coated filters on uncoated lenses...

[Carioca]

Hi,
I have purchased a couple of uncoated barrel lenses and would like to know, if the use of coated (ex. skylight filters) could slightly improve image quality, in terms of flare and contrast?
As all other lens surfaces inside those lenses are uncoated as well, I do not expect miracles, but I'm curious to know if some of you eventually had some testing done regarding this matter.

Thanks,

Sidney

[BrianShaw]

I doubt it, but a decent lens hood might.

[Ron Marshall]

There are reflections at every glass air interface. Coatings reduce the amount of reflection. Adding another glass air interface, ie. a filter will not improve image quality.

If you plan to use filters multicoated will be better than uncoated.

[Robert A. Zeichner]

Every air to glass transition involves a loss in transmission that varies from as much as 4% for uncoated surfaces to less than .1% for the very best multicoated surfaces. Single coated lenses usually exhibit around 1% loss. In addition to these transmission losses (which translates to reflections bouncing around inside the lens) there are other factors that contribute to veiling glare. Among these are how well the inner surfaces of the lens barrel are finished and/or baffled and the condition of the glass surfaces themselves (haze or oxidation from age or condensation of lubricants that have evaporated from mechanical parts of the lens housing).

That said, putting any filter in front of an uncoated lens cannot possibly improve the situation, only reduce the amount of additional degradation.

The best way to maximize the contrast and minimize the negative effects of veiling glare is the use of a proper shade. In an upcoming edition of PhotoTechniques, there will appear an article I wrote that investigates this very issue and offers a very effective and affordable solution to the problem. The improvement I was able to achieve resulted in nearly a full contrast grade of additional local contrast.

[Doremus Scudder]

I was always under the impression that by using a multicoated filter on an uncoated lens one could reduce the amount of possible flare. The reasoning goes like this. Since the multi-coating on the filter cuts out 3+% of the possible flare, the uncoated elements would have less flare to deal with, i.e. 4% of <0.1% rather than the original 4%. Could someone point out the error in this for me please, since I cannot seem to find it.

Best,

[JW Dewdney]

Doremus,
I think that makes a certain sense - especially when you consider that most COATED lens assemblies only have coatings on the OUTSIDE of the assembly. Like M&Ms (american candy). There's no real reason to have it internally. The film, I'm sure, really doesn't understand that the filter is an added accessory - and as far as IT'S concerned, it's a lens with a coated flat front element.

Robert's argument - while it sounds pretty good - doesn't take into account that multicoated lenses are only coated on the outside.

[Marko]

I was always under the impression that by using a multicoated filter on an uncoated lens one could reduce the amount of possible flare. The reasoning goes like this. Since the multi-coating on the filter cuts out 3+% of the possible flare, the uncoated elements would have less flare to deal with, i.e. 4% of <0.1% rather than the original 4%. Could someone point out the error in this for me please, since I cannot seem to find it.

Best,

Doremus,

Your logic is correct but it applies to each glass-to-air surface. Every glass-to-air surface without coating in the optical path has a potential for full flare.

Using multi-coated filter would defnitely reduce flare, but only compared to using non-coated filter. Compared to no filter at all, I think it would introduce yet another glass-to-air surface and it would therefore increase potential for flare.

{edit}

Trying to simplify: flare is a property of glass, not light itself. Every time the light touches glass, most of it gets transmitted (and refracted in the process), but part of it gets reflected as well. The greater the incident angle, the greater the reflection. When that reflected light hits the back surface of the glass element it already passed previously, the same thing happens again, part of it gets reflected back again and we see that as flare.

Again, this happens only on the boundary of two materials with a significantly different refraction indices, such as glass and air in this case. The greater the difference, the more pronounced the effect. The role of coating is to both reduce this difference and to absorb reflected light.

{end edit}

[Robert A. Zeichner]

Doremus,
I think that makes a certain sense - especially when you consider that most COATED lens assemblies only have coatings on the OUTSIDE of the assembly. Like M&Ms (american candy). There's no real reason to have it internally. The film, I'm sure, really doesn't understand that the filter is an added accessory - and as far as IT'S concerned, it's a lens with a coated flat front element.

Robert's argument - while it sounds pretty good - doesn't take into account that multicoated lenses are only coated on the outside.

The reasons for coating lens surfaces are as valid inside the lens assembly as on the outside. Reflection of light rays happens at every air to glass surface as Marko articulated in his post. The transmission losses are cumulative and the reflections echo back and forth as they pass through one glass/air boundry further exacerbating the problem.

