UV Filters Useless or Worse?

Gearsau · Jul 10, 2018

Leonard Shepherd said:
Gearsau said:

Agree. I do have the Nikon NC filters on my lenses , as, sometimes, I would have been photographing in a dusty environment . Saves the front element.

Click to expand...

What do you mean by "saves the front element"?

Dust is easily removed with a blower brush.

Many lens have fluorine coatings to help remove dust, rain and smudge removal

Leonard Shepherd said:

My Nikon 300mm f2.8, Nikon 500mm f4 and Nikon 800 mm f5.6 have a clear glass filter on them. Factory fitted.

Click to expand...

Perhaps you do not fully understand.

The lenses you refer to have very easily damaged soft ED or FL elements directly behind the front meniscus glass.

They are likely behind a meniscus because Nikon regards ED or FL as being unsuitable as a front glass.

A meniscus glass is slightly curved to reduce the risk of reflection back onto the sensor and a double image from the back of a flat front glass.

There is only one curved meniscus filter I have used, being the removable front glass for the 200-400 VR series.

SoupOrPhoto · Jul 10, 2018

Why Toi said:
SoupOrPhoto said:

*I agree with everyone who said that they would use a clear filter in bad environmental conditions (salt or sand spray, e.g.).

Click to expand...

Without a discernable effect with a quality filter and not knowing when I would encounter sand spray when out, I just get lazy and leave my B+H/Hoya filter on my lenses. So far so good.

I certainly understand that. Just watch out for reflections off of the sensor—especially shooting dark scenes with a bright light. This a discernible effect.

This is why Nikon makes a meniscus “filter” for the few lenses that need them (lenses with front elements made with more fragile glass, according to Leonard Shepherd, above).

Leonard Shepherd · Jul 10, 2018

SoupOrPhoto said:
This is why Nikon makes a meniscus “filter” for the few lenses that need them (lenses with front elements made with more fragile glass, according to Leonard Shepherd, above).

To add a bit more detail Nikon have supplied some lenses with a front meniscus for at least 15 years. From memory Canon introduced front meniscus about 5 years earlier.

In the background although multi coating works, it only works over a very narrow angle relative to the surface of the glass.

It follow that when a bright highlight is reflected off a sensor it normally passes safely out of the front of a lens by passing through the curves of individual lens elements.

If the sensor reflected highlight detail instead hits the back of a flat glass rear surface of a flat glass filter added to the front of a lens it is more likely to get reflected back toward the sensor creating a double image.

Nano coating is effective over a much wider angle of incidence than multi coating.

The lenses we are now discussing have light travelling parallel to the optical axis after passing through the rear "light bending" element in the rear of these lenses. I presume this is why the flat glass rear slot in filters work OK without nano coating.

--
Leonard Shepherd
In lots of ways good photography is much more about how equipment is used rather than the equipment being used.

SoupOrPhoto · Jul 10, 2018

Leonard Shepherd said:
SoupOrPhoto said:

This is why Nikon makes a meniscus “filter” for the few lenses that need them (lenses with front elements made with more fragile glass, according to Leonard Shepherd, above).

Click to expand...

To add a bit more detail Nikon have supplied some lenses with a front meniscus for at least 15 years. From memory Canon introduced front meniscus about 5 years earlier.

In the background although multi coating works, it only works over a very narrow angle relative to the surface of the glass.

It follow that when a bright highlight is reflected off a sensor it normally passes safely out of the front of a lens by passing through the curves of individual lens elements.

If the sensor reflected highlight detail instead hits the back of a flat glass rear surface of a flat glass filter added to the front of a lens it is more likely to get reflected back toward the sensor creating a double image.

Nano coating is effective over a much wider angle of incidence than multi coating.

The lenses we are now discussing have light travelling parallel to the optical axis after passing through the rear "light bending" element in the rear of these lenses. I presume this is why the flat glass rear slot in filters work OK without nano coating.

Leonard Shepherd · Jul 10, 2018

SoupOrPhoto said:
I have not had this type of lens, but I cannot think of a reason to use a clear filter in the slot.

Slot in rear filters supplied with Nikon lenses have been NC in the 19 years I have used Nikon.

One reason might be NC in Nikon's view is better than UV ;-)

The rear light bending element of this type od lens is usually a few inches in front of the lens bayonet, and is less prone to dust with a slot in filter behind.

Exchanging the slot in NC for a rotating Pol seems to be the only way to get a Pol effect with these lenses.

