Understanding the 24 PC-E

[fossum]

<p>I have been using the 24 PC-E for landscapes, and mostly using the shift movements to achieve "right" perspectives or for stitching several photos together without the use of a pano head.<br>

Today I photographed some stones in a stream, and I found that due to the close focus, I didn't achieve enough DOF to render the distant leafs with sufficient detail. So I went back this evening and tried the tilt function. I obviously need more experience, because the results are not what I predicted.</p>

<ul>

<li>I set focus distance to about 30cm. Aperture was f/8. I checked focus using Live View on D700</li>

<li>I then moved my focus point towards the distant leafs, zoomed in, and used the tilt function (downward movement)</li>

<li>The tilt was "straight angled" compared to the sensor frame</li>

</ul>

<p>I could not see any significant difference in sharpness on the LCD as I was tilting, so I ended up <strong>tilting the lens to its maximum</strong>, and took the shot.<br>

Reviewing it on the screen, I notice that the <strong>closest subject was sharp</strong>, but that sharpness didn't cut straight through the scene, but rather "jumped" 4-5 meters, rendering the middle part "blurry", the leafs sharp (except from at the top of the image, which was also rendered blurry).<br>

So I am <strong>looking for some advice on my way to learn this lens</strong>. I have made it work at my desk, but I simply haven't understood the lens enough to use it in a cold river with wet boots! So as said, I am thankful for any advice!<br>

I will try to add a photo, albeit small, to illustrate.</p>

<p>Best regards,<br />Richard</p>

<div></div>

[Two23]

<p>I don't have this lens, but I have been using a 4x5 field camera for some years now. I have an idea of what might be going on. You have to be able to visualize what the lens is doing, in relation to your sensor. With a regular lens, the plane of the sensor and the plane of the lens is parallel, sort of like this: l l. With a view camera or tilt lens, the sensor stays fixed of course, but the focus plane of the lens sort of "lays down" onto your subject. It looks sort of like this: l /. (Use your imagination here, LOL.) If you were taking a shot of something more or less flat, such as a field of flowers, it would be in the plane of focus from near to far. However, that's not what you have here. In your shot above, the plane of focus did "lay down" on top of the creek, but those tree stick up through it. That's why the top of the trees wasn't sharp--the plane of focus went "under" that. One thing you need to change about your thinking is this. A tilt/shift lens does NOT have more DOF, it simply allows you some freedom where to place it. You can angle it with a tilt/shift, instead of only having it 2D front to back like a regular lens. Don't know how clear I am here, but basically you need to think of DOF being sort of like a rectangular "box," and a tilt lens merely has the ability to angle the box, not make the box bigger.</p>

<p>Kent in SD</p>

[john_macpherson]

<p>And to add to Kent's observations - the amount of tilt needed is often much less than you'd think. Its not easy to see on the LCD but with some practice you'll get it. My first experience with the 24t/s was with another very experienced photogrpaher and we were both completely flummoxed by it - its not the easiest nor most intuitive lens to use.</p>

[jeff_conrad]

<p>I don't have this lens, either, but I have the Canon TS-E 24 mm, and I've worked with LF. Effective use of tilt takes a bit of learning, but once you understand what's happening, it's not all that difficult.</p>

<p>When the lens is tilted, adjusting focus rotates the plane of focus about an axis below the lens. Focusing toward infinity rotates the PoF away from the image plane; at infinity focus, the PoF is almost perpendicular to the image plane. The tilt controls the distance of the rotation axis from the lens; with a short lens such as the 24, a little tilt goes a long way. The best illustrations of this I can think of are the diagrams on Merklinger's <a href="http://www.trenholm.org/hmmerk/">web site</a> . You might also want to look at the Wikipedia article on the <a href="http://en.wikipedia.org/wiki/Scheimpflug_principle">Scheimpflug principle</a> .</p>

<p>When the lens is tilted, the DoF is wedge shaped rather than between two parallel planes; the greater the <em>f</em> -number, the wider the angle of the wedge. One important characteristic of the DoF is that it's equally distributed about the PoF on any plane parallel to the image plane. For example, if the camera back is vertical, the DoF will be the same on both sides of the PoF where it passes through a distant tree, so to make the best use of the DoF on the tree, the PoF should pass through the middle of the tree. The DoF is also discussed in the Wikipedia article.</p>

<p>To best position the DoF, then, you need to do two things:</p>

<ol>

<li>Determine where you want the PoF. In most cases, you want it to pass through the middle of a distant tall object. The same would be true for a close object, but because the rotation axis is usually below ground, it's often tough to find a suitable near object. In many cases, you can simply pick the nearest object that you want sharp.</li>

