24mm Tilt & shift

[paul_russell1]

Finally splurged on one, having fun with the shift, not having so much fun with

the tilt, even slight downward tilting seems to render the top area soft, it's

less of an issue when I come down a couple of stops, but I thought that was the

point of the lens. I know that every situation will be different but do any

users of this lens have any advice on best usage?

 

Please only reply if you actually have this lens or have experience with it,

i've read all the websites and theory, with the best will in the world I need

to speak to somebody who has had some hands on with it.

 

Cheers

[mark u]

What are you trying to shoot? Guessing, I'd assume a landscape - but if you want advice, it helps to know the subject plane orientation you are trying to deal with.

[aginbyte]

I use it almost exclusively for the architectural work on Romanesque churches. Use it 95%

of the time, including seven weeks in France in September and October. Great lens, must

be babied a bit. I use it in two ways, mostly to correct for perspective distortion in tall

structures but also to bring near elements in focus where I have close and far elements in

a shot. A perfect example would be shooting down a cloister from close in. The near

columns will normally be out of focus and the far ones in sharp, but with the shift, it is

possible to keep both sharp. The key to this is to understand the Scheimpflug principle.

What it means is that the plane of focus is at an angle, not parallel, which is what you need

for that kind of shot. Check out Wikipedia for a good explanation. You don't have to do

the math, just understand the principle.

 

It is also very good for selective focus, where you can have foreground and background

out, and just one particular area in perfect focus. This can be done with telephotos, but

with the TS you do it without compression. That can be useful.

 

There is one other great use for the shift, and that is to use it for panoramas, which I love

to do. Lock the camera down on the tripod, shift completely to one side, shoot, then shift

completely to the other.

 

Hope this helps. If you want, ask specific questions.

[howard_slavitt2]

I have this lens, and have previously owned a view camera. The problem with tilting with the lens is that it's almost impossible to see the effects of tilt in the viewfinder. Thus, you need to tilt based on experience (or knowledge) but not based on what you see. In general, I find that very little tilte and stopping down to f11 gives the best results to hold everything in focus. There's a spreadsheet that someone did floating around the web based on Scheimpflug calculations that shows, based on tripod height, how much tilt you need. I think the chart is based on infinity focus. In general, I use that chart, but only use about half of the calculated tilt and stop down to f10 or f11. My procedure is a bit more complicated than that based on a lot of practical tests I ran. If you want to get best results, my recommendation is to search for the chart (you should find it pretty quickly), then run various tests with notes on how much you tilted, focus point, etc., and examine the results to determine a procedure that works for you. . . . It's a special lens, and worth the effort, IMHO. Also Live View may make it easier to use in the future.

[obakesan]

Hi

<P>

I had one for about 10 years ... loved it, sold it 3 months ago. Its hard to say from your post, but as you say the top is going soft, I guess that this means you've not sure how to use it. Like Mark says tell us more about what you're trying to do ...

<P>

as a shot in the dark, remember that the tilts are effectively center based. So as the top moves away, the bottom moves closer.

<P>

Think in terms of a 'plane' of focus moving about the center with the normal DoF ratio changes around the distance from the focal point (thinner when close, thicker when further away)

<P>

[w_t1]

"best usage" = tripod, although I sometimes treat mine as a normal lens without movements and hand hold. Most people use the "going soft" creatively, intentionally....ever seen portraits with the 90 TS lens?

[lester_wareham]

I did some research on the tilt function and wrote it up on a web page, if you don't mind equations and graphs it might be useful.

 

http://www.zen20934.zen.co.uk/photography/tiltshift.htm

[jeff_conrad]

I'll echo Mark's and Chris's comments. I've had a 24 TS-E for a dozen

years, and haven't seen a problem with softness when tilting. Admittedly,

I seldom tilt the lens, because unless I'm extremely close to a foreground

object, I get adequate DoF without tilt.

 

<p>

I also agree with Howard that the effect of tilt is usually tough to see.

My approach is a bit different, though. The proper tilt depends so much on

the specifics of the scene that I haven't found calculating the tilt too

useful. If I can't see the effect at all, I usually assume tilt isn't

needed.

[qtluong]

I'm with Jeff. If you cannot see the effect of the tilt, then it is not needed. When you come into a situation where you need it, what you see in the viewfinder will get you in the proper ballpark, and stopping down to about f11 take care of the rest. Check <a href = http://www.largeformatphotography.info/how-to-focus.html>how to focus the view camera</a>.

The same principles need to be understood and applied with the TS lens.

[robin_sibson1]

I've had a TS24 for some years, and have used it on film, 1.6-factor digital, and FF digital. It is fairly easy to use the shift movement - even hand-held with practice, supplemented with a bit of fine-tuning in post-processing. The main hassle is the need to meter before shifting, so I always use it in M mode. I agree with other posters that the effect of tilt is quite hard to judge - indeed, almost impossible through the viewfinder of a 1.6-factor camera. It is just about possible on FF with a screen with reasonable 'bite', and the Type D screen you might well be using anyhow is a good choice. It's surprising how little tilt is needed to make a difference, and the large depth of field of the 24mm lens at the sort of apertures used with tilt means that it is fairly forgiving, so a rule-of-thumb approach is reasonably effective in many cases. The availability of Live View is probably going to make a big difference here.

[pbizarro]

When you tilt the lens, the focus plane will tilt too, so you may end up with elements in your image that end up "poking in and out" above and below the tilted focus plane (flowers, trees, etc).

 

Therefore, you must stop down the lens, say to f/8. That should be enough to extend the depth-of-field on both sides of the focus plane, so that it covers the elements in the picture.

