PDA

View Full Version : My silly little polariser thread



arthurking83
22-06-2009, 2:03am
When bad can actually be a good thing!

A CPL(or polariser) on an Ultra Wide Angle lens!
You may read comments by some folks that using a polariser on a very wide angle lens is not normally a good idea, as it causes weird blue sky effects!

http://www.ausphotography.net.au/gallery/files/1/0/_DSE7079.JPG

like that. And of course it's true(hey I just posted an image with that weird effect didn't I? :D) but many of the comments I've read from various posts(not only here but on many forums) regarding that phenomenon seem to be of the opinion that it always causes that effect, which is simply not true!
ie. the wording in the advice seems to be advocating that you will see this effect, not that you may see this effect. Two different ways of describing what should be a simple hint or tip or caveat.

So it does happen and you will almost likely see it too on your UWA lens, and having a higher quality polariser makes no difference.

But I sometimes put that effect to work for the common good! :D
Sometimes you may come across a gully or creek or valley where the dynamic range from a blue sky to the darkest shadows in the deep red earth is just too much for your camera to handle and GND's are all but useless really. You maybe able to regain 1 stop more DR using a good GND, but then the trees or hills or whatever that line the sides of the gully/valley will also get affected by the GND too. So another stop or two of exposure range can be gained by using that ugly blue sky(UBS) phenomenon to your advantage.
As I did in this recent shot from Bendigo(actually out of Bendigo near Maldon.. but that;s not the point! :p)

http://fc00.deviantart.com/fs47/f/2009/172/5/0/Golly_by_arthurking83.jpg (http://www.ausphotography.net.au/forum/showthread.php?t=33530)
Image links to the thread. In that thread the third image also benefited by that UBS effect, but not as much as this particular one.

Problem is that you can't rely on it. When you can't rely on it(UBS) to happen then you will almost certainly see normal polarisation, so you still get some exposure benefit.. but UBS seems to concentrate the darker blue in that weird triangle pattern, and that seems to be maybe up to another stop of exposure latitude.
How this helps is that you can overexpose the part of the sky that you'd like to make blue, and recover it later in PP.
Over expose by 1 stop and you should recover that colour in your editing software, but more importantly you will be a lot closer to a better exposure value for the shadow areas.
Technically I stuffed that gully shot as I didn't over expose the sky by enough. But that's not really the issue here. I did bracket the shot though so I did get a brighter exposure version of that scene, but this exposure looked nicer on viewing the histogram, and the brighter version(that was +0.7Ev compared to this version) had over exposed foreground spots, and the shadows were still recoverable without losing any color info. That's all that's important.
So the point of it all :confused013 ... you can maximise your dynamic range by thinking outside the square.
Where you see problems, it could actually be a solution in disguise :th3:

As for that first image example the UBS shot. That was a slightly underexposed version of a bracketed scene. Even though I bracket, I still don't HDR my images.
But with the +0.7Ev shot the UBS issue was easily manageable would have been recovered without fuss.
If it's too dark you have trouble doing so and the resultant image will have data loss/blue channel noise and a few weird mottled effects between the (now) recovered darkest part and the pulled back brighter parts of the sky.

But what I do if I can't avoid the UBS issue, is to rotate the polariser until the darkening in only on one corner, and then use the GND turned at an angle that balances out the other corner. it's simple and usually effective(if you have a GND of course! :p)
ie. in the upper image the exposure I used to process I had the UBS spot nearer the upper right, and the GND turned about 15° anticlockwise to darken the left corner.

Hence, in a few of my images I stack polarisers and GND and more GNDs and maybe even another GND!.... just for the hell of it, and because I have a few of them.
That corner is the one furthest from the sun(as you'd expect :rolleyes: ) so you you want to use GNDs in that manner just try to think about that element next time you start setting up your composition. ie. if you are facing the 'wrong way' then if you can turn round and face the right way. :D

I'm still yet to find a practical use for stacking polarisers though... but I'm working on it :p
I still only get a brown double v effect(at it's most extreme effectiveness and you can rotate the entire stacked set of polarisers to vary the orientation of that V within the frame.
Reality is that it looks uglier than the UBS effect... so we'll call that one the UIERBVS
(UglIERBrownVSyndrome).
I just can't think of any practical use for it :crzy:

hope someone can use this hint/tip. It helps me sometimes.. not perfectly all the time and with creative use of filters(eg. I still have to stack a GND on occasion too) but as long ats the resultant image turns out with a good exposure value(histogram) that's all that matters


:)

David
22-06-2009, 8:25am
Thanks Arthur for that wealth of education about Vignetting and how to overcome in on UWA shots.

