Use of Filters for Protection

[2F/2F]

I don't use them because they seem to be a magnet for dust, fingerprints, and glare, which cause more of an effect when suspended half an inch or so from the front element than they do if on the front element itself. I use a lens hood instead. My only damaged front element happened when a filter shattered in a collision that would have just hit the rim of the lens if the filter was not there. The broken glass was jammed into the front element, and caused one large scratch and two smaller ones. I use them when there is mist in the air, if I happen to have one, but no other time.

 

[Q.G.]

The answer will inevitably be: sometimes it makes a huge difference (overall loss of contrast, multiple ghosting etc.), and sometimes it's totally unnoticeable. I simply don't enjoy having to think about whether there will or won't be ghosting from my filter and wondering I will be able to see it in the viewfinder or on the ground glass. These things have a way of not presenting themselves until you're just about to make a print.

Well, if you just remember that filters don't 'ghost' (they can make things worse, yes), but that it is light bouncing off glas surfaces that is doing that, and that your lens consists of nothing but glass surfaces, you'll also know that you should be watching for that without filters too.

Which leads to:

N.b. if you do need a filter, then you need a hood as well. (I hood almost all the time anyway, but my point is that a filter makes a hood all the more important)

Think about it:
You are using a hood all the time. Because a hood is very important (it indeed is).
The downside of using a filter is that it makes using a hood important, and you then must use a hood all the time.

Hmm...

(You noticed how i glossed over "almost". That's because it's hard to comprehend how one can worry about filters on a lens, and not use a hood all the time).

Any way you slice it, a filter- multicoated or not- introduces two extra reflecting surfaces between yourself and your subject.

That is undeniably true.

Also, if you are going to use a filter all the time, then you will want to put a multicoated b+w on each and every lens. You won't want to be screwing the things on and off all the time, sharing between lenses, because sooner or later you will cross threads and have to do a circumcision.

So buy filters for each one of your lenses. Where's the problem?

I don't get the logic that a filter is cheaper than a new lens. So? Being careful doesn't cost anything at all. Just to be annoying, I think I will start saying that every time these discussions pop up. Nah, better not, I am annoying enough already :rolleyes:

"Being carefull costs nothing at all" ...

 

[keithwms]

:rolleyes:

 

[Ed Sukach]

I've heard the idea that "a filter introduces two more glass-to-air surfaces, therefore adding to the possibility of increased flare, "ghosting" (? hard to see how, with fairly decent geometry), decreased contrast ... and a number of other optical disasters.

There IS a parallel: it is common in our perception of "quality" to link the number of elements in a lens to overall quality. A six-element lens will be "better" than a five-element lens ... even though there ARE two additional glass-to-air surfaces. The argument could be made that the prestigious lens manufacturers take more care in manufacturig, use more well-defined glasses, but I think all that is offset by the complexity of a
cohrently curved surafes, and its integration into an optical design.

Bottom line: elements MAY be added to improve the overall performance of the lens, with due consideration of their negative effects. They are usually well worth it.

One thing puzzling...

A few here have responded that they will only use a protective filter if it is misting, foggy, or raining.
Uh...
wouldn't condensation occur equally readily on the front element of a lens as it does on the front surface of a filter?

 

[2F/2F]

One thing puzzling...

A few here have responded that they will only use a protective filter if it is misting, foggy, or raining.
Uh...
wouldn't condensation occur equally readily on the front element of a lens as it does on the front surface of a filter?

I am not worried about condensation. That happens from changing temperature, and aside from making gradual shifts in temp, there is nothing you can do about it. I have never run into it except in my viewfinder when shooting in the snow after getting out of a warm car. I use filters in foul weather to physically keep water off of (and *out of*) the lens.

 

[david b]

On every single lens I have, there is a UV filter.

 

[David A. Goldfarb]

I've been photographing on the beach where the wind is coming in, and I'm wiping salt water off the filter every few minutes. That's a case where I'd rather be cleaning a filter than the front element of the lens, so it makes sense to use a filter. Otherwise, I pretty much always use a lens shade to reduce flare and protect the lens, and I keep my lenses capped or cased (some of my old lenses don't have caps) when not in use.