 

Also take into account that the lens is designed to cover a bigger than normal view (for 35mm), to cater for the shift and tilt. This means that the border of this larger than normal circle will be optically lees than optimum, when compared with the centre. By tilting and shifting, you are using this border area.

[obakesan]

Quang

 

with all due respect, I'll go out on a limb here and say that I suspect that Jeff means that <i>its hard to spot in the viewfinder</i>.

<P>

I've found that it can be veryhard to see on my EOS 630 (which has a reasonable sized screen to see) what's in and out of the plane (especially in dim light), unlike the ground glass of larger formats (which I'm sure you know) one just can't get a loupe on the viewfinder (without an expensive angle finder) to see for sure. Problem is, it shows in the enlargements.

<P>

I've had to do some 'in out back and forth' tilting to be sure where things are, and still found that the enlargments aren't just where I wanted (despite it looking like it was)

<P>

of course if Jeff didn't mean that then .. well what can I say ?

<P>

<i>shrug</i>

:-)

[paul_russell1]

Cheers chaps, guess its a case of overshoot until I get used to its peculiarities. Ta muchly

[andy_jones2]

Here's the table someone referred to above. I put it together as an aid/cheatsheet when I

was learning TS and 4x5!; it makes the assumption that you are on a horizontal plane (e.g. an

airstrip) and you want it all to be in focus to the horizon As noted also, the tops of tall trees

for example, especially close ones, will be above the top of the DoF wedge. Use it only to get

a feel for the magnitude to expect or as a starting point - eventually experience will replace

the need for the table. Even better, read the relevant (BTW excellent) articles Quang-Tuan

referred to !

 

http://www.fototime.com/F4958E37DE63F33/orig.jpg

[jeff_conrad]

I may have been a bit quick to equate “I can't meaningfully apply

it” to “I don't need it.” Perhaps I should have said

“If I can't meaningfully apply it, I'm probably better off not

messing with it.” A little tilt usually goes a long way, and if it's

not set right, a little bit can significantly reduce sharpness. Many

scenes, especially those with height near the camera, are better off with

no tilt.

 

<p>

Seeing what's sharp and what isn't in small-format viewfinders has always

been a problem, with or without tilt. It's not hard to see why: if we

assume the final image is an 8×10 viewed at normal distance and an

8× enlargement of a FF 35 mm image, one is working with a CoC in

the viewfinder that's 1/8 the final-image CoC, so by definition it's

impossible to see what is sharp and what is not. Compare this with using a

4× loupe on a 4×5 groundglass, where one is viewing twice the

final-image CoC. I've never bothered with an angle finder because it only

allows magnifying the center of the image, and I'm usually interested in

the edges. Live View may indeed help, but it's not an option I currently

have.

 

<p>

I may be making this seem more hopeless that it really is. Although it's

tough to see whether something will be acceptably sharp in the final image,

it's usually possible, with reasonable accuracy, to see what is sharpest,

at least when setting focus (<var>f</var>-number is another matter ...).

 

<p>

For a simple situation, such as an inclined camera at a certain height

above the ground focusing on a flat subject, a formula may work better than

a visual adjustment of tilt. However, I've seldom found it quite that

simple in practice. Although it's easy to calculate the relationship

between tilt and the orientation of the plane of focus, the measurements to

determine what the PoF should be can be quite a chore. Moreover, unless

one is photographing a carpet, the orientation of the PoF is only part of

the problem. For a scene that has significant depth in the distance, one

depends on DoF as well as an inclined PoF to get everything sharp. Some

years ago, I wrote a program to calculate optimum tilt and

<var>f</var>-number for a given “cloud” of points, but the

measurements are so tedious that I've hardly ever used it. If Canon were

to incorporate a mini Sinar e algorithm (perhaps with eye-controlled

point selection ...) into a future body, amazing things could be done.

Since they'll probably never even offer DEP mode again, I'm not holding my

breath. Consequently, I'll usually set tilt visually—provided, of

course, that I can see what I'm doing.

[jeff_conrad]

Paul,

 

<p>

You may just need to get a better feel for how to set the tilt. It's

certainly worth looking at diagrams to understand what's happening. I

think Merklinger's are excellent, though for me, his methods seem

unnecessarily complex. Eventually, you need to develop a feel for how

what's shown in side view in a diagram actually looks through the

viewfinder. The key thing to remember is that when you have set tilt, the

focus ring rotates the PoF rather than moving it toward or away from the

camera.

 

<p>

As several others have suggested, I highly recommend reading QT's article

on focusing the view camera, which is probably the most usable introduction

to the subject that I've seen. I have an <a

href="http://www.eosdoc.com/manuals/?q=Tilt-Shift">article</a> that's more

EOS specific (I've just noticed that many of the links need updating ...),

though not as comprehensive.

 

<p>

Finally, there's nothing like actual practice to develop a feel for using

the lens. Start with a simple situation such as Andy's airstrip (or even a

carpet), and practice getting the entire subject plane equally sharp. I

use Howard Bond's method of setting tilt and adjusting focus to check

sharpness; QT describes an approach that sets focus and adjusts tilt to

check sharpness. Either approach should give the same result; pick

whichever works best for you. After a few tries, you'll probably find it

fairly simple, and be ready for some of the additional challenges that QT

describes in his article.

[andrew robertson]

The Canon Angle Finder is a most useful accessory for using a tilt shift lens. Any eyepiece

magnifier should help you see the effect of tilt better, but it also really helps to see it in

your mind as well. This only comes from practice.