View Full Version : Does a filter degrade the image? You decide!
StewartR
27-07-2008, 14:41
We seem to have had quite a few "what filter" or "should I or shouldn't I filter" threads here recently. Of course a protective filter will degrade an image to some extent, but the key issue is whether it's noticeable. Some people claim that they can tell the difference; some claim that the difference is obvious. But I don't think we've seen any with-and-without comparisons.
So here's one.
I took a shot of part of my garden, using:
* my trusty old Canon 350D
* a brand new Canon EF-S 10-22mm f/3.5-4.5 USM
* a brand new Hoya Pro 1 Digital protector filter
and then I took the same shot without the filter. Or maybe it was the other way round?
Both were shot in RAW and processed identically - the same RAW conversion parameters, then the same levels adjustments, USM sharpening, etc. As well as the whole frame, I've prepared a coupe of 100% crops - again processed identically.
In the pictures below, numbers 1B and 1C are crops from 1A. Similarly, 2B and 2C are crops from 2A. So your challenge is to decide:
Were 1A/1B/1C taken with the filter, and 2A/2B/2C without the filter?
Or perhaps 2A/2B/2C were taken with the filter, and 1A/1B/1C without?
Can you tell the difference?
1A. Full frame. With filter, or without?
http://StewartR.smugmug.com/photos/339645649_sDoLQ-L.jpg
1B. Full frame. Without filter, or with?
http://StewartR.smugmug.com/photos/339645951_3Vf3J-L.jpg
2A. 100% crop. With filter, or without?
http://StewartR.smugmug.com/photos/339645832_ET6D8-L.jpg
2B. 100% crop. Without filter, or with?
http://StewartR.smugmug.com/photos/339646121_kAZBb-L.jpg
3A. 100% crop. With filter, or without?
http://StewartR.smugmug.com/photos/339645740_bo9Fg-L.jpg
3B. 100% crop. Without filter, or with?
http://StewartR.smugmug.com/photos/339646025_pXz3j-L.jpg
So tell me what you think. Which is which, and why? Or can you not see any difference?
Well I cant see any difference... and if there IS any difference I think Id rather have my lens protected than worry about that tiny amount. :thumbs:
I agree with Janice, :thumbs: buggered if I know :shrug:
Going by physics it Must degrade the image but in real life you probably wont notice unless it's a pants filter :)
My puter is pulling the web through my mobile at the mo, compressing all the images to state near death but I'll plump for set one without and two being with the filter.
Not a fan of them myself, although in that shot I'd not really expect to lose or gain an awful lot either way. Bring the sun round to fall across the lens and then perhaps my risk might pay off. :)
strobemonkey
27-07-2008, 15:08
Difficult to judge because of your subjects. Do it on a large brick wall please.
My guess:
1B, 2B, 3B are with filter
In the pictures below, numbers 1B and 1C are crops from 1A. Similarly, 2B and 2C are crops from 2A. So your challenge is to decide:
Were 1A/1B/1C taken with the filter, and 2A/2B/2C without the filter?
Or perhaps 2A/2B/2C were taken with the filter, and 1A/1B/1C without?
Can you tell the difference?
Which is which, and why? Or can you not see any difference?
I don't get your file numbering....
Veracocha
27-07-2008, 15:19
All A without?
StewartR
27-07-2008, 15:25
Difficult to judge because of your subjects. Do it on a large brick wall please. I would respectfully suggest that it you need photos of large brick walls to be able to tell the difference, then the difference is too small to worry about.
But if you think there's a different type of everyday subject which would accentuate the difference, just name it. I'd be happy to consider taking another couple of shots.
strobemonkey
27-07-2008, 15:36
I would respectfully suggest that it you need photos of large brick walls to be able to tell the difference, then the difference is too small to worry about.
I think that's true.
But if you think there's a different type of everyday subject which would accentuate the difference, just name it. I'd be happy to consider taking another couple of shots.
How about macro shots with a macro lens???