Modern zoom lenses with as many as 18 to over 30 elements would be unusable were it not for sophisitcated coating techniques throughout the lens. Transmission loss would be enormous. As a matter of lens making history, the first experimental coatings were applied only to the interior surfaces of the lens as they were so fragile, they would be destroyed if touched.

The purpose of multicoating is to broaden the bandwidth of light over which the coating can effectively improve transmission through an element, therby reducing loss to a small fraction of a percent. It seems to me that this would be as important inside the lens as outside since the losses are cumulative.

[GPS]

Good quality binoculars have commonly multicoated surface on all its lenses.

[BrianShaw]

... especially when you consider that most COATED lens assemblies only have coatings on the OUTSIDE of the assembly. Like M&Ms (american candy).
Is this really true? I'd be interested in finding source material to support/refute this. Somehow, I don't think that Zeiss or the other "good" lensmakers ignore the importance of coating internal lens elements.

[James E Galvin]

I have had many lenses apart, photographic, microscope, telescope.... All of them that were coated were coated on all glass to air surfaces, external and internal. Inner cemented surfaces are not coated, as the small index of refraction difference between glass and cement reflects much less than glass to air (the main reason for cementing). Each surface adds flare. A coated surface adds less flare. So a coated filter adds less than an uncoated one, but still adds. To reduce flare, use a shade.

[Dan Fromm]

Come to think of it, Jim, IIRC there's been a rash of separation problems with some very recent Schneider lenses. One of the explanations brought forwards is that for some very good reason Schneider didn't coat the outermost 1 mm of surfaces to be cemented, and this very tiny gap provided an entry for something evil that attacked the cement. Possibly just another urban legend ...

[JW Dewdney]

Well - I'm only speculating (somewhat wildly) here - but it seems to me that any gain from attempting to coat interior air-contact surfaces would be easily lost in terms of the losses you would incur transmission-wise, which is to say that coating multiple interior surfaces would really cut down on the effective speed of the lens. I would GUESS the loss to be somewhere around 1/30th of a stop per surface. But hey - I'm talking through my hat. It'd be interesting to have an official statement from a respected lens manufacturer about this.

[James E Galvin]

I haven't heard of surfaces to be cemented being coated, but that doesn't mean that it isn't done. It is usual for the coating to not reach the edge of an element, the element is in a holder while coating, masking the edge. But when looking at reflections in a lens to determine its construction, the dim reflections, even in an MC lens, are the cemented surfaces. So antireflection coatings on cemented surfaces are not needed. I would think also not desired, as the adhesion of a coating is an issue. The two pieces of glass have different thermal expansions, so the cement is stressed with temperature variations, indeed lenses larger than perhaps 3" are not cemented for this reason. A coating with poor adhesion would then increase the possibility of seperation. And any gap at the edge is much thinner than the cement. The seperated lenses I have encountered have been a half century old, cemented with balsam. The elements were not coated on the cemented side. Interesting that some new lenses are having a problem, don't have an answer.

[Mark Sawyer]

As a polarizing filter reduces reflections in front of the lens, would use of a polarizing filter reduce reflections (flaring) within the lens?

[James E Galvin]

The coating is not a filter. It reduces reflections by constructive interference. For a single coating, it is a 1/4 wave thick layer of magnesium fluoride, a transparent material, applied by evaporation of the MgF. The light not lost to reflection is transmitted. So coating increases the transmission of the lens.
Light reflected at an angle is polarized, with a maximum near 45 degrees. The polaroid filter can remove the polarized part, reducing the reflections. The reflections in the lens are at small angles, and a variety of angles as you go around the lens. No help.

[JW Dewdney]

James - don't you think it's possible in any scenario - that a polarizing filter in front of a lens could block out stray light that would otherwise have caused some visible flare? It would appear to me that the success of this solution would be completely circumstantial - though generally speaking, I'd have to disagree with your final resolution for this very reason. To my mind, polarizing filters can deliver crisp, high contrast images BECAUSE they're reflecting a good portion of non-image forming light (depending entirely on the ability of the user to identify the correct angle of 'polarization' and the specific scene/camera orientation involved). Though generally - I see no fault with your logic. Just your conclusion.

[James E Galvin]

If bright reflected (and thus polarized) light is coming from the scene and causing flare in the lens, you are right. Dim down the bright light with a polarizing filter. I was thinking of the reflections in the lens, from a "normal" scene, that is one without bright reflections.

[BrianShaw]

This has been a great discussion, but frustrates me because of the lack of "real proof". I'm sure much of this is concealed as proprietery process, and I'm sure there is quite a bit of variability depending on the quality/cost of lens itself.