Lens manufacturers do try to reduce the number of glass-air surfaces, even with the best modern coatings, by using aspheric elements, which usually make it possible to replace two elements with one.

 

[Q.G.]

Lens manufacturers try to make lenses that comply with demands.

Those could be for using less glass (to save money).

But also to produce good lenses.
And then they really do not skimp on glass elements.

Moreover, the demands for fast autofocus also require rethinking lens designs.

The result is that when you compare the modern computer optimized IF lenses to the simple thingies that were made not that long ago, you'll see that the average number of elements per lens has gone up quite a bit.

Never a complaint heard. But this filter-thingy keeps resurfacing.

If you want to avoid flare, use a hood.

 

[Q.G.]

[...]
There IS a parallel: it is common in our perception of "quality" to link the number of elements in a lens to overall quality. A six-element lens will be "better" than a five-element lens ... even though there ARE two additional glass-to-air surfaces. The argument could be made that the prestigious lens manufacturers take more care in manufacturig, use more well-defined glasses, but I think all that is offset by the complexity of a
cohrently curved surafes, and its integration into an optical design.

Bottom line: elements MAY be added to improve the overall performance of the lens, with due consideration of their negative effects. They are usually well worth it.

Just so noone gets the wrong end of the stick:
Elements may also be left out to improve lens quality. (And i mena improve lens quality, not reduce the risk of flare).

There is no hard and fast rule, no fixed relation between number of elements and lens quality.

 

[Curt]

There should b a third option - Sometimes

.

I don't always, (b u t) a while back in Yosemite, up the road to Half Dome, next to the lake there is parking lot, there are stairs leading up to a path leading to a great view of Half Dome. The stone stairs are wonderful, covered with pine needles and grains of sand. I saw a squirrel out of the corner of my eye, I turned to see it, my feet semi-turned, my camera was around my neck on a long strap. I turned, it turned away and the lens struck the rock siding of the stairs. Metal bent, glass shattered. All was saved except the filter which was shot. The lens survived in perfect condition. I was POed, the squirrel was a great bystander. He/She probably was laughing inside, it was shaken not stirred.

 

[Ed Sukach]

It isn't difficult to find specific, rifle-shot instances that contradict every/ any "sweeping generality".

1. Have lens designs been modified by eliminating an element for the SOLE purpose of reducing lens-to-glass surfaces? It is possible. I would submit that it must be rare. Can anyone cite a specific exampke?

2. Have lens designs been modified by the elimination of an element to improve overall optical performance? Possible- even more rare. An example?

3. Have lens designs been modified by the introduction of an additional aspheric element to reduce cost?
"possible" here is somewhat of a stretch. I guess there must be a possibility - at least if the added element is of molded plastic. Aspheric glass is **expensive** and difficult to produce. Any examples?

Again, I would view "examples" as "exceptions to the rule" - provinng that there ARE exceptions, not disproving the rule itself.

Now for the shouts of, "Too technical!! Too technical!!
It is NOT my intent to "complicate" the discussion, but at the same time. distortions in the basic concepts (n.b. "BASIC") of design must be avoided.

Optical Design is a wondrous thing!!!

 

[Uncle Goose]

I have a filter on all my Bronica lenses, simply as protection. I have already seem a lot of lenses with damaged front lens coatings because the users were stupid enough not to protect their lenses. It's like when using certain powertools, what's cheaper, a new pair of safety goggles or eye surgery?

 

[Philippe-Georges]

Outdoors, I would not feel secure whit out, but in the studio, I do carry filters but for an other reason than pure protection. Always a hood too!
Uncle goose has a very good motivation...

Philippe

 

[David A. Goldfarb]

Kodak made a disposable with a plastic aspheric lens to reduce cost, so yes, that's been done.

The Planar design is over a hundred years old, but didn't come into its own until the advent of lens coatings, because it had too many glass-air surfaces, and the loss in contrast outweighed the improvement in resolution over the Tessar and other simpler designs before coatings.