Try shooting into the sun or with the sun in the periphery, or take a night shot with point sources of light (such as street lights or car headlights) or a landscape shot with glistening water, or fireworks & bonfires, I'm sure we'll all be able to see a difference then :)
Not a clue which is with and which without, but A's are slightly darker and richer coloured than B's. The shadows on the grey bowl also look slightly crisper.
StewartR
27-07-2008, 17:48
So let's have a look at the scores on the doors so far.
Janice ... can't tell
Cobra ... can't tell
mrgrubby ... can't tell
dazzajl ... A=without, B=with (but probably guessing)
strobemonkey ... A=without, B=with, but would prefer shots of a brick wall
Veracocha ... A=without, B=with
sprog ... probably can't tell
Kerioak ... A's look better than B's
Any more takers? I'm inclined to let it run a little longer before revealing the answer. And I'd really like to see some of the "difference is obvious" brigade have a go at it...
1B. 2B, 3B with filter i think, they are ever so slighty softer at 100% crop.
Bugger all in it ***
dazzajl ... A=without, B=with (but probably guessing)
Nope, right or wrong I'm happy to stand by that as a firm decision.... but it's not on resolution that's for sure. :D
For the life of me I can't tell which are with and which are without.
However, there is a very slight difference between the two (2A & 2B). If you look at the large leaves on the bottom of each picture you'll notice that those in 2A are darker than those in 2B! :shrug:
Still, both are of the same standard. So, whichever is with a filter - from these two - I wouldn't care!
Would be intersting to also know what type / brand of filter you've used; and if possible a comparison in a dark setting (say in doors, or at night) with a lot of side light(s) (like in a street) which can be problematic :shrug:
A looks darker than B, in the first image anyway.
I couldn't say which is running a filter :shrug:, but I'm gonna go with the A's since adding glass never made anything lighter.
The difference isn't worth a fart for this exercise.
Maybe with fullsize images, and double hard pixel peeping at atomic level, you might see a difference.
Defiance
27-07-2008, 18:43
Firstly, thank you Stewart for running this test as I was one of those who posted asking what was the best filter to use to protect my lenses.
As for the difference, I've looked at them, gone way, looked again, and I still can't tell the difference.
strobemonkey
27-07-2008, 18:48
A bit off topic for Stewart:
Could you please kindly do a test between EF 50 1.8 and an "L" zoom lens that cover 50mm.
Both at 50mm and same apperture, say f4 and f11. Remove exif data and we'll guess which image is from what lens...
agree with janice , but the As look very very slightly richer in colour.
I would respectfully suggest that it you need photos of large brick walls to be able to tell the difference
Dumb newbie question; why a brick wall? :thinking:
Dumb newbie question; why a brick wall? :thinking:
its a good way to test for Barrel / Pincussion, Vignetting and sharpness.
its a good way to test for Barrel / Pincussion, Vignetting and sharpness.
Thanks :thumbs:
I'm going for A is with and B is without.... Mainly to do with colour saturation - A is marginally more saturated... B less so. Can't notice difference in sharpness though as most parts of the image may or may not be moving due to the wind
Try shooting into the sun or with the sun in the periphery, or take a night shot with point sources of light (such as street lights or car headlights) or a landscape shot with glistening water, or fireworks & bonfires, I'm sure we'll all be able to see a difference then :)
Yep, this is what I was thinking as the filter obviously won't have the same coating(s) that the actual lens will have.
So let's have a look at the scores on the doors so far.
Janice ... can't tell
Cobra ... can't tell
mrgrubby ... can't tell
dazzajl ... A=without, B=with (but probably guessing)
strobemonkey ... A=without, B=with, but would prefer shots of a brick wall
Veracocha ... A=without, B=with
sprog ... probably can't tell
Kerioak ... A's look better than B's
Any more takers? I'm inclined to let it run a little longer before revealing the answer. And I'd really like to see some of the "difference is obvious" brigade have a go at it...
I'm with Kerioak, the A's have slightly deeper colours, have a look at the sky in 1 & 2.....
Steve
birdy961
28-07-2008, 12:22
Not a clue, cannot see any difference between them :shrug:
There's a minor difference in the colours and I think I can only see it because I'm deliberately looking for the difference in the two shots side by side (or one on top of the other). I have no idea which is with or without the filter though.