Leonard Migliore · Jul 10, 2018

SoupOrPhoto said:
Interesting, thanks. I had not thought about rear filters. I have not had this type of lens, but I cannot think of a reason to use a clear filter in the slot. (I have not noticed reflections with my high-quality CPL--coatings?)

The rear filter is in an area where the light is converging to the focus. The lens is designed to have a certain thickness of glass (the filter) there. If the filter is not there, there will be a change in the focal point and also in the spherical aberration. You would probably only notice this at wide apertures but if you bought a 300mm f/2.8 you really want it to be sharp at f/2.8.

SoupOrPhoto said:
I don't understand your last paragraph. It seems like the problem would be worse with rear filters because there are more angles of reflection off of the sensor that could reflect from the filter and hit the sensor again (because it is closer). What am I missing?

When you have a filter in front of a lens, the lens acts as a collimator for light coming off the sensor; it concentrates the light and puts it back on the sensor. If the filter is between the lens and the sensor, the reflected light bounces away (diagrams would help but I don't have any right now).

SoupOrPhoto · Jul 10, 2018

Leonard Migliore said:
SoupOrPhoto said:

Interesting, thanks. I had not thought about rear filters. I have not had this type of lens, but I cannot think of a reason to use a clear filter in the slot. (I have not noticed reflections with my high-quality CPL--coatings?)

Click to expand...

The rear filter is in an area where the light is converging to the focus. The lens is designed to have a certain thickness of glass (the filter) there. If the filter is not there, there will be a change in the focal point and also in the spherical aberration. You would probably only notice this at wide apertures but if you bought a 300mm f/2.8 you really want it to be sharp at f/2.8.

Oh, so you do use a clear filter for these lenses. Makes sense.

Leonard Migliore said:
Leonard Migliore said:

I don't understand your last paragraph. It seems like the problem would be worse with rear filters because there are more angles of reflection off of the sensor that could reflect from the filter and hit the sensor again (because it is closer). What am I missing?

Click to expand...

When you have a filter in front of a lens, the lens acts as a collimator for light coming off the sensor; it concentrates the light and puts it back on the sensor. If the filter is between the lens and the sensor, the reflected light bounces away (diagrams would help but I don't have any right now).

Michael Benveniste · Jul 10, 2018

Leonard Migliore said:
The rear filter is in an area where the light is converging to the focus. The lens is designed to have a certain thickness of glass (the filter) there. If the filter is not there, there will be a change in the focal point and also in the spherical aberration. You would probably only notice this at wide apertures but if you bought a 300mm f/2.8 you really want it to be sharp at f/2.8.

Quite a few of the older Nikkors offered the option of a gelatin holder in lieu of a drop-in filter. My old 300mm f/4 AF is one such lens. I also have the drop-in polarizer for that lens, and the glass in that filter is thicker than the L39c which came with the lens. AFAIK, Nikon never made an NC filter for the 39mm filter holders.

A few years ago, I tested that lens with and without the filter in place. I detected no difference in sharpness or in spherical aberration. I did notice a slight change in displayed and recorded focus distance. ImageJ showed that a loss of about 1.2% of transmitted light, and a color shift in the 25-50 degree range. That's consistent with what I've measured with a B+W 010 MRC UV filter. Nor would I expect a flat piece of glass to cause any changes to spherical aberration.

Nikon does note that when using a drop-in polarizer, "The distance scale is shfited from the correct position. The closest focused distance is extended slightly. The memory set position may change slightly when using focus preset." All of this is consistent with a thicker filter. Without a filter, the change is in the opposite direction.

Leonard Migliore said:
When you have a filter in front of a lens, the lens acts as a collimator for light coming off the sensor; it concentrates the light and puts it back on the sensor. If the filter is between the lens and the sensor, the reflected light bounces away (diagrams would help but I don't have any right now).

Unfortunately, the light can get reflected back onto the sensor either way. It's still a flat piece of "stuff" that's parallel with the sensor. But the drop-in filter is smaller and the light from the sensor is hitting it at less of an angle so it's less likely to cause problems.

--
Light travels at 2.13085531 × 10^14 smoots per fortnight. Catch some today!

Why Toi · Jul 10, 2018

Leonard Migliore said:
When you have a filter in front of a lens, the lens acts as a collimator for light coming off the sensor; it concentrates the light and puts it back on the sensor. If the filter is between the lens and the sensor, the reflected light bounces away (diagrams would help but I don't have any right now).

