAY WHAT??: How well would this Nikon lens work on that converted Canon M2?

 

NO RETURNS: Every camera/lens we convert done per particular order. Therefore there are NO returns allowed as this is a custom build/modified item.

BUYER BEWARE: No further comment.

 

WILD HYPERBOLE: Simply put we are the ONLY service provider who can totally rebuild stock lenses to make them UV grade! No need to buy $5,000+ worth Fused silica lenses! You have got one included in this SUPER LOW PRICE!

REASON: We all are capable of taking a lens apart and replacing it with new elements which can pass UV. There is nothing new there. It has been done before. Now, Image-laboratory might be the only "service provider" who is currently doing this, but he is definitely NOT the only service provider who can do this.

Also, I do have to wonder what the heck this "rebuild" does to the original stock lens? I immediately note that it must (unavoidably??) introduce a degree of chromatic aberration into the rebuilt lens. I see NO mention or discussion of this.

 

WILD HYPERBOLE with editoral interspersed response:

PLEASE READ THIS PART CAREFULLY

  1. HOW DID WE MODIFIED THE LENS AND THE CAMERA IS OUR INTELLECTUAL PROPERTY
    [EDITOR: Only if you are granted a patent.]
  2. USERS ARE NOT ALLOWED TO TAKE THE LENS APART OR ANY OTHER WAY TRY TO REVERSE ENGINEER THE LENS/CAMERA CONVERSION.
    [EDITOR: How would you ever know if I took apart my lens? Reverse engineering is commonly performed on all cameras & lenses. Absolutely unenforceable and silly to request.]
  3. SHOULD INFORMATION ABOUT CORRECTION BECOME AVAILABLE TO THE USER/BUYER THEY ARE NOT ALLOWED TO SHARE SUCH INFORMATION WITH ANYONE ELSE
    [EDITOR: Another silly unenforceable statement.]
  4. SELLER MAY PROVIDE BUYER WITH PARTIAL INFORMATION ABOUT CONVERSION. SUCH INFORMATION IS PRIVILEGED AND CONSTITUTE INTELLECTUAL PROPERTY OF THE SELLER. THE MERE FACT OF SHARING IT WITH THE BUYER DOES NOT CONSTITUTE ALLOWANCE TO SHARE OR DISTRIBUTE IT BY THE BUYER/USER.
    [EDITOR: I hope the seller chooses buyers who can keep such "secrets".]

Link to comment
IMAGE-LABORATORY: BE ACCURATE in your ads and your business will grow. If your ads remain misleading and inaccurate, then Ebay will have something to say about it. I get a lot of complaints about you.
Link to comment

I might also point out that it is impossible to create the transmission profile shown in this ad/listing/graph with any kind of absorptive filter stack: eBay item number: 191832155399

Certainly not using BG39 or BG40 or any other BG type glass and still maintain Red/IR suppression.

This graph shown can not represent any UG11 + BG39 stack formula, nor any other such stack using any absorptive glass made by earth people, no matter what specific stack recipe you mix up.

BG glass cuts off at 300nm, and BG39 cuts off higher than all the other BG glass.

The filter therefore would not transmit anywhere close to 300nm. BG40 and S8612 are much preferred for this kind of task.

post-87-0-03940200-1459057050.jpg

Link to comment

Thanks, Steve!

 

To give the benefit of the doubt, the ads seem to indicate that a narrowband 330nm filter (with apparent strong IR supression) was used rather than a U-330. I edited my comment about the dandelion photo above to be more accurate about this. And marked the edit location.

 

But the naive buyer will see 330nm quoted in the photograph and then see the U-330 mentioned everywhere else and conclude that he/she will be able to replicate the photo results if they request (or separately purchase) a U-330 filter (or conversion). Quite clearly this cannot happen.

Link to comment

The Yongnuo EF 50mm F/1.8 seems to have problems autofocus in live view:

 

http://petapixel.com/2014/12/25/first-review-40-yongnuo-50mm-f1-8-nifty-fifty-clone/

 

http://www.henrysnote.com/2015/01/yongnuo-yn-50mm-f18-review.html

 

I can't imagine the conversion to UV capable has helped this issue.

 

I'm am curious whether this manufacturer used a UV friendly cement in the elements of this lens. That would make conversion to UV compatible a lot easier.

Link to comment

Thanks, Steve!

 

To give the benefit of the doubt, the ads seem to indicate that a narrowband 330nm filter (with apparent strong IR supression) was used rather than a U-330. I edited my comment about the dandelion photo above to be more accurate about this. And marked the edit location.