V8burble
28-07-2008, 12:36
For what it's worth, I would say A=with, but the difference in this situation is obviously too small to say for sure either way. :shrug:
CScottMcQueen
28-07-2008, 12:36
Im in the 'the colours seem better in the A's camp' especially on the the 1st and 3rd image.
I'm running a calibrated monitor if that makes any difference but for all that does my eyes are knackered :)
No idea, and there is such a minor difference in the colours that if others had not posted about it I would not have noticed them either!!
Les McLean
28-07-2008, 12:49
I think it quite often boils down to the quality of the filter, I've used (in the distant past) cheapo UV filters and experienced degradation, particularly flare in my images.
A good quality filter (as you have used), you would be hard pushed to tell any difference. (IMO)
If I was to guess I would say A had the filter.
However.... hardly any noticeable difference. So much so, that I can't really tell which has it or not, but a gut feeling goes with A. They just seem slightly darker
jerry12953
28-07-2008, 14:56
Well, come on then, Stewart....how long are you going to keep us in suspense for ? :bang: :nuts: :naughty:
I don't think we need to worry about which is which. What this just demonstrates is that there is hardly any noticeable difference to the images yet alone anything that indicates a drop in image quality. I imagine there are short falls in certain lighting etc but on the whole I'm impressed with the results.
Flash In The Pan
28-07-2008, 15:06
:bang:After scratching the front element on a £300 lens last year I put a quality UV one on every one of my lenses now :bang:
contrastingLight
28-07-2008, 15:19
Ill too go with A`s filtered and B without, the colour is slightly richer on the A`s and would suggest a UV filter to my untrained eye............:shrug:
When are we gonna find out the suspense is painful.
macky799
28-07-2008, 15:31
b with, a without :thumbs:
StewartR
29-07-2008, 07:59
OK, time to let the cat out of the bag.
1A/2A/3A are with the filter. 1B/2B/3B are without.
Some people got it right, some people got it wrong, and some couldn't see any differences. But I think we can all agree that any differences which do exist are entirely marginal.
The only difference that I can detect is that the colours are slightly more saturated in the A's, ie with the filter. But that could be an artefact of the processing. I processed the two images identically. If I had processed each one on its merits (assuming of course that I had the skill for that, which I don't!) then I expect this difference would disappear.
So where does this leave us? I think it's reasonable to conclude that, in decent light, if you use a good quality filter (this was a Hoya Pro 1 Digital), then your image quality will not suffer.
Is anyone up for doing another test with more challenging lighting?
Woohoo I was correct...
I think testing for flare would be a good idea. Maybe induce some and then remove the filter to see how it alters the effect.
Could be worth testing for vignetting on something wide like a 10-22 lens? (you previous pics may have answered this if they were at 10mm)
EdinburghGary
29-07-2008, 08:25
Stewart, would love to see more of the same - other types of shots, in bright sunlight etc.
Gary.
Even with the answer I still cannot see any perceivable difference :thumbs:
( please excuse my gramma she has not been well for sometime :D)
Can you tell the difference?
This is an old thread but I havn't read it before.
At this time of posting, I have not gone to the end to find the answer.
I think that "A" is without filter.
They sky is more saturated. There is more contrast in the foliage.
Andrew
This is an old thread but I havn't read it before.
At this time of posting, I have not gone to the end to find the answer.
I think that "A" is without filter.
They sky is more saturated. There is more contrast in the foliage.
Andrew
So now, having read the thread and found that I am wrong, I can only conclude:
I was expecting the filter to degrade the image. On the contrary, I believe that it enhances it. (in this case...)
Andrew
ccsharry
30-01-2009, 12:17
on 2A and 2B, you see more fringe defects on the 2B if you look at the high contrast edges... um... filter improves quality? interesting!!
DigitalRelish
30-01-2009, 12:27
It's worth pointing out that unless the testing environment is under full control then you've got other variables at play e.g. a darker leaf could be caused by the shadow of a bird overhead or a slight breeze altering the angle at which a leaf reflects the sun changing how bright and how saturated it looks.