<li>Having chosen the two points through which you want the PoF to pass, you need to adjust the lens so that it does so. One approach is to adjust focus and tilt simultaneously until both the near and far points are sharp. Some folks swear by this procedure, while others swear at it; I'm one of the latter, and prefer a procedure by Howard Bond, which I summarize in an <a href="http://www.eosdoc.com/manuals/?q=Tilt-Shift">article</a> on the EOSDoc site. A transcription of Bond's article is available on the <a href="http://www.largeformatphotography.info/">Large Format Page</a> .</li>

</ol>

<p>With the PoF set, you need to determine the <em>f</em> -number. Once again, there are at least a couple of approaches. One is to stop the lens down to working aperture and adjust the aperture until everything looks sharp. Another is to use the lens DoF scale, much as you would with any manual-focus lens:</p>

<ol>

<li>Focus on the nearest point you want sharp; note the marked distance. </li>

<li>Focus on the farthest point you want sharp; note the marked distance. </li>

<li>Set the focus so that the two noted distance marks straddle the index mark and note the <em>f</em> -number that is at or outside the two marks. </li>

</ol>

<p>Here, the near and far points will probably not be the ones you used to set the tilt.</p>

<p>Two caveats:</p>

<ol>

<li>When the lens is tilted, the marked distances usually don't meaningfully relate to distances from the camera, so don't try to correlate them with actual distances—do it strictly by assessing focus.</li>

<li>The “near” and “far” points aren't necessarily those nearest and farthest from the camera, but rather the least and greatest angular distances from the top of the image plane. At first, it may be a bit tough to find these points, but with a bit of practice adjusting the focus back and forth, you'll develop a feel for what's happening.</li>

</ol>

<p>I don't have the Nikon lens, so I don't know whether the markings are far enough apart to be of much use. On the old Canon 24 mm lens (which I have), the scale is very readable; on the new one, the markings are so close together that the scale probably isn't much use.</p>

<p>I'm sure this all look quite complicated, but once you do it a few times, it's actually pretty simple. I have an updated version of the EOSDoc article with diagrams and a lot more “how to” information. The EOSDoc site has been essentially static for quite a while, so I doubt the new version will ever get published, but I'd be glad to send you a copy if you'd like.</p>

[paul_kent2]

<p>My guess would be that you used far too much tilt, placing the wedge of sharp focus too parallel with the ground. The tops of the trees are out of focus suggesting this. You could also have stopped down more. In field conditions I rarely need more than a few (2-4) degrees tilt, only using more for particularly sharp perspectives or close focusing. You can practise common landscape conditions at home using objects on a coffee table for your close objects and a far wall to represent infinity. When using large format I often double check my focus before changing locations, carefully louping the ground glass. On digital I carefully review the shots paying particular attention to corners and any objects likely to be close to the edge of the region of sharpness (e.g. particularly tall trees, nearby rocks).</p>

[jeff_conrad]

<p>At the resolution shown, it's hard for me to determine exactly where the PoF is (though it obviously passes below the top of the tree on the left). When near and far are sharp but the middle is soft, it's usually because the focus isn't properly set, placing the PoF to high or too low (in this case, I'd say the former). When a scene has height as well as depth, it's impossible to get everything in focus, so you need to rely on the DoF, and this usually means running the PoF through the vertical middle of as many objects as possible. Because it's sometimes tough to find a near object on which to focus when setting tilt, I often set the tilt by passing the PoF through two points at the near limit of the desired DoF. Although this will lead to the right tilt (the PoF and the near and far limits of DoF all intersect at the same line), it won't be the correct focus (i.e., angle of the PoF). Once the tilt is set, the focus needs to be adjusted so the PoF passes through the middle of the region you want sharp. I once tweaked the tilt of an image looking across a cove bounded by 20–30 foot cliffs using this approach, and forgot to reset the focus (the wind was blowing and burning my eyes ...); the result was a sharp flower on my side of the cove, a sharp building on the opposite side, and badly blurred water in the middle.</p>

<p>I can't say whether you used too much tilt, but it's easy to do. A quick sanity check: the distance from the lens to the apex of the DoF wedge is given by</p>

<p><em>J</em> = <em>f</em> /sin <em>θ</em> ,</p>

<p>where <em>f</em> is the lens focal length and <em>θ</em> is the tilt. With a 24 mm lens, 1° tilt puts the apex 1.38 m below the lens. If the camera is higher above ground than this, it's usually too much tilt.</p>

