Examples of looking up trees shots with TIlt-Shift lenses

[steve_wagner1]

<p>Anyone have some good example of tall trees shots done with TS lenses that show the difference? Thanks</p>

[sumo_kun]

<p>Are you shooting right up the trunk of a tree? Or from further away but still looking up?<br>

I think if you search for movements in large format, it will give you an idea of what movements can achieve.</p>

<p>Here is a page which details the movements and their general usage<br>

www.toyoview.com/LargeFrmtTech/lgformat.html</p>

<p>This site has some examples but no before after type shots<br>

<a href="http://www.luminous-landscape.com/tutorials/movements.shtml">http://www.luminous-landscape.com/tutorials/movements.shtml</a></p>

<p>What are your trying to achieve?</p>

[leicaglow]

<p>Brett, Are you looking to straighten the lines of the tree trunk, extend the focus range of the trunk, or both?</p>

<p> </p>

[jeff_conrad]

<p>Brett,</p>

<p>I second Chuk's question. Bear in mind that, for a full-frame camera, TS lenses offer only a half-frame shift in landscape orientation, and only about 1/3 frame shift in portrait orientation (multiply these values by 1.6 or 1.5 for APS-C sensors), so they won't let you take the classic redwoods and rhododendrons in the fog shot while keeping the tree trunks vertical. In such a situation, you can make minor adjustments, either decreasing or increasing the convergence.</p>

<p>From a greater distance, you often would have enough shift to prevent convergence, (or divergence, famously described <a href="../equipment/canon/tilt-shift">here</a> by Frank Sheeran as “acid-trip trees”).</p>

[tien_pham]

<p>Jeff,</p>

<blockquote>

<p>TS lenses offer only a half-frame shift in landscape orientation, and only about 1/3 frame shift in portrait orientation</p>

</blockquote>

<p>What do you mean when saying "multiply"? Do you mean something else?</p>

<p>Could that be "divide" instead of "multiply"? I think when you multiply something w/ a number greater than 1 (1.5 or 1.6), the result will be larger. That means the tilt-shift lens would be able to move a greater distance. But the APS-C sensor is smaller, the sensor lies further away, thus the shift should be less.</p>

[phule]

<p>Since when does the sensor lie further away?</p>

<p> </p>

[jeff_conrad]

<p>The shift is the same regardless of format; because the APS-C sensors are smaller, the shift is a greater percentage of the frame. For example, a 12 mm shift on a full-frame 35 mm sensor in landscape orientation is half of the frame height. On a “1.6 crop” sensor the same shift is about 80% of the frame height.</p>

<p>I'd still go back to the original question: what is the objective? Regardless of format, shift is of limited use up close, but can be useful from a slightly greater distance. Tilt can be useful for getting the entire trunk sharp from a close perspective; if there are other branches or other plants close to the camera that also need to be sharp, tilt will not help much.</p>

[tien_pham]

<blockquote>

<p>Since when does the sensor lie further away?</p>

</blockquote>

<p>Sorry for the confusion. The sensor does not lie further away, but it is smaller than the FF, thus the coverage area of the sensor becomes smaller. I was thinking in terms of the left-right distance at the edge in my mind, for simplification of the reasoning when responding to the post by Jeff. Actually, I should be cleaarer when saying that in previous post.</p>

[tien_pham]

<blockquote>

<p>For example, a 12 mm shift on a full-frame 35 mm sensor in landscape orientation is half of the frame height. On a “1.6 crop” sensor the same shift is about 80% of the frame height.</p>

</blockquote>

<p>I still don't quite get it! If the APS-C sensor is smaller, the area is smaller, the area coverage of the shift should be less. How can a shift, say 12mm, can cover 80% of the frame height from 50% (half) of it? The sensor is smaller, hence smaller area (both height and width,) the shift is still the same (12mm), and you said the shift will be larger for the APS-C sensor?</p>

<p>Your 2nd paragraph is a moot point, depending on how the shooter orients the camera. If he/she orients his/her camera in the "horizontal" way, the tilt (up/down) may not be relevant. On the other hand, if he/she orients his/her gear side way ("vertical"), then shifting lelft-right may not be relevant.</p>

<p>I used the terms "horizontal" and "vertial" instead of "landscape" and "portrait" because one can hold the camera in the portrait format, but the picture taken is still in the lanscape format, especially if that person points up his/her camera to the sky. In case someone may wonder about them!</p>

[sumo_kun]

<p>I think what Jeff was trying to say is that 12mm is half (50%) of a 35mm frame height. That is how much a typical shift lens can shift with the camera in landscape. This is a fixed value. Now on APS-C, 12mm shift is equal to 80% of the frame height. That is why Jeff said multiply by 1.6. The physical distance the lens moves is the same but as a percentage of the frame height, its larger.<br /> I am having trouble visualising how you can orient the camera in portrait but take the picture in landscape... unless you are taking a picture of something sideways?</p>

[tien_pham]

<blockquote>

<p>The physical distance the lens moves is the same but as a percentage of the frame height, its larger</p>