I am going to take a month to digest it and another 3 to apply it effectively but this kind of detailed information you cant get at a workshop and having people on AP who have an amazing amount of experience and know how and share it freely makes this site GOLD for newbies to photography.

Thank You

jjphoto
22-06-2009, 9:56am
The polarising effect as seen by the camera is determined not only by the rotation of the filter but by the position of the sun (because of the way this polarises the light). In summer when the sun is almost over head then the polarising effect will be pretty much even in every direction (along the horizon) but when the sun is lower, so it's angle relative to the camera is more variable, then the effect will vary greatly relative to the suns position. The effect of the polarising filter is greatest when 90 degrees to the sun and diminishes to minimum effect when 180 degrees away from the sun (ie sun behind you). Your 'UBS' effect is an illustration of that simple rule because the UWA lens shows a broad area of the sky and there for the changing effectiveness of the polariser across the sky.

I think the polarising filter is possibly the only 'must have' filter, but of course there are many other useful filters. Yet many people seem to hate the polarising filter. I often think that only happens when they don't really understand how/when to use it.

JJ

arthurking83
22-06-2009, 10:28am
......

I think the polarising filter is possibly the only 'must have' filter, but of course there are many other useful filters. Yet many people seem to hate the polarising filter. I often think that only happens when they don't really understand how/when to use it.

JJ

:th3:

Going by the 90° technical specs of polarisation, this scene shouldn't have had the level of UBS that it did(except that I did underexpose, and undexposure has exaggerated the effect by a factor of 2or3 in the case above.
I think the most important factor aspect of trying to understand polarisation(in photography is to be aware that it changes with latitude too. Your position will determine how a polariser will work(as well as altitude).

I agree with the comment that in summer time you get nice even polarisation which is handy, and can make for very nicely contrasted images.

The problem I'm having with the polariser works best at 90° comments is that it may be kind of misleading. 90° in which dimension? The world is 3 dimensional and not only do you have to deal with direction, you also have to deal with elevation from the horizon.
The above UBS image is perfectly positioned at 90° to the sun. I'd say, at a guess that the sun may have had up to 20° of elevation, so how does the 90° theory affect my calculations when I'm driving along and need to position myself somewhere to keep the subject like a mountain in view?
The camera was pointed approx 90°(give or take a degree or two) yet UBS is very obviously intrusive here(unless the scene was a valley :th3:). rotating the polariser I could simply get the v shaped darkened section either on the left hand corner in a slightly wider arc, more central and slightly narrower in scope, or further to the right corner, slightly brighter and at it's less intrusive as it was spread out more(most likely due to the smaller angle with respect to the position of the sun, but also due to the much brighter sky in that location as well.

The best position I find for polarisation(in winter), is about 10-15° more(than the 90° positioning they recommend. ie. the recommendation is to have the sun over your shoulder, where I see better results (with the 10-20mm) if the sun is just behind my shoulder.. and with careful rotation of the pol you can usually even out the blueness of the sky.

old dog
22-06-2009, 11:15am
thanks Arthur for this explanation. I am looking at a cpl and nd`s and gnd`s for my kit. 77mm I think (because of the 12-24) and step down for the others. Still unsure what system yet.

nisstrust
22-06-2009, 11:56am
Peoples, any opinions on a singh ray 3-8 stop ND vari polariser?

http://www.singh-ray.com/varind.html

thinking of getting one....

arthurking83
23-06-2009, 2:53am
Peoples, any opinions on a singh ray 3-8 stop ND vari polariser?

....

Nisstrust! A ND filter is not the same as a polariser. Doens't give you the same effect.
An ND only allows a longer exposure time for a given aperture(where you may not want to close the aperture down for either DOF(ie. you want shallow DOF) or diffraction(stopping too far can cause lost image sharpness.

A polariser gives you contrast more than anything else, but also removes reflections and glare(which is the same as any other reflection too :D)

if you were referring to the VariNDuo (http://www.singh-ray.com/varinduo.html) filter that they have, or the BlueNGold polarisers they also offer??... all I can say is:

Damned expensive! :D

Of course with no experience with any SinghRay, I can't give an opinion other than how damned expensive they are!
(ps. I think I'm about to pull the trigger on a few of their standard GND filters though)

I suppose, though.. if their value added polarisers give any benefit over a standard polariser, then you'd have to weigh up the cost disadvantage over the improvement in the resultant images.