Refering back to my earlier post questioning a comment about which lens elements are coated (just the outside or all), I found a couple of interesting tidbits. Also, I might have mis-read the intent of the post stating that just the outside elements were coated - I read that as front and back, but perhaps it was meant to mean all but hte cemented itnerfaces.

Kingslake speaks of coating all air-glass interfaces but implies (relatively directly) that coating of cemented interfaces isn't of much value.

The Zeiss website has a step-by-step description of Zeiss lens production. They speak of a step where each lens ELEMENT is coated and inspected, then centered and glued into groups, prior to final assembly.

I also looked into some lens patents that I collected in the past and they speak of the lens design but do not speak of quality factors such as coatings.

[Dan Fromm]

Brian, FWIW, in some of its discussions of coatings the Vade Mecum refers to early post-WWII lenses from a variety of makers that were hard coated on exterior surfaces and soft coated on interior surfaces. They remark that these lenses are very vulnerable to overenthusiastic cleaning.

I've completely dismantled a Lomo RF-5 (450/10 process lens), which turned out to be a 4/4 double Gauss type. All surfaces coated. I've taken a number of Boyer tessar, heliar, plasmat and double Gauss types as far apart as possible without uncementing and all of the surfaces I found were coated.

It pains me to say it because he was a powerful force for good and a great man, but in some ways Kingslake seems to have been a little batty.

Cheers,

Dan

[BrianShaw]

Very interesting, Dan... thanks for the comments on your lens dissections. Re: Kingslake... I won't mention to him what you think of him ;-)

Where does one get this Vade Mecum... I saw a web stie offering it once but I'm not so sure if its for real or not. http://members.aol.com/dcolucci/sell.htm No contact information is provided, just a send PayPal option. I'm okay with trusting, but have no idea if this is "the source" or not.

[Ole Tjugen]

Dan Colucci is a good source for the Vade Mecum. No worries there. :)

The Vade Mecum is a great source for information on British lenses, fair for other lenses as sold in the UK (but beware of differing "Series" designations), and frustrating on anything else. On the whole, it's invaluable!

[GPS]

A little bit of the common sense: Imagine a projection screen in a room with a window. Through this window (representing a filter), from the outside world, you project on the screen (representing a film in a camera) a slide with a slide projector (the useful, picture making light). At the same time, sun rays (the parasite light) go through the window and fall on the screen, causing flare, diminishing the picture contrast etc. Now, if you have the window NON coated, more of the flare causing sun rays get reflected and less off the flare causing light falls on the screen. BUT, also more of the useful projector light gets reflected and less of it falls on the screen. If you have the window COATED then less of the sun flare causing light gets reflected (more transmited) and more of it will cause the flare on the screen. In the same time, the same is true for the projector light - less of it gets reflected, more transmitted. The question is - in what proportion those 2 lights in each case get mixed? It is surely possible that the sunlight can come from a direction where its rays get more reflected (grater loss of it ) on the window than the loss in the window from the projector light. Simply put, the windows (filter) glass can act, in some cases, as a mirror for the sun light (getting rid of it on the screen) but not for the projection light, comming from a different angle.
Of course, use of an uncoated filter (that acts better for reflection of the sunlight than more transmitting coated filter) in this way on a camera lens is highly impractical.

[Helen Bach]

The proportion of reflected light at an air-glass or glass-glass surface is, for light normal to the surface:

(1-x)^2/(1+x)^2

where x is the ratio of refractive indices of the two materials.

That means that the lower the ratio, the lower the reflection – hence it is less important at a glass-glass surface (where the maximum ratio is likely to be about 1.2, corresponding to 0.8% reflection) than at an air-glass surface (where the ratio is likely to be between 1.5 and 1.8, or 4% and 8% reflection).

If that surface is altered by the addition of an interposing surface, ie a coating, the overall amount of reflection can be reduced. This is separate from the interference effect that is dependent on the thickness of the coating. The ideal refractive index for a coating is the square root of the product of the refractive indices of the two materials comprising the surface.

So, for an air-glass surface the ideal refractive index lies between 1.22 and 1.34. For glass-glass surfaces it is around 1.64 for the greatest likely difference in RI.

Magnesium fluoride, MgF, has a refractive index of 1.38 so it is slightly higher than the optimum for many glasses, but close enough, and it is always between the RI of air and of glass. It is quite wrong for coating a glass-glass surface because it would actually increase the amount of reflection by increasing the ratio of refractive indices.


Best,
Helen

[Helen Bach]

Sorry, MgF2, not MgF.

I'm only referring to single coating, and the intention was to put some simple numbers to what has already been discussed.