 

But the naive buyer will see 330nm quoted in the photograph and then see the U-330 mentioned everywhere else and conclude that he/she will be able to replicate the photo results if they request (or separately purchase) a U-330 filter (or conversion). Quite clearly this cannot happen.

 

U-330? You are referring to the subsection pertaining to camera conversions, which is not what this ad/listing is really about. Item number: 191832155399

That ad/listing is for a UV-only filter stack using Schott UG11 + Schott BG39. The graph is erroneous, because no stack using those two glass types can have that profile.

That is an impossible transmission curve using any BG type glass.

Link to comment
That was just a raw transmission graph made without any filtration??
Link to comment
Andy Perrin
That nice flat line outside the UV makes it look like the seller just...set everything to zero out there. Arbitrarily.
Link to comment

That nice flat line outside the UV makes it look like the seller just...set everything to zero out there. Arbitrarily.

 

Yep, exactly. That's what it looks like to me, like a UG11 or U-340 graph with the Red/IR peak errased. Of course, it doesn't work that way when stacking.

Once you stack with BG glass, even extremely thin BG glass like 0.25mm, you truncate the U-glass transmission curve severally.

The only way to get a curve like the one shown would be to use dichroic coatings, such as the Baader U.

Nevertheless, A UG11 + BG39 stack will work, but not the way the graph shows.

Link to comment
enricosavazzi
(...)

WILD HYPERBOLE: Simply put we are the ONLY service provider who can totally rebuild stock lenses to make them UV grade! No need to buy $5,000+ worth Fused silica lenses! You have got one included in this SUPER LOW PRICE!

REASON: We all are capable of taking a lens apart and replacing it with new elements which can pass UV. There is nothing new there. It has been done before. Now, Image-laboratory might be the only "service provider" who is currently doing this, but he is definitely NOT the only service provider who can do this.

Also, I do have to wonder what the heck this "rebuild" does to the original stock lens? I immediately note that it must (unavoidably??) introduce a degree of chromatic aberration into the rebuilt lens. I see NO mention or discussion of this.

(...)

I quite agree with Andrea about everything in her post, and particularly on the ridiculous, unenforceable conditions attached to the sold item (no disassembly, no reverse engineering, no telling others what you've found out - or else...what? end up on Santa's naughty list?).

 

I also strongly doubt that Image-Laboratory can rebuild a lens to transmit UV. In theory, this could be done in one of four ways (see the following discussion for technical obstacles to all of them):

 

1) Removing the lens front element, removing its coatings (chemically?), remounting the front element. The rest of the elements are untouched.

My comments: Many optical coatings are both harder and chemically more difficult to etch than the lens glass, so removing the coatings without damaging the glass is no simple feat. The outermost coating of the front element, at least in some legacy lenses, is the coating most likely to cut UV. The legacy coatings on internal lenses was often kept simpler (single-layer coatings, or softer coating materials) to save money, since most UV was already cut by the first coating. I believe Björn was the first, or one of the first, to publicize this on Nikon E lenses. However, modern multicoatings cost near to nothing and are extremely hard, so all elements in modern lenses are multicoated and may cut UV to a higher extent than single-layer coatings.

 

2) Removing most or all elements and treating them to remove the coatings.

My comments: With the complex optical schemes in use since the mid-20th century, I would expect that most uncoated lenses become useless because of high flare.

 

3) Removing the cement holding together cemented optical elements. Possibly done in combination with (1) or (2).

My comments: If the cement is Canada balsam, removing it could increase UV transmission substantially. Canada balsam is relatively easy to remove (if you have plenty of time to soak cemented elements in a mixture of alcohols or mild solvents for weeks. I have done this, albeit not to increase UV transmission). Modern lens adhesives are often difficult or impossible to remove without using solvents much better than Agent Orange at killing anything that grows in the neighborhood (a poetic exaggeration, but you get the idea). The glued optical surfaces are often not coated, because the relatively small differences in refraction indices (glass-to-glue-to-glass) do not cause much internal reflections here. Removing the adhesive from a cemented doublet creates two glass-to-air surfaces that substantially increase flare and lower contrast, but if there is only one cemented doublet, it can still be used with some care. There may be problems caused by the air gap, like Newton rings if the gap is uneven, and more.

 

4) Replacing some or all of the optical elements with new ones designed to transmit UV.

