Lens and filter options for landscape photography

[jhenderson0107]

I currently own a Canon 5D mkIII and five lenses including the Canon EF 14mm f/2.8 L II, Canon EF 24-105mm f/4 L IS USM, Canon EF 70-200mm f/2.8 L IS USM, EF 100mm f/2.8L Macro IS USM and Canon EF 50mm f/1.2 L USM.  I am an amateur that shoots family portaits, woodworking projects and occassional auto and air shows, but I primarily enjoy hiking and landscape photography. 

I am contemplating the purchase of a new landscape lens (TS-E 24 II, TS-E 17, 24-70 II or Zeiss 21mm) and believe I need to incorporate use of filters to improve my landscape images.  The cost of the 100mm HiTech holder, grad filters,  circular polarizer and adapters tallies to over $1K.  As a point of reference, the cost of a 82mm quality variable density filter and circular polarizer from B&W is about $600. 

My understanding is that grad filters limit the dynamic range of portions of an image to work-around the limitations of the imaging sensor.  If sensors had sufficient dynamic range, use of grad filters would be unnecessary since the gradient effect and tonal modifications can be applied in post-processing.  I understand that the effects of polarizers and uniform darkening filters cannot be emulated in post. 

Here are my questions: 
1. My perception is that use of a Lee or HiTech filter holder and plates will require more setup time and more fragile than screw-on filters.  Do the virtues of a filter holder system, such as the HiTech outweigh the benefits of screw-on filters?
2. If I were to invest in a new body and lenses capable of higher dynamic range to augment my current gear, would this substantially alter the makeup, need for and cost of a filter system?  In other words, is there another equivalent technical solution? 

[neuroanatomist]

Here are my questions: 
1. My perception is that use of a Lee or HiTech filter holder and plates will require more setup time and more fragile than screw-on filters.  Do the virtues of a filter holder system, such as the HiTech outweigh the benefits of screw-on filters?
2. If I were to invest in a new body and lenses capable of higher dynamic range to augment my current gear, would this substantially alter the makeup, need for and cost of a filter system?  In other words, is there another equivalent technical solution?

1. Yes, IMO.  The problem with screw-in round grad NDs is they force you to put the horizon in the center of your image - which is probably not where you want it.  A rectangular filter gives you the flexibility to raise and lower the density gradation in the frame.  Note that you can pretty easily hold the filters in your hand for the shot, without the need for a holder system.

2. Theoretically, yes.  Practically, no.  At the risk of having this thread descend into chaos (aka M.R.  ), there are sensors that provide additional mathematically-determined DR based on a calculated noise floor, but that is not the same as real world, usable DR that allows you to distinguish details in the shadows.  But there is another, sort of equivalent technical solution that's much cheaper - HDR.  Done properly with an intent to blend exposures as opposed to tone mapping, you can get very good results.  Probably worth a try before spending $1K on filters.

[EOBeav]

But there is another, sort of equivalent technical solution that's much cheaper - HDR.  Done properly with an intent to blend exposures as opposed to tone mapping, you can get very good results.  Probably worth a try before spending $1K on filters.

Could not agree more. At the risk of negating Neuro's excellent response, technically tonemapping is a function of HDR, while blending exposures utilizes a different approach. Both lend themselves to loosening up that dynamic range through the use of multiple exposures.

BTW, is it common to use tilt shifts for landscapes?

[neuroanatomist]

Could not agree more. At the risk of negating Neuro's excellent response, technically tonemapping is a function of HDR, while blending exposures utilizes a different approach. Both lend themselves to loosening up that dynamic range through the use of multiple exposures.

BTW, is it common to use tilt shifts for landscapes?

True...I just lumped them together practically, since the software packages that support it almost universally offer both options.

A TS-E is great for landscapes - the tilt capability allows you to achieve a very deep DoF without needing an exceptionally narrow aperture.

[Random Orbits]

Could not agree more. At the risk of negating Neuro's excellent response, technically tonemapping is a function of HDR, while blending exposures utilizes a different approach. Both lend themselves to loosening up that dynamic range through the use of multiple exposures.

BTW, is it common to use tilt shifts for landscapes?

True...I just lumped them together practically, since the software packages that support it almost universally offer both options.

A TS-E is great for landscapes - the tilt capability allows you to achieve a very deep DoF without needing an exceptionally narrow aperture.

+1.  Sometimes I wish it had 2 shifts though  -- one to fix converging verticals and one for panos.

[wayno]

But there is another, sort of equivalent technical solution that's much cheaper - HDR.  Done properly with an intent to blend exposures as opposed to tone mapping, you can get very good results.  Probably worth a try before spending $1K on filters.

Could not agree more. At the risk of negating Neuro's excellent response, technically tonemapping is a function of HDR, while blending exposures utilizes a different approach. Both lend themselves to loosening up that dynamic range through the use of multiple exposures.