I'm not taking any sides in the debate, but just saying that the differences that can be perceived may not be down to solely the addition of a filter.
I agree, it would be interesting to see a scene compared when you've got sunlight glistening off a reflective surface or other point sources of lights.
The sky at the top of the crop in 2A looks very slightly more grainy to me in comparison with 2B. Otherwise I can't notice any difference between the photos, maybe the As are fractionaly more saturated? :thinking:
I realised the old filters I was using did me no favours, bad flares with the cheap Hama filter on the Sigma lens, and after a quick test I was looking at least a stop, probably more with the filter on the 70-210. So on the basis of this test, I've bought Hoya Pro 1 Digital filters for all my lenses which arrived yesterday. As I shoot at rallies mostly, I'd rather chip a filter than the front element on a lens! ;)
Would be interesting to see some tests in difficult lighting or direct light, if I ever get the chance I might do a couple...
Mattfeliks
30-01-2009, 14:48
I think we need to see some more challenging lighting situations.
OK then - with more challenging lighting!
Marrakech, late at night with bright bulbs all over the place.
Spot the difference:
http://farm4.static.flickr.com/3125/2472102068_99e8181ee3.jpg
http://www.flickr.com/photos/dvf-woz/2472102068/sizes/o/
http://farm3.static.flickr.com/2025/2472103428_b6b37b4d62.jpg
http://www.flickr.com/photos/dvf-woz/2472103428/sizes/o/
;)
Excuse the slope - they were test shots taken after I noticed something fishy going on with my pictures that night.
Admittedly it's not a GREAT filter, but I decided to never use a protection filter again once I saw these two shots.
I'm really glad I noticed early in the night because I got some classic shots and they would have been ruined otherwise!
wontolla
30-01-2009, 15:18
Well I have certainly learned something from this thread! (I use a UV on all my lenses) BUT, if ever I want to take a picture of a brick wall, I will take it off!
Well I have certainly learned something from this thread! (I use a UV on all my lenses) BUT, if ever I want to take a picture of a brick wall, I will take it off!
You can see my pictures can't you? Not really a brick wall is it? :)
Both sets of shots are very interesting in their own ways. But if you have 4+ lenses that's around £200 worth of filters to kit them all out, and unless you are really clumsy you are unlikely to damage more than one front element (that's without taking into account front element damage which damages the lens as well).
So is the replacement of a front element cheaper than £200?
And from looking at Stewarts photos the filters appear to give a slight loss in contrast, but sharpness wise there is no difference IMO.
EDIT: Also the 350D is 8(?)MP, how about trying it on a 12,15,24mp camera, there may very well be a difference there. (not trying to be picky but after having a look at a few of the photozone retests with the 50D the change in resolution can make a difference.)
StewartR
01-02-2009, 08:05
OK then - with more challenging lighting!
Marrakech, late at night with bright bulbs all over the place.
Spot the difference: That's interesting.
But if it was a cheap filter, then all it goes to show is that cheap filters degrade the image, which I think we all knew anyway. If it was an expensive filter though...
StewartR
01-02-2009, 08:08
Also the 350D is 8(?)MP, how about trying it on a 12,15,24mp camera, there may very well be a difference there. (not trying to be picky but after having a look at a few of the photozone retests with the 50D the change in resolution can make a difference.)Yes, the 350D is 8MP.
I don't think there's much point trying a 24MP camera, because that will be full-frame and the pixel density will be therefore be pretty much the same as the 350D.
If we're concerned about sharpness, then re-testing with a 12MP or 15MP crop-sensor camera would be of interest, but I don't have any of them.
malla1962
01-02-2009, 09:22
Most of my lenses have filters on them and never had a problem with IQ.
i would'nt be without a filter..L series lenses are too expensive to risk any damage
Holden Caulfield
01-02-2009, 13:05
When I started working in the photographic trade years ago the buzz phrase then was ‘lines per millimetre’. A retailer in the south had the bright idea of using a device like a projector which when a lens was attached would project a wall size image of a test chart, he would then proceed to measure how may line per millimetre could be observed at different apertures.