With improved coatings, it is possible to use more elements to make a better corrected lens, so additional elements add something optically to the lens formula. Unnecessary filters only reduce contrast and add nothing optically.

Some lenses like large fast teles have their own multicoated protective glass--sometimes two layers--particularly if the large front element is not only large but exotic glass and thus very expensive to replace, and the lens is marketed mainly to sports and nature shooters who sometimes work in environments that can be hazardous to lenses.

 

[Joe Grodis]

Trick question... Generally I will use a filter on a pricey Nikkor or Zeiss lens. A cheap plastic quantaray I don't bother.

 

[Prest_400]

It isn't difficult to find specific, rifle-shot instances that contradict every/ any "sweeping generality".

2. Have lens designs been modified by the elimination of an element to improve overall optical performance? Possible- even more rare. An example?

The Olympus Zuiko OM 85 f2.0; It has a pair of designs, the later one eliminated one of two cemented elements for a single one.
http://www.mir.com.my/rb/photography/hardwares/classics/olympusom1n2/shared/zuiko/htmls/85mm.htm

 

[keithwms]

Ed, IMHO it is inappropriate to speak of reflections deep within the barrel in the same context as reflections that you can get between the filter and front element. The front element sees way more light than actually forms the image.... that's why we hood...

Anyway, there are good reasons why doublets are cemented and not air spaced.... but this is really beside the point. Again, the stray light at (and beyond) the front element is the issue here.

People just need to do what works for them, and if they do offer an opinion, they should back it up as transparently as possible! I have stated my reasoning very clearly ( I think): a filter, no matter how multicoated and German, introduces two extra surfaces in front of the front element. That's it! Either that means something to you or it doesn't! And you probably don't want to see a detailed ray diagram to demonstrate what I am saying, so let's just let it go

Part (b) of this logic is that filters are not considered in the lens design... they could not be! That's because the thickness of the lens, the dispersion of the glass, the spacing form the front element etc. are all variables that you'd need to know in order to do it right. And lo! The filters are -flat!-... not curved like the front element, so that calculation would be really dicey. Now, there are examples of filters that are considered within the overall lens design: for example, the lenses with slip-in filter compartments within the barrel. You get a set of filters with the lens. But again, the front element is what sees a lot of extra stray, non-image-forming light.... etc.

There is one recent notable exception to this rule that lens designers don't like to see filters in their lenses. The leica m8 has the issue that they couldn't place a [IR-reflecting] hot mirror between the sensor and the rear element. so for that one camera, one is obligated to put a hot mirror over the front element to avoid purply blacks.... just because the digital sensor has extended IR sensitivity. Now, I don't know about you, but I'd be hopping mad if somebody told me I had to use a hot mirror over my front element. But anyway, that is the only example of which I am am aware that the factory actually suggested that a front filter could improve the image. And clearly it is a major compromise forced by the camera design itself.

I admit that I have a hard time considering the cost of a lens in this discussion. If we're talking about ways to promote the ultimate image quality, then the discussion is simpler: don't use filters unless absolutely necessary, and always hood. Simple. Indisputable. Count the interfaces.

If we start considering everyone's individual finances and how that influences the way they shoot, well then these discussions always go off in some meandering direction. Personally I don't understand the fear of using superfantastic lenses unfiltered because of what may happen if they are careless. That which you are afraid to use as the lens designers intended should be at home in the velvet box. Sorry to be blunt, but really.... (here we go! And this isn't without ample forewarning!)....

It doesn't cost anything to be careful.

---

Just be obnoxious, I'll just add one other curveball to this: sometimes you actually want flare and such. Here is such a case, something I shot last autumn with an rb fisheye. Fisheye into the sun. Lotsa stray light and weird flare and such... and totally intentionally set up to pick up all those things, but... wait, there is an illustrative question after you look at the image...

(there was a url link here which no longer exists)

Now the question: how much of that do you think I was able to see on the ground glass?
.
.
.
.
.
Answer: almost none of it.

Thinkaboudit

 

[Philippe-Georges]

When flair due to reflections on filter surfaces is an issue, then one can consider the Pentax SMC - or B+W multi coated filters, to name a few, not cheap but effective (I hope so...).