BTW, is it common to use tilt shifts for landscapes?

IMO, with the exception of the 0.1% of people who use HDR effectively, grad NDs will always look better and more natural than HDR.

[Lloyd]

If you are considering filters, the TS-E 17 does not lend itself to the use of filters.  There are several who have made some very interesting devices to adapt filters to this lens, but I don't believe that there is an off the shelf device to hold a filter for the TS-E 17.

[jhenderson0107]

IMO, with the exception of the 0.1% of people who use HDR effectively, grad NDs will always look better and more natural than HDR.

The results of my HDR attempts place me squarely in the general statistical population.  They were...not beautiful. 

I borrowed the TS-E 17mm from CPS last week and while I enjoyed using it, it's too wide for my intended landscape uses.  Creating well-framed landscape shots on my current 14mm is nearly impossible for me - all of my attempts become incoherent, sprawling colorscapes.  The TS-E is a little narrower, but I need to become proficient with a 24mm lens and some filters before attempting anything wider.  Once I routinely obtain good results at 24mm, I can graduate to use of my wider lenses and perhaps I'll spring for a TS-E 17mm then.  HiTech makes a 165mm lens holder suitable for the Canon 14mm and perhaps I could adapt it to a 17mm lens also. 

[EOBeav]

IMO, with the exception of the 0.1% of people who use HDR effectively, grad NDs will always look better and more natural than HDR.

The problem with grad ND's is that if you have something in the foreground that sticks up into the top of the image, it's going to be unnaturally darker at the top than at the bottom. Bracket and blend, and you'll have no worries. Once you stick a filter on, you're stuck with what comes out the other end.

[PackLight]

Here are my questions: 
1. My perception is that use of a Lee or HiTech filter holder and plates will require more setup time and more fragile than screw-on filters.  Do the virtues of a filter holder system, such as the HiTech outweigh the benefits of screw-on filters?
2. If I were to invest in a new body and lenses capable of higher dynamic range to augment my current gear, would this substantially alter the makeup, need for and cost of a filter system?  In other words, is there another equivalent technical solution?

1, Yes more set up time. With graduated Lee type filters you can move them up and down as needed to get the exact exposure you need. You can stack these and place them at different angles to take the light down on a slope or say the side of a mountain. You can stack them with less fear of vignette. They are much more usable. They are expensive and they take take more time.
2, Technology is good and there are things you can do as others have said, but getting it right at the camera is always better.

I would go for the TSE 24mm or the Zeiss 21mm.

[jimjamesjimmy]

a 5d 3, 5 awesome lenses and considering buying more gear, yet not sure what a nd grad filter does? 

[scrappydog]

1. My perception is that use of a Lee or HiTech filter holder and plates will require more setup time and more fragile than screw-on filters.  Do the virtues of a filter holder system, such as the HiTech outweigh the benefits of screw-on filters?

Tilt shifts require more time to set up and shoot, regardless of whether you use rectangular ND filters or screw-on types (e.g., you must check the corners in LiveView, manual focus, use camera metering in default position, etc.).  The rectangular ND grads are more fragile, and more prone to scratches, but they are essential IMO to doing landscape work.  Usually, I hand hold my ND grad filters rather than use my Lee Foundation Kit holder.  With the TS-E 24, this is simple.  It is a lot more problematic with the TS-E 17 because of the bulbous front element -- any backlighting will reflect off the ND grad back into the front element.  The rectangular ND filters are preferable to screw on filters because, as Neuro pointed out, your horizon line may not be in the center of the frame. 

I have two ND filter recommendations, in the minimum: (1) get 4x6 rectangular filters, and (2) get a 2-stop soft and 3-stop hard.  The 4x6 size will permit you to hold them and have room for your fingers.  The 2-stop soft is more pliable to landscapes that are not straight horizons, and the 3-stop hard will cover the horizon shots.  You can still add ND filter effects in Photoshop, but this does you no good if your shot exceeds the dynamic range of your camera.  I recently shot my TS-E 24 II (with a 5DII) in a large greenhouse and the ceiling was about 5+ stops brighter than the ground level -- I had to stack 2-stop and 3-stop ND grads and meter down to not clip the highlights.  You cannot resurrect clipped highlights in software.  HDR is an option, but there are times when it is not viable.  You can also composite shots, but again, there are times when this is not viable.

2. If I were to invest in a new body and lenses capable of higher dynamic range to augment my current gear, would this substantially alter the makeup, need for and cost of a filter system?  In other words, is there another equivalent technical solution?