Think of it this way. 70-200mm/2.8 lenses.
The Nikon is 2 1/2 times the price of the Tamron, yet I'm more than happy with what the Tammy gives but the VR is the key issue for me. Would I pay $1.5K more for the VR feature. YEP! :D(of course hindsight is a wonderful thing, isn't it? ;)

back to the glare/reflections topic. If you require the use of a polariser and an ND filter say because you want to remove glare and produce milky water flows, then this type of filter may make some sense, otherwise you'd have to stack polarisers and ND filters. That could lead to other IQ issues, such as vignetting or loss of sharpness(where it may count). But what I see the issue being is that glare can be dealt with with exposure and you won't get glare in a milky looking stream of water anyhow, so it seems kind of redundant. I reckon(by estimate) that reducing exposure so as to not over expose the glare from those wet rocks, you risk losing just under 1 stop... maybe closer to 2/3rds of shadow exposure. That's just a rough estimate from what I've seen.
Using a polariser on shiny foliage to reduce the glare of that shiny foliage, usually gets me about 0.7ev of extra details in the shadows.

Polariser_1
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3857&gal=gallery&type=full

No_Polariser_1
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3858&gal=gallery&type=full

Polariser_2
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3860&gal=gallery&type=full

No_Polariser_2
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3859&gal=gallery&type=full

All images were taken with the Sigma 50mm lens and then a quick unscientific shot was taken with the polariser giving maximum effect, and then, without removing the Pol, I rotated it to a point where it seemed like zero effect. Maybe or maybe not the same as not having the polariser on, I really don't know or can't remember now.. but as it was an unscientific test and this is the SILLY polariser thread, we'll just accept that rotating the polariser to minimum effect is the same as not having one on(save for about a 1 stop loss of light, as polarisers are known to give that exposure effect)

In image 1 the lens was directed at about 40-50° to the directional position of the sun. That is, the sun is just in front of my right shoulder. I had to put the lens hood on correctly so as to avoid flare. So even at less than 90° to the sun, you can still get maximum polarisation of the subject. The reason the polariser was fitted, was to minimise/eliminate the glare of the foliage... shiny green stuff.
That's blatantly obvious in images #1, and slightly less so in images #2, which were at 90° to the position of the sun. But the obvious factor is that the glare is gone in the polarised images. As I'd see in on a normal day(as I wear polarised sunnies as standard procedure, even in wet overcast conditions).

In fact!.... polarisation is a strangely interesting topic of mine(because of the sunnies that never leave my face, unless I'm gazing wantingly into the eyes of my beloved :cheeky33: !... Ooops I have no beloved, but that's not the topic at hand) and that translates easily into photography related interests too. The reduction in glare is the main reason, which is related to increase in contrast too, and that means that in sopping wet driving conditions even on an overcast day, I can see just that little bit further ahead.

If I can see further ahead, then it's going to be obvious that the camera/lens must be able to also!? :confused013
I can't actually see further ahead.. it doesn't magnify the scene, I can make out more detail further ahead.

in both image samples(which are not perfectly matching frames, but close enough) you can make out a little more detail in the shadows, and the over exposed leaves are now more vibrant, less distracting.
What I did: was to keep exposure constant via the cameras meter, but in both cases the camera gave me 1/3 stop less shutter speed in the polarised scenes, as it tried to under expose the glare on the leaves(spot metering -1.3Ev on the glare of a leaf)
Each of the images in each of the series was metered on the exact same location, to maintain some consistency and the camera was allowed to expose as I ... kind of wanted it too aperture priority mode, and in both of the series, and a few others I did on that day, the difference was always a 1/3rd stop exposure difference(via shutter speed) always slower in the polarised shot.
exif data is all still intact, for those interested and can find fault with my faulty testing procedures!
What I should have done: was to over expose the glare on the leaf in the polarised shots, maybe even over exposed it by about 1 stop or so, recovered that detail ion processing, but the slower shutter speed that would have given would have been great to get more detail in the shadows too.
Can't really do the same thing in the non polarised shots, as the glare is already over exposed!
So one thing a polariser does, is to give you slightly better dynamic range, as well as more vibrant colours. colour contrast, as opposed to tone contrast.

I think I've already mentioned elsewhere, I have never seen any difference in quality from my 15yo polariser which I thought was a Marumi, but is in fact a Hoya. Cheap $40 Hoya from 15 years ago, and I suspect no coatings whatsoever!