Philippe

 

[pgomena]

I generally don't use a filter for protection. The exception is when shooting near the ocean, where I will use a UV filter to protect the lens from salt spray.

I worked as a photojournalist for a few years and never used a protective filter. I just left the lens hoods on the lenses all the time and put them hood-down in my sectioned Domke bag with the rear lensdcap on. Never scratched a single lens. Early on, I had an experience where a skylight filter caused nasty reflections when I photographed lights at night. Never used one as general protection after that.

Peter Gomena

 

[Ed Sukach]

The Olympus Zuiko OM 85 f2.0; It has a pair of designs, the later one eliminated one of two cemented elements for a single one.
http://www.mir.com.my/rb/photography/hardwares/classics/olympusom1n2/shared/zuiko/htmls/85mm.htm

There is no question that this lens design was modified over time ... but there is nothing here to indicate the motivation for the modification.
"Overall performance"? .. less flare? Increased resolution?/ contrast? Perhaps, but reduction in weight, a reduction in cost to manufacture through the utilization of new glasses... could be equally important.

The reduction in manufacturing cost alone, by the substitution of single element for a cemented pair would/ should be substantial.

 

[Ed Sukach]

Ed, IMHO ...

Let me emphasize: I respect your opinion ... you are perfectly entitled to explain your reasoning, defend your point of view. Much of what we learn "comes hard" and it is only natural that we become obsessed with our points of view.

....it is inappropriate to speak of reflections deep within the barrel in the same context as reflections that you can get between the filter and front element. The front element sees way more light than actually forms the image.... that's why we hood...

"Way" more? Certainly, a greater amount. There will be additional chance for strange off-axis reflections from the additional glass surfaces to cause flare ... but VERY little from geometrically parallel glass surfaces - remember that THICK, highly refractive fused quartz optical flat introduction - and the net result of *NO* discernable effect at all - nada!

People just need to do what works for them, and if they do offer an opinion, they should back it up as transparently as possible! I have stated my reasoning very clearly ( I think): a filter, no matter how multicoated and German, introduces two extra surfaces in front of the front element. That's it! Either that means something to you or it doesn't!

It does! Apparently, not quite the same meaning or with the same intensity as it does to you.

... And you probably don't want to see a detailed ray diagram to demonstrate what I am saying, so let's just let it go ...

I hope you realize that you have just thrown me into my own private Brier patch. I am moderately familiar with Ray Traces and their role in lens designs of the past. Ah... fond memories of Snell's Law... I might just have a Slip Stick (remember those?) around here somewhere ... I DO have a 15 place Curta.

It would be a telling Ray Trace... parallel rays reaching a flat, refractive plane ... NO refraction at 90 degrees to the optical axis, either entering or leaving.

Part (b) of this logic is that filters are not considered in the lens design... they could not be! That's because the thickness of the lens, the dispersion of the glass, the spacing form the front element etc. are all variables that you'd need to know in order to do it right. And lo! The filters are -flat!-... not curved like the front element, so that calculation would be really dicey.

Dicey? Not really. Just enter neutral values into Snells.... as simple as anything I can imagine.

Personally I don't understand the fear of using superfantastic lenses unfiltered because of what may happen if they are careless. That which you are afraid to use as the lens designers intended should be at home in the velvet box. Sorry to be blunt, but really.... (here we go! And this isn't without ample forewarning!)....
It doesn't cost anything to be careful.

Afraid? Me? After all these years of working with Intralaboratory Standards? No thanks. I don't think I'll start being afraid at this point in my life.

 

[keithwms]

Where is that ignore button? Oh yes, there...

 

[Q.G.]

Ed, IMHO it is inappropriate to speak of reflections deep within the barrel in the same context as reflections that you can get between the filter and front element. The front element sees way more light than actually forms the image.... that's why we hood...

Well ... it's not really inappropriate at all.
Every extra glass surface offers an opportunity to reflect. And light bounces back and forth between all surfaces.

