Ultrawide angle lenses & graduated ND filters

[© Tom Cooper not going to Views]

I just obtained a Rokinon 12mm f/2.0 lens for my Sony NEX-6.  I don't recall ever being more pleased with a lens.  But...

 

While it is possible to put a polarizer on a lens this wide, the results (re: the sky) are generally not satisfactory, since the polarizing effect will apply unevenly.  So the best bet is to use a Graduated Neutral Density filter.

 

Anybody have any suggestions as to how many stops I want for a graduated ND filter?  I will be using it to tame overly bright skies, and will probably end up with two or maybe three filters so I have a variety of options.  My initial reaction is to just get a 1X and a 2X, which should give me three options (1X, 2X, and both for 3X).  I don't think using more than two would be a good idea with a lens this wide, and even using 2 might be a problem with vignetting.

 

But then I got to thinking...is it just about the density of the filter?  There are two other issues that come to mind.

 

First. just how gradual should the graduation be?  What looks like a smooth transition in a short tele might look like a solid line at this focal lengh.  I don't see that being spec'd anywhere.

Second, most graduated ND filters I have seen assume the lens is pointed close to the horizon.  I don't do that very often, and expect to do it a lot less with this lens.  If I get close to the ground and point the camera high in the sky, I would expect the use of a center-horizon ND filter to be obvious by the presence of a bright band of sky below the filter.  Pointing down low would have the opposite effect, where the sky is darkened, but so is a good deal of the foreground.

 

Apart from investing in an expensive kit that allows me to slide ND filters to specific positions in front of the lens, does anyone have any insight or advice on my twin delemas?

 

Thanks,

 

Tom

[Colin Hughes]

Graduated filters are OK with level horizons - seascapes, flat land, etc. - but they are a problem otherwise. A more satisfactory solution is to blend two images, one exposed for the sky and the other for the land (if you Google "blending exposures". you should find many tutorials on the subject. An alternative, is to use software, such as Google Nik suite, to imitate the graduated filter effect but restricting it to the sky portions of your image. Also, newer cameras have an HDR function which can automatically combine bracketed exposures

[Lady GooGoo La La]

Yub polarizers no go!

I found that in my experience a 1 stop ND grad is almost useless. I use exclusively soft grads, in both standard threaded glass filters and cheap plastic Cokin series P, the plastic filters scratch very easily and may not be truly neutral. The Cokin "P" series like the Lee plastics (better but more expensive) are big and bulky but do allow stacking.

I like my ND 400 welding goggle filter the best. ;-)

With the soft grads I dont really notice an abrupt lighting change but maybe I dont look close enough.

The HDR approach maybe easier as Colin suggested, even when I use the ND grad filters I still cant catch all the tones, I prolly just dont really understand how to use em. I seem to get the best results with a 2 stop ND grad when I use grads instead of HDR.

La La La

[Lady GooGoo La La]

If a polarizer works for you the other alternative is to use a slightly narrower lens with a polarizer and stitch them together. Send the 12mm lens back!  ;-P

[© Tom Cooper not going to Views]

Unfortunately, a polarizer with stitched images of the equivalent width would not work.  The problem is that the polarizer's effect on the sky itself is not even, and varies based on the angle between the camera, the sky, and the sun.  The darkest area of blue sky is always going to be 90° from the sun, and that is true whether a single 12mm image is shot or two 16mm images are stitched together. :(

 

No matter how much research I do on this topic, I keep ending up back at careful HDR as the best solution.  I prefer in-camera solutions when possible.  While my camera can do 3-shot HDR, it only produces a JPG (no RAW) for HDR, so I don't get the lattitude I like with a 16 bit image.  Also, the only control I have over the HDR process is the number of EV stops between the three (locked in to 3) images.  I was hoping to avoid always having to carry a tripod and all the computer time that goes with doing the HDR right.

 

Tom

[Kevin Childress]

Tom,

I'll throw in a couple comments for what its worth; maybe this will be useful down the road. 

Your last question in your original post mentions a kit that allows you to slide ND filters to specific positions in front of the lens. In my opinion, if one makes an investment in using large grad NDs, a filter holder that allow vertical adjustment of the filter is indeed necessary. I've only seen two other "solutions".  1) The photographer hand-holds the filter in front of the lens and hopes the graduation is in the right place. I watched a guy do this one day when attempting to photograph a gorgeous rolling-mountain sunset. He would look at each take on the LCD, wasn't happy with the position of the grad, and try again. He fiddled with the filter so much that he lost the light and never got the shot. 2) Using gaffers tape to attempt to tape the filter in place. Same issue as above and much harder to tweak ... 

Deciding which density of a filter to purchase was very frustrating for me in the beginning. My first was a 2-stop Cokin filter. All of the images shared below are in conjunction with that same 2-stop Cokin filter. Images #1 and #2: This is only to illustrate the apparent density of 2 stops against a white background and against a black and white print. You can get a pretty good idea from these examples of what 2 stops looks like and how gradual the transition is from Cokin. I've seen that not all filter manufacturers have the same graduation - some have softer, more graduation transition than others. For me, the softer the grad, the better.  Image #3 is the most important example here. Bottom line: read as many reviews as you can about color shift with a given filter. As you can see with this Cokin filter, "neutral" density is not always quite neural. Image #3 is a screenshot of an unprocessed raw file, except for me manually setting the white balance from the building's white walls. This was photographed on a perfectly cloudy day - the entire sky was gray. Every bit of the blue and purple in the sky is color shift from the filter. Clearly that filter isn't neutral at all. I've stopped using the Cokin filters for this reason. "Neutrality" is definitely one area where you get what you pay for.  Finally, the orange lines on image #3 are placed approximately where the graduation occurs on that filter.

That 2-stop Cokin is also made from a very soft resin which scratches very easily. So I would also look into the physical material of the filter and read what other's have experienced in regards to scratching. 

[© Tom Cooper]

Kevin,

 

I just realized this reply was here.

 

First off, I want to point out that the net effect in images 1 & 2 are of a 4-stop filter, not a two stop filter.  That's because the incident light is reduced 2 stops going towards the paper/print, and then reduced 2 more stops coming back out.

 

Second, as I mentioned in a comment on one of your other photos, I think I'm going to concentrate on using judicious HDR than ND grads.  That avoids the issues of rate of gradiation, placing the gradiation, issues with shape of the horizon, defects in the filter, and non-neutral color balance.  Oh, and if it's hosed up, I can only blame myself.

 

Having said that, well-done HDR obviously requires a quality tripod.  I've been using the same department-store tripod for nearly 20 years, and it is well past it's useful lifespan.  I have about three hundred to drop on a new one and expect to be picking something out this weekend.  :-)

 

Tom