Sure about this explanation? The flat glass at the front does not bounce light straight back to the sensor as light rays coming in and out are actually on 'bent' paths, unless it's approaching a 0 degree FOV. For light bounced back from the sensor, it'll exit (except on axis) will incident the flat front filter at an angle, expanding to an angle that matches the FOV of the lens eg. Wide angle. As such, a tiny percent of that ray (subject to coating efficiency (will reflect off at an equal angle. Hardly the straight on straight off analogy here.

ARClark · Jul 10, 2018

I do nature photography, and I've had a couple lenses saved from damage by “useless” filters. Frankly, I don’t use them for IQ, I use them for protection.

Alan

Leonard Migliore · Jul 10, 2018

Why Toi said:
Leonard Migliore said:

When you have a filter in front of a lens, the lens acts as a collimator for light coming off the sensor; it concentrates the light and puts it back on the sensor. If the filter is between the lens and the sensor, the reflected light bounces away (diagrams would help but I don't have any right now).

Click to expand...

Sure about this explanation? The flat glass at the front does not bounce light straight back to the sensor as light rays coming in and out are actually on 'bent' paths, unless it's approaching a 0 degree FOV. For light bounced back from the sensor, it'll exit (except on axis) will incident the flat front filter at an angle, expanding to an angle that matches the FOV of the lens eg. Wide angle. As such, a tiny percent of that ray (subject to coating efficiency (will reflect off at an equal angle. Hardly the straight on straight off analogy here.

The trick is, you have a lens in the way. A lens is a device that transforms angular deviation into lateral deviation. Here is a sketch:

The red rays entering from the left parallel to the optical axis are all sent to the optical axis on the focal plane. Their angular deviation of zero is transformed to a position on the optical axis.

The blue rays entering from the left at a 5 degree angle to the optical axis are all sent to a point f*tan 5 from the optical axis. That's what a rectilinear lens does.

Now, what happens if some light bounces off the sensor? It's a flat mirror so the reflection is a diverging beam at an angle of 5 degrees, like this:

The lens accepts this diverging beam and transforms it to a group of rays parallel to the incoming beam because that's what a lens does.

Now let's see what happens if there's a filter in front of the lens:

The much-weakened beam is directed back through the lens but since it's a parallel beam, it's focused on the sensor. The location of the focal point is directly opposite the initial focus. Since the light is concentrated by being focused, it is often visible if the initial light came from a strong source. You see these all the time when people put filters on their lenses and shoot at lights.

--
Leonard Migliore

Why Toi · Jul 10, 2018

Leonard Migliore said:
The trick is, you have a lens in the way. A lens is a device that transforms angular deviation into lateral deviation. Here is a sketch:

The red rays entering from the left parallel to the optical axis are all sent to the optical axis on the focal plane. Their angular deviation of zero is transformed to a position on the optical axis.

The blue rays entering from the left at a 5 degree angle to the optical axis are all sent to a point f*tan 5 from the optical axis. That's what a rectilinear lens does.

Now, what happens if some light bounces off the sensor? It's a flat mirror so the reflection is a diverging beam at an angle of 5 degrees, like this:

The lens accepts this diverging beam and transforms it to a group of rays parallel to the incoming beam because that's what a lens does.

Now let's see what happens if there's a filter in front of the lens:

The much-weakened beam is directed back through the lens but since it's a parallel beam, it's focused on the sensor. The location of the focal point is directly opposite the initial focus. Since the light is concentrated by being focused, it is often visible if the initial light came from a strong source. You see these all the time when people put filters on their lenses and shoot at lights.

Anyway, I fully appreciate that the sensor will bounce a small percentage of light but the scale (for a high quality filter) is hardly that relevant given the said ray will need to pass through the double digit glass/air interfaces within our lenses before reaching the front filter. Seriously, when the light source is that strong, I'd be more concerned with the flares and contrast loss it induces as well as the potential burn to the sensor than what the front filter is doing. Oh wait, or I can just take that front filter off in those rare situations... :-D

lb77 · Jul 11, 2018

I use protective filters on most of my lenses when not using polarisers. As to UV filters i see no perceivable difference in IQ with them and without them except in very specific conditions like ghosting of strong lights at night etc.

I feel more comfortable cleaning the filters instead of a very expensive lens.

SoupOrPhoto · Jul 11, 2018

Why Toi said:
Leonard Migliore said:

The trick is, you have a lens in the way. A lens is a device that transforms angular deviation into lateral deviation. Here is a sketch:

The red rays entering from the left parallel to the optical axis are all sent to the optical axis on the focal plane. Their angular deviation of zero is transformed to a position on the optical axis.