My comments: This requires an extremely sophisticated redesign of the optics to take into account the different parameters of the new elements. Most likely, none of the existing elements can be re-used, and all need replacing. The next, unsurmountable obstacle is that the lens barrel is built to accept elements of specific thicknesses and spacings. It would be a major feat of engineering to redesign an optical formula using different glass types while still using the same element thickness and spacing. If Image-Laboratory has the resources for doing this, they would be much better off making and selling their own lenses. I am sure Nikon, Canon and even the X-files would be interested in buying these modified lenses and reverse-engineer them.

 

Lens Performance Degradation after Disassembly/Reassembly

Now to the difficult part. Most commercially produced optical elements are slightly decentered or otherwise out of specification. Removing optical elements and then reassembling the lens, almost always, randomly rotates, offsets and shifts the elements with respect to each other and, most of the time, degrades the overall lens performance. Differences of the order of micrometers do matter in a multi-megapixel system. Small optical elements are particularly vulnerable to this. Disassembling and reassembling a microscope objective, for example, is a sure way to ruin it (unless it was already a poor performer from the start). Camera lenses are less sensitive, but without an optical bench and collimator, and a time consuming realignment, they also become worse after a manual disassembly/reassembly pass. I know that we all have opened and reassembled lenses for good or bad reasons, but believe me, there is a difference afterwards, or else the lens manufacturers would not need to test and align lenses on their production lines. On the other hand, a well-aligned lens full of molds may still perform worse than a misaligned, cleaned-up lens, so if the lens is a loss anyway, just go ahead without qualms.

 

If Image-Laboratory wishes to provide evidence in order to prove me wrong on the unfeasibility of converting a lens to increase its UV imaging performance (other than the old and well-publicized trick of removing the front element coatings of certain legacy lenses), that would be a chance to learn something really interesting.

 

There is, by the way, one special case in which it is feasible to (sometimes) increase the UV transmission of a lens without replacing elements with custom-made ones or messing with the coatings. As far as I know, this only applies to video lenses that have a built-in IR-cut and UV-cut flat filter mounted at the rear of all elements. Remove this filter, and the lens transmits NIR (usually quite well) and sometimes UV (if allowed by the glass and coatings). This filter is sometimes found at the rear of videocam lenses for very small sensors. The Gopro Hero series is an example, they have no built-in IR- & UV-cut filter in the camera, but there is one such filter at the rear of the lens. To save money, the filter is increasingly often a dielectric coating on the rear lens element, so no luck removing it, but you can always replace the stock lens with one designed for NIR surveillance video, without a built-in filter.

Link to comment
Thanks Enrico for your detailed post. I discussed this matter with Klaus and his thought, with which I agree, is that the entire optical assembly was replaced with the one from UV-capable lens. Note that the rebuilt lens is marketed as F/4 (or F/3.5 in the transmission chart), comparing to Yongnuo's original F/1.8. If that is indeed true, I will not be surprised if he is using JML 50/3.5 as a donor.
Link to comment

Enrico, thank you very much for your excellent run-down of the infeasibility of "lens rebuilding". This is a very valuable contribution.

 

I have taken the liberty of adding bolding to your main points for enhanced readibility so you will see an "Edited by Andrea B." tag at the end of your post.

 

**********

 

Alex: I will not be surprised if he is using JML 50/3.5 as a donor.

 

Hence the stern warnings not to take that thing apart lest the scam be discovered !! :D

 

Oh my heavens, I hope nobody sends Image-Laboratory some nice expensive lens for this rebuild disaster thinking that they are going to get back most of their original lens "parts".

Link to comment

One thing is apparent, this eBay add has generated some interesting discussion. I too communicated with this seller when he first posted the Canon M & lens to point out that while a plot was referenced there was no transmission plot shown. To his credit the plot was quickly added. Further messages suffered some of the earlier communication issues that some time ago made me feel that English might not be his native language.

 

I offered some constructive points which didn't seem to be appreciated so I stopped trying. Unfortunate really, but some folks are just like that.

 

People who think they know it all are particularly irritating to those who do. B)

 

One thing I still find interesting is that he states his lens not only auto focuses but also maintains full aperture control. This suggests that it may not be an entire optical block but a split assembly using the existing aperture. Googling teardowns of the 50mm Canon vs Yongnuo lenses are easy to see how this could be done if the spacing of donor lens groups could be maintained.

 

Alex,

Speaking of autofocus UV-capable lenses I expect you have seen this posting from last year on the MFlenses forum. There is also a

showing another similar Minolta AF to Sony adaptation, perhaps you can understand the speaker as I certainly cannot. Using an autofocus capable adapter from a mirrorless Sony to one of these Minolta AF empty lens carriages ought to work similarly. This concept reminds me of how you mounted a 35mm f/.3.5 optical assembly on a focusing helicoid which you earlier shared with us in another post in this forum.