He would then put a written test report in with the lens providing a so called value added extra.
For a while we had a parade of customers (usually students) who would ask how many LPM our lenses were and why could we not test them!!
One day one particularly obnoxious individual came in to the shop and demanded that we test the LPM on his camera lens.
I pointed out that ‘in the real world’ the lens would deliver far more quality than the film emulsion could resolve.
I suspect that the same is true of digital sensors.
I suspect that the same is true of digital sensors.
I don't think that is actually correct.
It has been muted that we are rapidly approaching/have reached a scenario where even top quality lenses are struggling to resolve as much detail as the higher MP sensors can resolve.
There have been a number of articles discussing the fact that DSLR sensors have all but reached their limit which can only be surpassed with an increase in lens quality.
swanseamale47
01-02-2009, 15:20
Both sets of shots are very interesting in their own ways. But if you have 4+ lenses that's around £200 worth of filters to kit them all out, and unless you are really clumsy you are unlikely to damage more than one front element (that's without taking into account front element damage which damages the lens as well).
So is the replacement of a front element cheaper than £200?
And from looking at Stewarts photos the filters appear to give a slight loss in contrast, but sharpness wise there is no difference IMO.
EDIT: Also the 350D is 8(?)MP, how about trying it on a 12,15,24mp camera, there may very well be a difference there. (not trying to be picky but after having a look at a few of the photozone retests with the 50D the change in resolution can make a difference.)
Let me see if I have this right. Lets all nip out and buy a 24mp camera to see if we can tell the difference with or without a filter by taking photos of a wall, something most of us will never do in real life anyway?
The test allready proved many cant see the difference, and that even when comparing imges next to each other, in the real world withoput the comparrison it would be pretty much impossible to know if a filter had been used. Wayne
Thanks for doing this Stuart, I use UV filters and have wondered wether it was affecting image quality. It looks like I would not be able to tell the difference
Pete
That's interesting.
But if it was a cheap filter, then all it goes to show is that cheap filters degrade the image, which I think we all knew anyway. If it was an expensive filter though...
Actually, Stewart, that's exactly what I was talking about when I asked if you could do your test in low light.
When I used my Nikon 24-70 AF-S with a Nikon NC filter (at night, with weird lighting conditions ... like Streets of London with the Christmas and New Years lights on) I would get these weird UFO light interferences. It was this which made me decide to ditch using filters
dalex_257
03-02-2009, 13:28
Admittedly it's not a GREAT filter, but I decided to never use a protection filter again once I saw these two shots.
I'm really glad I noticed early in the night because I got some classic shots and they would have been ruined otherwise!
I agree with you, very much having had the same experience in the night shots with my Sigma 10-20 with protective filter (Hoya Pro 1) - never used it since on that lens unless the lighting is sufficiently diffused. I was not as lucky as you and only discovered the nasty flares and haze post factum when it was too late.
My view on that, is that there is very little point to spend all those money we do on excellent glass to achive highest possible quality only to slap a piece of glass in front of it inhibiting that quality we sought. Each lens optics designed with certain performance goals and the lens optical path is usually calculated quite thoroughly for the elements and coatings it uses. Those calculatons and the lens performance do not take into account additional elements in that optical path whether coated or not. Any filter, even the most expensive, will degrade performance and inhibit the lens characteristics as they were calculated by manufacturer.
The fact that we don't see the difference in one of the many possible conditions is not reflecting the possibility of this happening in different lighting conditions. I would only be using the filter where I know for sure that I won't have that degradation significantly visible (not even in postprocessing) in the light conditions I'm shooting. Otherwise the filter comes off for me. For example for this Sigma 10-20 and my filter I established by trial that I can only safely use it when the light is very diffused and soft - like in a very cloudy weather. So I was ok using it yesterday with the filter during the snowfall with no significant effects. But I'd never use this Siggy with the filter in a bright sunny day ;)
vBulletin® v3.8.4, Copyright ©2000-2012, Jelsoft Enterprises Ltd.