<p>Tilt is most effective when the region of interest is a good fit to a wedge-shaped DoF. In this scene, it may be tough to get much advantage from using tilt—there is significant height as well as depth, e.g., the tall tree on the left and the rocks at the bottom. Now the height at the bottom may not seem like much, but the DoF is narrow near the camera, so even a little height can be a challenge. The only way to see if tilt helps is to try it with and without, and see which lets you use a smaller <em>f</em> -number (or gives you greater DoF with the greatest <em>f</em> -number you can live with.</p>

[fossum]

<p>Thank you all for very detailed (and partly complex) explanations!<br>

Kent: I actually think I understood what you said about the rectangular box. It makes sense. If I have understood this correctly, the DOF field tilts forward (in this case), but does not extend, and I will only have that much DOF back/forth of the tilted "rectangle".</p>

<p>Jeff: You know a lot more about this than I probably ever will! At the same time, I acknowledge that I better learn it or get rid of the lens. Thank you for all the linked articles and the detailed explanation; I don't have time right now to read through them, but it certainly looks promising! I'll check back in the evening, and tell you whether I understood something more.<br>

Again thanks!</p>

[jeff_conrad]

<p>Richard, the DoF is wedge-shaped, something like |/_{<big>⁄</big>} ; the vertical bar is the image plane, the first slash is the near limit of DoF, and the second slash is the far limit of DoF. Try to imagine that the two slashes intersect, and that both are tilted more to the right. The tilt sets the distance of the intersection below the camera, the focus determines how far the wedge is tilted to the right, and the <em>f</em> -number determines the angle between the near and far limits (it increases with <em>f</em> -number). Kent may have established a new diagrammatic convention ...</p>

 

<blockquote>... or get rid of the lens</blockquote>

<p>No need. There is a bit to digest, but it's not nearly as bad as it might first appear. It's much easier to describe with a few diagrams than it is with words. I'll see if I can upload a better diagram of the DoF.</p>

[jeff_conrad]

<p>Let's see if a JPEG will display ...</p>

<p>The diagram shows focusing with the rear standard of a view camera, but unless you're doing close-up work, the difference between rear and front focusing isn't significant.</p><div></div>

[2Oceans]

<p>Richard, The surface you photographed is irregular. Did you try stopping down further to see where that gets you? Good hunting. Andy</p>

[paul_kent2]

<p>What Jeff wrote - the dof is wedge shaped. My point about practicing is that you will quickly learn the tilt needed for the scenes you typically photograph.</p>

[bill_stice1]

<p>Richard, a good exercise in learning how to use tilt is to take a regular old flat ruler. Put it on a table. Set the camera up with lens untilted and at about a 45 degree angle to the ruler. Focus near the center of the ruler. Now tilt the lens and watch how the amount of ruler in focus changes. You will be surprised at how little tilt is needed.</p>

[fossum]

<p><strong>Jeff:</strong> I really appreciate your efforts to explain this in even further detail. I think it is sinking in - veery slowly, after reading, and re-reading this thread. I think my best shot from here is to take the lens out this weekend and see if I can gain some understanding on how it works.</p>

<p>With regards to the lens I am using, Nikkor 24 PC-E, it's the newest model, and has a very small DOF-scale.</p>

<p>I'll check back later and report whether I've learned something or if I've thrown the lens downstream in pure frustration.</p>

<p><strong>Paul:</strong> With regards to the aperture used, I wanted to see how well the tilt worked on DOF, while retaining optimal sharpness for this lens (which is somewhere between f/8 and f/11 on the D700).</p>

<p><strong>Edward</strong>: Yes, I tried that without stopping down. The nearest rocks are about 20-30 cm from the sensor (I can't remember exactly), and using f/16 didn't give enough DOF.</p>

<p>Thanks again for helping me understand this (to me) complicated process.</p>

[jeff_conrad]

<p>Richard, as I mentioned, this is a fairly tough scene to get within the DoF, so I don't think you should assess the value of the lens by this experience alone.</p>

<p>The only way I could think of approaching it would be to set the tilt using the near limits of the DoF and see if that works. To me, those limits appear to be the tree on the left and the rock on the lower right, but you'd see right away they really were the right points.</p>

<ol>

<li>Adjust both tilt and focus until those two points appear sharp; this sets the tilt. Note the marked distance (as best you can, anyway); this is the near limit of DoF. My guess is that this would lead to a fairly small value of tilt, placing the apex of the DoF wedge below ground level.</li>

<li>Find the far limit by adjusting focus (don't touch the tilt). It's tough for me to guess where it would be in this picture, but after a couple of tries, it should be fairly easy to see by observing what happens in the viewfinder; you could then confirm it using Live View.</li>

