[lotech]

Hi there,

We used to put UV or 1B filter on film camera to block unwanted UV light, AFAIK UV is quite easy to block even with standard glasses used on windows, so that we will not get tanned indoor, and also can not erase EPROM behind windows (if you are in electronics field you know it), but I don't know which part in UV spectrum causing the bluish cast on film, and how digital sensor responds to it, wonder if UV filter still needed on digital camera.

[DeadJohn]

No, a UV filter is not needed because digital sensors generally have very low sensitivity to UV.


You'll read many opinions on UV filters. I feel that using one when not needed adds an extra piece of (often cheap) glass, which creates more risk of flare and low contrast.

Some people insist UV filters will protect the lens from damage. I think the chance of it helping is very small. Modern glass coatings are very resistant to damage. For me, a filter for protection only makes sense when taking photos in a hazardous environment such as salt spray, sandstorms, or spray paint.


(Use a lens hood. They also offer minor protection, but improve image quality while doing so.)

[BigDave]

Most of the effect for the UV filter was to protect the lens from debris. It helped a little with cutting off some haze, but a polarizer was always better. The near UV (360-410nm) would contribute to the blue cast. Now there is filtering on both ends (UV and IR) on the sensor. So the filters are not needed, but can still be useful for protection in some instances. Below is a typical CMOS sensor spectral response. We humans see in the 400-700 range +/- a little).

[lotech]

Thanks guy for the info., yes I used to put a UV or 1B on the lens simply for protection from physical damage to the lens, and I always use metal hood to protect the lens when bumping around in crowded places, also blocking my hands from touching the front glass and smudged it.

[Alex645]

Hi there,

We used to put UV or 1B filter on film camera to block unwanted UV light, AFAIK UV is quite easy to block even with standard glasses used on windows, so that we will not get tanned indoor, and also can not erase EPROM behind windows (if you are in electronics field you know it), but I don't know which part in UV spectrum causing the bluish cast on film, and how digital sensor responds to it, wonder if UV filter still needed on digital camera.

I concur with @DeadJohn & @BigDave. Unlike film, UV is a nonissue for digital sensors. I also agree that a lens hood is preferable for protection BUT IF you don't have a hood, I think a multicoated clear filter is wise. From working with hundreds of students per year, I would estimate about 2-3% would have incurred damage to their front element if they didn't have a filter on their lens. The other 97-98% would deem it unnecessary although as a prime shooter without WR protection, that filter does help a bit.

What you mention about window glass is true for about 98% of UV, but not 100%. That 2% can cumulatively damage skin and was significant enough for film. Same with photos behind picture frame glass. Standard glass, although it blocks most of the UV, will not protect images from fading prematurely.

Bottom line: UV is not needed on a digital camera for its intended purpose.

[hypermodern]

Looking at BigDave’s spectra, we could use a violet enhancer in to 400-420nm range!

[clickclick]

Fully acknowledge there's no need like on film, but shooting around water with and without, I still feel a high quality UV filter can help with haze. I feel high quality is key to keep from degrading the image, and shooting into light is always problematic by adding another reflective surface, so something along the lines of B+W is the way I'd go. I've also played around without having a UV filter on my lens from a protection standpoint. I do a lot of walking around outdoors. I can't always blow the front clean. I'd rather be wiping a filter than the front of my lenses, especially given the costs of some of the glass involved. I see too many used lenses with "cleaning marks" for sale to say there's no advantage to protecting the front element. Yeah, in a perfect world, but I don't seem to take photos in a perfect world. Just my .02 and worth what you paid for it.

[BigDave]

Looking at BigDave’s spectra, we could use a violet enhancer in to 400-420nm range!

And the graph also explains how Reds can get over saturated too! Maybe some day there will be a sensor that provides even, or neat even, spectral response. Hey SONY, how about it!

[clickclick]

Not to get this topic too sideways, but I've been intrigued by some of the special filters out there. The light pollution filters are one such that have caught my attention:

Light Pollution Reduction Filters | B&H Photo Video

[BigDave]

I really love the marketing these companies do. I am not sure how a magenta filer (as it seems) which reduces/neutralizes green in an image, has an overall effect on "light pollution". Most city lights have a great deal of yellow (some effect) and now more commonly Blue (a Yellow filer would be needed) as the predominant color in the source. Just enough red and green so we see white.

Here is a good link to the testing of said filters. Color balances change, yes, but not sure if any of them "reduce" the overspill of the city lights...

9 Light Pollution Filters Tested: Do They Really Work? Part 1 | B&H Explora

regards,

[RGlasel]

Maybe some day there will be a sensor that provides even, or neat even, spectral response. Hey SONY, how about it!

If only it were that simple. Our natural spectral sensitivity is not even, not even even across different levels of intensity. Add to that complex formula the built-in NI (natural intelligence) most of us have and not only would a sensor with even spectral response "see" different images than we do with our stereoscopic eye-cameras, but as we age we learn to post-process man-made images in our head to match what our brains remember scenes look like in real life and images from a more "natural" sensor will look artificial, until we convince our brains otherwise. The same thing could happen if someone marketed a sensor with even more uneven spectral sensitivity, all it would take is a few Youtube influencers to tell us how our old images are not as good as the new images from Brand X cameras.

[hypermodern]

If only it were that simple. Our natural spectral sensitivity is not even, not even even across different levels of intensity. Add to that complex formula the built-in NI (natural intelligence) most of us have and not only would a sensor with even spectral response "see" different images than we do with our stereoscopic eye-cameras, but as we age we learn to post-process man-made images in our head to match what our brains remember scenes look like in real life and images from a more "natural" sensor will look artificial, until we convince our brains otherwise. The same thing could happen if someone marketed a sensor with even more uneven spectral sensitivity, all it would take is a few Youtube influencers to tell us how our old images are not as good as the new images from Brand X cameras.

And then you would have to find a way to output the image. There are a lot of greens that even ProPhoto RGB can't output.

[BigDave]

Our natural spectral sensitivity is not even, not even even across different levels of intensity.

Very true. the photopic response (what we humans see) is skewed to be sensitive to green mostly (for normal color vision) and most sensors would try to mimic that, I am sure. Based on the graph I presented, green is a predominant sensitivity, but WOW, Red really overshadows the green! Maybe we need a cyan filter in there, to reduce the red and bias the sensor toward the blue some!

[clickclick]

Color balances change, yes, but not sure if any of them "reduce" the overspill of the city lights...

I'd love to have a candid sit down with lead filter designers from someone like B+W and find out how much of these things are marketing versus believing they really can make a difference. It would seem that regardless of the part of the light spectrum being affected, it's going to be brighter where there's a visible light source.

[DeadJohn]

Fully acknowledge there's no need like on film, but shooting around water with and without, I still feel a high quality UV filter can help with haze. I feel high quality is key to keep from degrading the image, and shooting into light is always problematic by adding another reflective surface, so something along the lines of B+W is the way I'd go. I've also played around without having a UV filter on my lens from a protection standpoint. I do a lot of walking around outdoors. I can't always blow the front clean. I'd rather be wiping a filter than the front of my lenses, especially given the costs of some of the glass involved. I see too many used lenses with "cleaning marks" for sale to say there's no advantage to protecting the front element. Yeah, in a perfect world, but I don't seem to take photos in a perfect world. Just my .02 and worth what you paid for it.

For lessening haze, a CPL filter is more effective than a UV filter.

A little bit of dust and smudges on the front of a lens are invisible in photos. Cleaning lenses or filters too often is a common way things get damaged. I don't know about your outdoor conditions, though, and maybe your climate has a lot of sticky dust that needs frequent cleaning.