The blue rays entering from the left at a 5 degree angle to the optical axis are all sent to a point f*tan 5 from the optical axis. That's what a rectilinear lens does.

Now, what happens if some light bounces off the sensor? It's a flat mirror so the reflection is a diverging beam at an angle of 5 degrees, like this:

The lens accepts this diverging beam and transforms it to a group of rays parallel to the incoming beam because that's what a lens does.

Now let's see what happens if there's a filter in front of the lens:

The much-weakened beam is directed back through the lens but since it's a parallel beam, it's focused on the sensor. The location of the focal point is directly opposite the initial focus. Since the light is concentrated by being focused, it is often visible if the initial light came from a strong source. You see these all the time when people put filters on their lenses and shoot at lights.

Click to expand...

Anyway, I fully appreciate that the sensor will bounce a small percentage of light but the scale (for a high quality filter) is hardly that relevant given the said ray will need to pass through the double digit glass/air interfaces within our lenses before reaching the front filter. Seriously, when the light source is that strong, I'd be more concerned with the flares and contrast loss it induces as well as the potential burn to the sensor than what the front filter is doing. Oh wait, or I can just take that front filter off in those rare situations... :-D

I was unable to get a picture at night with the moon in it without taking off the UV filter on the lens. The effect was also noticeable in a dark indoor scene with a bright window in the background. It worked exactly the way Leonard described.

These are not the most common shooting situations, granted, but the extreme situations highlight a potential pitfall of the flat, clear glass. I suggest that you try it and make sure that your filters are not causing unwanted effects. Admittedly, the UV filters that I used to use were not the best quality.

SoupOrPhoto · Jul 11, 2018

lb77 said:
I use protective filters on most of my lenses when not using polarisers. As to UV filters i see no perceivable difference in IQ with them and without them except in very specific conditions like ghosting of strong lights at night etc.

I feel more comfortable cleaning the filters instead of a very expensive lens.

That ghosting--light reflecting off of the sensor then being reflected back by the filter--is exactly what made me stop using clear filters. I figure that it is most noticeable in these extreme situations but must also be happening with standard use. It may be less noticeable with common scenes but it is clearly something that happens. Nikon even warns about this (see post above).

I have not scratched a lens in years (since giving up clear filters). Lucky, maybe. And the lenses are easier to clean than most of my filters.

Why Toi · Jul 11, 2018

May I ask what kind of filter and sensor you were using? Have not encountered such problems here.

SoupOrPhoto · Jul 11, 2018

Why Toi said:
May I ask what kind of filter and sensor you were using? Have not encountered such problems here.

Both a D90 and a D750 with various mid-range filters (not the cheapest, not good ones--Tiffen and Hoya, I think). I have tested and not had the issue with the B+W Kaesemann CPL.

Why Toi · Jul 11, 2018

SoupOrPhoto said:
Both a D90 and a D750 with various mid-range filters (not the cheapest, not good ones--Tiffen and Hoya, I think). I have tested and not had the issue with the B+W Kaesemann CPL.

Maybe that's the differentiation as I have only used mid-upper end filters. As we all can agree, the specific coating can make a difference. Further and as I understand it, the more budget end of the models may only have coating (or inferior coating) on the object side ie. Often without coating on the sensor side.

So I think as suggested by many here, the quality of the filter used matters in this discussion.

gkal · Jul 12, 2018

Leonard Shepherd said:
SoupOrPhoto said:

On a recent thread, the OP discovered a UV/protective filter on the 200-500 causing mischief in images.

I have had this experience a number of times and have come to think that the filters are, at best, useless.

Click to expand...

I have only had one incident where a "cheap" filter very seriously degraded image quality.

An inferior filter can degrade image quality with any lens.

From the retailers perspective the profit margin an a filter can be 400% ;-)

historianx · Jul 12, 2018

Why Toi said:
May I ask what kind of filter and sensor you were using? Have not encountered such problems here.

Yes I am curious as well. I also use mid-upper end filters and have no issues with them, certainly no image degradation as you would get with merde cheapo "filters."

UV Filters Useless or Worse?

Senior Member

Senior Member

Forum Pro

Senior Member

Forum Pro

Forum Pro

Senior Member

Veteran Member

Forum Enthusiast

Veteran Member

Forum Pro

Forum Enthusiast

Forum Enthusiast

Senior Member

Senior Member

Forum Enthusiast

Senior Member

Forum Enthusiast

Well-known member

historianx

Guest

Keyboard shortcuts