If all we had was one filter, no elements behind it, the problem of flare would be greatly reduced.
Put a sinlge element behinfd it, and the number of reflections increases very rapidly.
Put yet another lement behind that, and again, an even greater increase in flare causing reflections.
Etcetera.
You do have to consider all surfaces, not just concentrate on the thing that happens to be in front.

Anyway, there are good reasons why doublets are cemented and not air spaced.... but this is really beside the point. Again, the stray light at (and beyond) the front element is the issue here.

Nope.
Image formation, not flare reducation dictates the distance between elements.

People just need to do what works for them, and if they do offer an opinion, they should back it up as transparently as possible! I have stated my reasoning very clearly ( I think): a filter, no matter how multicoated and German, introduces two extra surfaces in front of the front element. That's it!

And you're perfectly right.

But the effect of that can be exagerated, can't it?

[...] Part (b) of this logic is that filters are not considered in the lens design... they could not be! That's because the thickness of the lens, the dispersion of the glass, the spacing form the front element etc. are all variables that you'd need to know in order to do it right. And lo! The filters are -flat!-... not curved like the front element, so that calculation would be really dicey.

The effect of flats in front of the lens, i.e. where incident pencils are rather parallel, is very, very, very small. As long as all light that hits the flat has (aqpproximately) the same direction, the results is really next to nothing. A very slight focussing difference, easily compensated for by focusing.

There is a difference between what a flat does to rays parallel to the optical axis and oblique rays. But that ony begins to become detectable at rather small lens to subject distances.

In general, there is nothing to worry about.

Now, there are examples of filters that are considered within the overall lens design: for example, the lenses with slip-in filter compartments within the barrel. You get a set of filters with the lens. But again, the front element is what sees a lot of extra stray, non-image-forming light.... etc.

And those filters are indeed critical. They play a part in how light is guided through the lens, i.e. where the angles often are considerable.
A different proposition, compared to filters in front of the lens.

There is one recent notable exception to this rule that lens designers don't like to see filters in their lenses. The leica m8 has the issue that they couldn't place a [IR-reflecting] hot mirror between the sensor and the rear element. so for that one camera, one is obligated to put a hot mirror over the front element to avoid purply blacks.... just because the digital sensor has extended IR sensitivity. Now, I don't know about you, but I'd be hopping mad if somebody told me I had to use a hot mirror over my front element. But anyway, that is the only example of which I am am aware that the factory actually suggested that a front filter could improve the image. And clearly it is a major compromise forced by the camera design itself.

Which (again) can be grossly exagerated.

It doesn't cost anything to be careful.

Yes.

Remember, though, that you only need to be carefull if there is a real risk.
Take a moment to think about that.

And if there is a real risk, there is a real risk.
And then: it costs less to replace a filter than to have a front element replaced.

 

[keithwms]

:rolleyes: Some progress has been made after all these pages!

Those who advocate routine use of the 'protective' filters are now apparently agreeing that they also need to use hoods and pricey multicoated filters.

Now, you'll need a pricey mc filter for each and every lens, unless you're willing to screw and unscrew them every time you switch lenses. I did some quick math and it'd cost me about $5000 to put b+w filters on my lens collection..... Considering that even after dropping a few lenses in such a way that the filter threads were bent but have still never managed to actually damage a front element... and I and many others get along just fine with lenses several $k in original value with slightly marked front elements that still perform excellently.... ach never mind

I'd like to take one last opportunity to reject the notion of normal incidence at the front element. Solid angle. Hood. But we can probably go back and forth on the optics ad nauseum and never really touch on the question of whether a front filter has protective worth because of the many non-optical factors at play. Many of those factors are quite individual.

For those interested in exploring the optics further, I would suggest consulting Mr. Fresnel, who has much more to say about all this than Mr. Snellius. Just please don't set theta_i to 90 degrees and say I told you so.

In summary, it seems that we've established that...

(a) people are going to do whatever they like anyway, and

(b) no amount of optics nor logic is going to affect decisions that have already been made...

 

[Q.G.]

You left out:
(c) some people see problems were there are none.

Now which side of the argument is that supporting?
Hmm ...

Maybe both?