There may be a solution with your existing gear.  When I use my TS-E 24 II, I use my 77mm filters (the TS-E is my only 82mm lens, but I have several 77mm lenses).  I have a B+W UV on the TS-E; I hand-hold my 77mm filters, which rest against the 82mm UV filter.  I wasn't sure if it would work, but it does.  Make sure you don't get your fingers in the shot though.  I have done this with my Hoya FL-W magenta, Singh Ray Blue-N-Gold, and and B+W circular polarizer with positive results.

So, the bottom line is that if you have 77mm filters, you can get by with the TS-E 24 II (short of a UV filter, which you would have to get an 82mm filter).  You don't need $1,000 worth of kit to balance your exposure; a couple 4x6 ND grads should be sufficient.  If you want to get a Lee Foundation Kit, this is a good deal, but not necessary to get started.  Of the lenses you mention, I would recommend the TS-E 24L II because (1) it is very sharp, (2) it has perspective correction, (3) it is more versatile than the TS-E 17 (I have both and I can attest to this), (4) it complements your existing lenses, and (5) I have never shot the Zeiss, but it does not have perspective correction.

[Arkarch]

ND Grads, HDR, spot burning/dodging (NIK Vivesa for example), and all out PS layer blending are all reasonable techniques based on situation.

My preference is ND Grads where possible.  It is, as you said, expensive and time consuming.  Its not something to pull out if you are with family or friends who just want to take the shot and move on.  Using ND Grads means working the scene.  But if you have the time and the budget -  the results are worth it.

However sometimes the geometry of the subject or composition does not work with filters.  Then its the software tools. 

I tend to first look at Photoshop or if its minor like shadows, NIK Vivesa first.  Layering for the big areas; Vivesa burns/dodges for the smaller areas.  Before anything else, make sure you have done all you can with LR / Aperture, etc to have a good source and an best-case histogram.

If HDR is the only way, I might want to question the photograph first.  Most HDR unfortunately will end up looking unnatural.  Its very hard to be subtle; tho sometimes it can be a neat look if that is your intent.

All that said, in the right, experienced, skilled hands, HDR can work for many situations.

[Kernuak]

I'll start with the easier angle. I recently sold my 17-40, because I rarely found a use for it after getting a full frame camera, it was simply too wide for the type of landscapes I usually shoot. I would therefore suggest something in the 24mm sort of range, particularly as you also mentioned finding your 14mm too wide. I also wouldn't rule out the 24mm L f/1.4 MkII, it's just as sharp as the 24mm TS/E and Zeiss 21mm when stopped down beyond f/2.8 according to data I've seen. It's certainly sharp enough for me and is in a totally different league to the 24-105. It's also cheaper than the TS/E and probably slightly cheaper than the Zeiss. I went through a similar decisions making process over the past couple of years and after a protracted period, went for the conventional 24mm because of its potential for Aurora photorgaphy and not for any useage or technical reasons. All three lenses are about the best you can get in that range and there is very little difference between the three in terms of image quality. The Zeiss probably has a look that many prefer, the conventional 24mm can be used for low light photography and not just landscapes and the TS/E of  course is useful for shorter exposures and can be used more in the aperture sweet spot when tilted.
In terms of filters. I am certainly not a fan of HDR, I've used it in the past and it can be used effectively, but it takes time to learn how to get good results and even when done well, there are some scenes where it doesn't suit at all (e.g. snow or solid blue skies with a high contrast transition such as in silhouettes). The way that the majority use it, it is more like a cheap man's method of filtration and not the occasional tool it should be. Also, because of its widespread use, if you want to look different to the masses, then I would avoid it. As for cost of filters, $1000 seems very steep (especially for Hi-Tech). While it would be nice to have a full set of filters, I don't think it is necessary. I would pick and choose, then add to your collection gradually. If you shoot a lot of seascapes or flat landscapes, get a set of three hard grads, if you shoot a lot of mountains or hills, get a set of soft grads. If you were really stuck for cash, then you could achieve a fair bit from one or two 3 stop grads (a hard and a soft maybe). I would also get a circular polariser, but leave any solid ND filters for a later date when you want to start experimenting with slower shutterspeeds.
One thing to bear in mind though. With the Lee system (I'm not sure about Hi-Tech), you can add a second adaptor to the front of the first. This allows you to angle the filters in two separate directions, for example to filter either side of a mountain or to use a CPL in one direction and a grad in another. However, with the 100mm filter set, the corners of the second adaptor are visible in a 24mm lens.
Filters will allow you more latitude in dynamic range and even if the filtration isn't strong enough for the final output, often they provide enough DR compression to preserve enough detail for recovery in your RAW convertor, where you can add further graduation.

[jhenderson0107]

Thanks to all for your helpful responses and suggestions.  Per the advise above, I ordered just two 4x6 grads and a B&W slim screw-on circular polarizer - better to add more later only as needed.  I opted for the TS-E 24mm II.