<li>If possible, use the lens distance scale to set the focus in between the two limits; remember that the marked distances probably don't meaningfully relate to distances from the camera. If the scale is too small, I'd try adjusting focus so that the middle of the distant tree on the left (with the browning leaves) is sharp. Again, I'd check with Live View at the greatest possible magnification. With the lens wide open, the rock in the foreground may lose some sharpness, but recall that it's the near limit of DoF, so this is to be expected.</li>

<li>Choose the <em>f</em> -number. If the DoF scale isn't usable, you need to stop down until everything seems to be sharp, checking the DoF limits in Live View.</li>

<li>If <em>f</em> /16 or so doesn't do the job, you might try adding a bit of swing by rotating the tilt a bit to the left (so that the PoF slopes a bit from upper left to lower right). This probably would help with the tree, but might cause problems with the stream. And you might find that you have new near and far DoF limits, and need to readjust tilt and focus. I don't really have a systematic method for determining rotation; when I use it, I do so by trying to envision the best left-to-right slope for the PoF, and check to see if that works.</li>

</ol>

<p>If none of this works, you probably can't get there from here. The last possibility might be to reduce the height, as by using landscape orientation.</p>

<p>Be assured that no one masters using tilt and swing overnight, even for relatively easy subjects. This is a tough scene, even for those experienced in using movements, and about as tough an introduction as I could imagine. I've run into similar situations (flowers in orchards and among oak trees), and sometimes have just given up. But at least for me, those situations aren't typical, so I just move on to a scene that's more cooperative. I think you'll find the same holds for you.</p>

[jeff_conrad]

<p>Someone wrote</p>

<blockquote>I can't say whether you used too much tilt</blockquote>

<p>Well, had he read the OP a bit more carefully, such as noting the bold type, he'd have had an easier time making the call.</p>

<p>You used way too much tilt, as Paul suggested. With 8.5° tilt, the apex of the DoF wedge was 162 mm below the camera, so it would have been impossible to focus so that the PoF was in a good position for both the treetop and the stream. The right tilt for this scene would probably be well below ground; it may seem counterintuitive, but with practice, you'll eventually see that it works. Thinking about the points through which you want the PoF to pass will usually avoid setting too much tilt; using the near limits of the DoF is one way to get pretty close to the proper tilt on the first attempt.</p>

<p>Maximum tilt has an application, but it's usually selective focus rather than landscapes. In most cases, angular DoF decreases with tilt, and that angle can sometimes be smaller than you might think.</p>

<p>This scene is nonetheless challenging, so I'm not sure that using much less tilt would have worked. But it would have made a much better starting point.</p>

[Two23]

<p>Learning tilt is a bit tough at first. I would suggest you start practicing on scenes that are much more 2-dimensional first, such as a flat field, or even short objects placed in a parking lot. Once you learn how to set up for a 2D scene, start trying to find the middle PoF to "lay" the tilt into with taller objects. It was the 3D portions of your photo that were causing you the most trouble. The suggestion of playing with table top compositions is a good one, to learn. Once you learn the idea of it, it becomes more intuitive. </p>

<p>Kent in SD</p>

[fossum]

<p>I said I'd be back, and here I am.<br>

Ok, thanks to your help I am back to loving this lens! While I can't claim to have mastered the tilt shift completely, I'm definitely <strong>getting a grip on it</strong>. You were so right: only a slight amount of tilt is needed. I'd like to get a better feeling on <strong>how tilt and aperture work together</strong>, but I guess I'll learn with time.<br>

I'd like to post some bigger photos, but I am still looking for an adequate service (or I might just build the page myself), so I'll have to post them here (limited size):</p>

<div></div>

[fossum]

<p>Here the forest carpet is slightly off focus, but the leaves above are sharp. The tree on the opposite side is pin sharp.</p>

<p>I just want to thank you all for great answers to my questions, you especially, Jeff. I don't think there is any other forum online where one could find this kind of knowledge, variety and decent conversations.</p>

<p>Now I just have to start saving for something like a 85 TS!</p>

[carl_becker2]

<p>Jeff Conrad, thanks for your help with this question. I have had my eye on one of these for some time. Now I have a bit of knowledge before I pull the trigger. I may just have to do so soon.</p>

[khris_rino]

<p>Its actually not hard. Heres how I do it.<br>

I rack the focus back and forth and compare near and far points. As I'm doing this I'm not attempting to focus accurately but just getting a feel for where the approx focus point is. The point is to adjust the tilt angle till the focus points for near and far points are the closest before getting into critical focussing. If I'm not getting enough dof I then stop down further till everything looks good.<br>

Btw I use a 1.3x crop camera without live view so D700 with full frame + live view should make it waay faster for you. If you've just relied on the LCD I'd also recommend spending a day or two looking through the viewfinder and practising focus. Makes it much easier. Cheers!</p>