 

Speaking of "empty lens" our member Johan has an interesting post on his site about using an "empty lens" with a working aperture-control upon which to reverse mount a decoupled macro. I wonder if one were to mount the optical block of one of known UV capable manual focus lenses in front of a working aperture-controlled AF "empty lens" would the location of the aperture behind the lens still be usable?

Link to comment

One thing I still find interesting is that he states his lens not only auto focuses but also maintains full aperture control. This suggests that it may not be an entire optical block but a split assembly using the existing aperture. Googling teardowns of the 50mm Canon vs Yongnuo lenses are easy to see how this could be done if the spacing of donor lens groups could be maintained.

 

That will involve some coding of the lens firmware to work properly. Youngnuo is F/1.8-22 and UV-lens is F/4-16, if I remember correctly.

 

Alex,

Speaking of autofocus UV-capable lenses I expect you have seen this posting from last year on the MFlenses forum. There is also a

showing another similar Minolta AF to Sony adaptation, perhaps you can understand the speaker as I certainly cannot. Using an autofocus capable adapter from a mirrorless Sony to one of these Minolta AF empty lens carriages ought to work similarly.

 

Sure, I know about that and was thinking about the same mod myself, mainly for video with full-time autofocus. Unfortunately, cheap lenses for Sony E-mount are not strong enough to autofocus the UV lens that I want to use for this, while mechanically better and stronger Sony E-mount lenses are too expensive just to be destroyed.

 

Speaking of "empty lens" our member Johan has an interesting post on his site about using an "empty lens" with a working aperture-control upon which to reverse mount a decoupled macro. I wonder if one were to mount the optical block of one of known UV capable manual focus lenses in front of a working aperture-controlled AF "empty lens" would the location of the aperture behind the lens still be usable?

 

It may make the whole system telecentric or hypercentric, it may also affect other optical properties of the lens, but one has to test either actual lens+aperture combo in real life or use one of the lens design software.

Link to comment
enricosavazzi

(...)

This suggests that it may not be an entire optical block but a split assembly using the existing aperture.

(...)

I thought much the same after reading Olexandr's comment. It would be difficult to modify the iris of a "foreign" lens to be still operated correctly by the camera, or to measure exposure correctly at full aperture. The placement of the iris with respect to front and rear optical subassemblies is somewhat critical, but not extremely so.

Speaking of "empty lens" our member Johan has an interesting post on his site about using an "empty lens" with a working aperture-control upon which to reverse mount a decoupled macro. I wonder if one were to mount the optical block of one of known UV capable manual focus lenses in front of a working aperture-controlled AF "empty lens" would the location of the aperture behind the lens still be usable?

It might or might not work. If the aperture is displaced too much toward the rear compared to its optimal position, it might function as a "variable vignetter" instead of an ordinary iris. As Olexandr also says, it might turn the lens into a telecentric one if placed at a distance from the lens equal to its focal length.

Link to comment

Unfortunately, cheap lenses for Sony E-mount are not strong enough to autofocus the UV lens that I want to use for this, while mechanically better and stronger Sony E-mount lenses are too expensive just to be destroyed.

 

Old Minolta AF lenses with iffy glass should be cheap enough but the adapters Sony LA-EA3 for self motorized lenses and LA-EA4 for screw driven lenses are rather pricey.

Link to comment
That is true, John, but in that case I would rather use one of the more recent Sony DSLT bodied with screw-drive autofocus motor and removed translucent mirror.
Link to comment

Sure Alex, a Sony body with screw drive would operate a screw drive lens. Not being a Sony owner, I did not know the autofocus would function without the translucent mirror.

 

I initially thought the translucent mirror in the LA-EA2 & LA-EA4 was a reflex mechanism which flipped out of path but am now uncertain. If not that would seem to render the EA2 & EA4 useless for UV since the translucent mirror is likely not transmitting UV well and without the mirror the autofocus sensors in the EA2 & EA3 are blind.

 

So is the solution for a mirrorless Sony to use the LA-EA1 or LA-EA3 and a motorized self driven lens?

 

Added later:

I also recently ran across this Leica M to Sony E autofocus adapter. Adding a LTM to M adapter flange to an LTM lens should open the possibility to autofocus the L39 lenses as well.

Link to comment

I tested an unmodified Sony A55 for UV - translucent mirror did produce some funky reflections and flares....

 

Motorized self driven lenses are either too cheaply built (SAM) or to expensive (SSM) for my purpose.

 

Techart adapter has limited focus travel in my opinion.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...