</blockquote>

<p>True. I uderstand now. We were refering different views of the same thing!</p>

<p>It is not very hard to orient your camere in a "portrait" orientation and still get a picture in the landscape view. This is because there is no reference for one to recognize the photo as the portrait or landscape format. Pointing a camera to the sky for example. You cannot tell the orientation.</p>

[jeff_conrad]

<p>Since we're talking about trees, I assumed the shift would be vertical regardless of camera orientation. I suppose distinguishing between rise/fall and shift (cross to Brits) would make this more clear, but small-format users and manufacturers usually don't make the distinction.</p>

<p>I think perhaps we're getting carried away with the details. My original intent was simply to indicate that a TS lens, even a 17 on full-frame 35 mm, won't capture tall redwoods while keeping the back vertical. But we still don't know whether that's what Brett is considering, or even whether he's interested in rise/fall or tilt.</p>

[tien_pham]

<p>Jeff, you are correct in saying about the shift. It was me who really carried the discussions away! Sorry about that. I just want to understand any differences in those values or any perpectives, if there is any.</p>

[ilkka_nissila]

<p>At least with the Nikon 24mm (and the newer Canon 24mm), the shift room is more than enough for practical use. Few large-format lenses of similar angle of view (i.e. 84 degrees) allow significantly greater shifting (in terms of angle of view; of course in absolute units of distance they provide more shift). I think the image looks already quite extreme when shifted 10mm. Normally I also want a bit of ground to be visible in the shot. In fact often I leave a slight bit of keystoning to make it look more natural, though I would not have to.</p>

[steve_wagner1]

<p>Sorry for ignoring the responses. Yes. I'm talking about a classic looking up a redwood shot, standing maybe 20 feet back, camera low to the ground, vertical framing at 17mm or so, with the base of the tree going to the bottom of the frame. Wondering how that would look with TS</p>

[jeff_conrad]

<p>Brett,</p>

<p>How it would look would of course depend on how you set up the shot. With a camera height of 5 ft, using the maximum rise of 12 mm on a full-frame camera from 20 ft, you could just about keep a 40 ft tree vertical; you'd include some foreground at the bottom of the image. Now 40 ft isn't a very big redwood; you'd need to point the camera up to capture more of the tree, as is usually done in such shots.</p>

<p>I doubt you'd want to keep even a 40 ft tree vertical, though. As Ilkka mentioned, using extreme shift to maintain vertical lines usually looks unnatural. A general rule of thumb (at least with buildings) is that if you need to look up more than about 20°, it's better to allow some convergence. Just how much is a matter of personal preference and the particulars of a given scene. With full 12 mm shift on a TS-E 17, the elevation of the line of sight is about 35°, which is quite a bit; shots of buildings with that much shift look unnatural to me. Even with the original TS-E 24, the line of sight is elevated about 25° at full rise, and at least to me, the top of a building looks as though it's reverse keystoned (even though the sides are perfectly vertical), so I usually allow some convergence. Trees, of course, aren't buildings, but I think the same principle is still largely applicable.</p>

<p>A lot of shots of redwoods and rhododendrons from the trails in the Northern California parks require looking up. I've never used a TS lens for this (even though I've had a 24 and a 45 with me), so I can't really comment from direct experience. Even with a 24, I'd probably want to be closer than 20 ft (and on many trails you have no other choice). As I've said, eliminating convergence would be almost impossible, and would probably look unnatural even if it were possible. It might be possible to reduce the convergence, but this isn't something I've spent much time with. I have used both the TS-E 24 and the TS-E 45 to capture small sections of trees when the flowers were of main interest. In those cases, convergence of the sides of the trunks would have been distracting.</p>

<p>Many trails in the redwood parks follow creeks, and sometimes the trails are fairly high above the creeks, requiring looking down to catch the bases of the trees and most of the flowers (the biggest challenge is often finding a view without foreground clutter). In those cases, falling front can be useful; to my eye, the divergence that obtains with a 24 mm lens is distracting (at the time, I didn't know this was the dreaded “acid trip tree” syndrome).</p>

<p>In some cases, tilt may help to get the entire tree sharp, but I've usually found that the many of the rhododendrons are fairly close to the ground (at least in comparison to the tops of the trees), so the regions I wanted sharp weren't wedge shaped. Of course, if you just want the tree, you might be dealing with an essentially planar subject, so the shape of the region of interest be less of an issue.</p>

<p>I probably haven't done much to answer your question because I'm still not sure what you're trying to accomplish. A more succinct answer might be that, at least for me, a TS lens wouldn't be the lens I'd normally use. Because of limited options for camera positions, I often appreciate the ability to adjust framing with a zoom. As always, of course, you may be envisioning pictures quite different from ones I take, so what I've said may not apply.</p>

[sravan]

I use T/S lents to stand just below old trees with dead branches and take shots of the tree closeup. I use the Tilt more to get the plane of focus better. I actually do not shift in this case as i want that convergence to be more.

Examples: