Arax Tilt & Shift lenses

On Araxfoto.com, they have generic tilt and shift lenses for a number of

cameras. I'm trying to figure out how useful the lens is, or what quality the

lens it. I'm leerious about the quality as it is a Kiev lens.

 

I can't imagine when I'd use shift as I abhor architecture, but tilt could be

amazingly useful, as I expect to do still life in the future. Any comments?

For nearly $700 US, I myself would spend more and get a TSE lens. The Arax is only a 35mm.

 

I have one of the shift (only) adapters for P6 lenses, but I haven't quite figured out what to do with a 180mm Sonnar on my Canon EOS 20D ;)

The 45mm Mir I have is really too soft to be of much use on a small format camera and that's still pretty long on an APS-sized sensor camera. For now, I (who 'adore' architecture) stick with my nice old Nikkor 35mm PC with an adapter. Reviews of the 24mm TSE are not very encouraging for my purposes.

Even if you don't like architectural shooting, the rear movements are an invaluable tool. For some reason Canon calls it "shift", but it's actually the equivalent of rear tilt on the big camera. Almost every shot I take with a monorail view camera has rear tilt and/or swing, even if only a few degrees. It can give you another POV without making it look like you are looking from an angle. It can correct for minor alignment/leveling issues more easily and attractively than moving the whole camera. It also controls the angle of the plane of focus, as do front tilts/swings. It's very useful for everything, and ideally small format shooters would all be able to have the time and money to use T/S lenses for a lot of our work.

 

I have no idea about the Kiev lenses of which you speak. I'm sure they work well enough for KGB work.

 

Keith

I had an Arsat 35mm f/2.8 PCS lens (shift only) for Nikon. Boy, that was a sharp lens! Don't know how it compares to the tilt/shift lenses, though.

Just a detail, but to avoid confusion:

 

The control called shift on the TS-E lenses is exactly equivalent to a front shift on a view camera. It translates the lens (not the camera's sensor) parallel to the film plane. Tilt is exactly equivalent to front tilt on a view camera. It bends the lens axis relative to the axis perpendicular to the film plane. You use tilt to adjust the plane of sharpest focus (Scheimflug effect). Shift is most commonly used in an attempt to keep parallel lines in architecture parallel. DSLR cameras with TS-E lenses have no movements equivalent to rear shift or tilt on a view camera.

 

Joe

There is another lens other than the 35mm (although it's buried).

 

http://araxfoto.com/specials/tilt-shift-35/

http://araxfoto.com/specials/tilt-shift-80/

 

I'm not really sure when I'd use one from when I'd use the other. I understand the difference distortion and angle of view, I'm just not so sure of when I'd use which.

 

So just to be safe, should I get a long-focal length lens just because I may not always want distortion.

George, how about an Arax tilt adapter for just a few bucks (GBP60 or so) - then you can use a huge range of Pentacon Six mount lenses on your EOS. That's what I'm using, and for the money it's hard to beat. I'm set up at the moment with 45mm, 90mm and 150mm medium format Pent Six mount primes that I use with the tilt adapter, and have spent very little money.

You might also want to look into the Hartblei Super-Rotator lenses.

There are reviews of these at luminous landscape, photozone.de, and elsewhere.

The Hartblei Super-Rotator line is first class. they work great and have excellent optics. I have used it both on 10D and 5D with great results.

 

You do have to know about the theory behind though to use it effectively.

I have the Arax T&S 35mm you can find some sample on my portfolio <a href="http://www.photo.net/photo/5007672"> here </a>

A true view camera shift does not change shapes at all. Rear tilt or swing are the only things that do that on a view camera. Therefore if the "shift" movement on a T/S lens alters the convergence of lines and changes shapes, then it is simulating a view camera's rear tilt or swing; not a view camera's shift movements.

 

Keith

"if the "shift" movement on a T/S lens alters the convergence of lines and changes shapes"

 

Stop trying so hard. Jon F.'s description of the TS-E lens is correct. Shift is the same as front rise/fall (and allows you to not tilt the camera and thus avoid convergence), tilt rotates the lens about an axis that lies in the lens plane.

 

FWIW, the (painfully pricey Zork) adapter gives you +/- 16mm of shift (much of that actually useful) with either Pentax 645 or (but not and) Mamiya 645 lenses on a 35mm camera. This is a lot more shift than Canon (and Nikon) give you, but you don't get tilt.

 

Also FWIW, the Mamiya 645's (manual focus) 35/3.5 on my 5D is as sharp as any lens I've used on my 5D. And the Pentax 645 35/3.5 is supposed to be even better.

Sorry Keith, but when you say

 

"A true view camera shift does not change shapes at all. Rear tilt or swing are the only things that do that on a view camera. Therefore if the "shift" movement on a T/S lens alters the convergence of lines and changes shapes, then it is simulating a view camera's rear tilt or swing; not a view camera's shift movements."

 

your just wrong. I owned and used the TS-E 24 and 90 for about 10 years, and use a rail type view camera now. The shift is just like Joe says its front plane lens shift plain and simple, the mechanism is quite accurately shifting the lens. With the rotation mechanism it can shift in any direction. The tilt is also the same as center swings, but because it can be rotated can be thought of as either swing or tilt.

 

Perhaps it could be called a swing shift lens but then people from the 70's might get the wrong idea about a 'swinging' lens ;-)

 

Dave .. have you tried the 24 on your 5D? The fella who bought mine is very happy with it.<div></div>

detail segments

 

I think there's some abberations showing on the top section, never noticed this on film so strongly. I don't know if this is because the 5D renders images better than film / scanning does or if its to do with the micro lenses not liking the slanted image (a 1D MII owner once mentioned this to me in discussion on the TS-E 24)<div></div>

"(and allows you to not tilt the camera and thus avoid convergence)"

 

Now that IS a key missing ingredient in what Jon said. The "shift" lets you avoid convergence.

 

You still have to "perfectly" square up the initial camera position to the subject, which you definitely do not do in standard practice with a view camera.

 

What I am saying is that to get this with a view camera, the first move is to angle the camera body. THEN you use tilts or swings to effectively achieve ("fake") the net shift that you could not get with direct shift of the standard alone. You almost never just use a shift to keep things square. You almost always angle the camera body, they you use tilts and swings to simulate that net shift. In the end, you don't end up with net tilt and/or swing, but you have used these movements to get to the net shift.

 

In cases with a view camera where you would use just a shift to keep things square, you are in a very mild situation as far as how much you need to alter convergence of lines. Thus my point that if you are seriously "fixing" converging lines with a T/S lens, then you are simulating what you would have to use tilt and swing for on a view camera.

 

Keith

Keith

<P>

Its been said that replying to other than the original poster is a mistake, but

as long as you accept that I'm not making any attack on you or your character here

we'll all be fine :-)

<P>

lets go back to your first two statements:

1 "<I>Even if you don't like architectural shooting, the rear movements are an invaluable tool.</I>"

<P>

and

<P>

2 "<I>For some reason Canon calls it "shift", but it's actually the equivalent of rear

tilt on the big camera.</I>"

<P>

I fully agree with the first statement, and your latter description of technique is of course what

ends up happening, however I can not see how the mechanism of the shift lenses is in any

way different to when I shift the lens of my toho up or down or left or right. I also <u>really</u> can't see how (keeping ones frame of reference on the camera) that it is at all equivalent to <i>rear tilt on the big camera</i>.

<P>

I'd be very interested to see your explanation of how this is so.

<P>

An old friend of mine who was raised on 8x10 cameras used to regularly tell me that my TS-E 24 was no substitute for a view camera, and he is of course right. But I think that he didn't put so much effort into examining these 'tiny' cameras and perhaps misunderstood them. 15 years ago I'd have said that you'd be very hard pressed to get a good quality large size priont out of 35mm, cameras like the 5D have rather changed that, which makes lenses like the TS lenses very handy photographic tools well worth understanding properly.

<P>

<SUB>{side note I've recently sold my TS-E 24 and 90 as I <u>don't</u> have a full frame DSLR and <u>do</u> have a 4x5 with 90 and 180mm lenses. If ever I get a full frame DSLR I will likely be getting those lenses again. I've found that a number of things conspire against me in the field in obtaining the full sharpness potential my 4x5 is capable of, which a DSLR would not suffer from; perhaps yielding better images in the end}</SUB>

<P>

Ok...I think what we are seeing is a difference in how we are describing the approach to movements. I came from a large format camera to small format T/S, and you came the other way if I understand you correctly.

 

The T/S lens shift is mechanically speaking the same as front shift on a view camera. My statement that it is not was factually incorrect. It came from the fact that I never think of view camera shifts in that way. I think of them as minor framing adjustments only. I was making my statement based not on the mechanical aspects of the movements, but on the thought process behind them. Basically, I was [quickly and off the cuff] saying you would use the shift on the T/S lens in situations in which you would use my above-described set up method on a view camera. The fact that a shift is effectively the same thing (although usually practically and physically impossible) is definitely true, although I have never in practical applications thought of it that way.

 

I would be hard pressed to find a view camera shooter who looks at a situation and thinks "I need to place the camera perfectly squarely (and off from the intended final composition) and then use such and such shift movements". Sure. This can work just fine. But I have not heard of anyone ever thinking about it that way in practice. Pretty unlikely unless in a very minor tweak that could almost be done with your eyes shut. The view camera user in practice would probably more likely think: "I need to point the camera at it from this angle, find my rough composition, and then square up my standards to the subject to get the convergence how I want it. Then I will fine tune my framing with shifts using the shift movement built into the standards."

 

So, when this person comes to a T/S lens, they think of it as having the same functionality in the end as an angled rail and rear tilt/swing; which it does. Some, like me, might even call it "the equivalent of rear tilt on the big camera". As far as setup method: no. In effect on the image: yes. Exactly the same. I used the word "equivalent" to make this point. Although the setup method is different in that is DOES start with setting up the camera perfectly square and off composition, the times you would use shift on the T/S lens are the exact times when you would use the above-mentioned angled rail and squaring the standards.

 

When the T/S user comes to a view camera, of course they might view it as doing a shift when correcting convergence, simply due to their usual methods and understandings. After all, it IS a shift in the end. It is a shift that is [almost always] achieved not by using the actual shift mechanisms built into the camera, but by "cheating" the effect of shift by using other movements.

 

In short, six of one, half dozen of the other, and I should not have made such definite sounding statements.

 

And I don't think I'm trying hard.

 

Keith

Keith

<P>

all very interesting (getting to understand how people set up and adjust their cameras that is)

<P>

FWIW (and perhaps confirming your suspicions) I normally approach architecture by first leveling the camera both vertically and horizontally. Then I sort out what I'll do to get the focus on things the way I want. I normally keep the back parallel to the structure and work on shifting the front (and perhaps dropping the rear).

<P>

<img src="http://d6d2h4gfvy8t8.cloudfront.net/6459563-md.jpg" alt="Butai (at Nezu Jinjya)" height="858" width="680" border="0">

<P>

I wonder if this is perhaps created by my experience of 35mm and TS lenses :-)

<P>

I've seen some people with field cameras angle their camera body up slightly, but then tilt the back and the lens forward. To me this equates to a different approach to shifting.

<P>

anyway, its interesting and fun playing with the view camera

<P>

:-)

A simpler way to look at view camera moves is to envision the rear and front planes in relation to the subject. A view camera, for example, can alter the angle of its rear to match the subject plane, in which case it would seem as though you are taking the picture in front of the object when in fact you might be at an angle.

 

There are many movements each with its own use, from correcting converging lines, perspective, focus, DOF, etc... A view camera is the mother of all cameras - I can't wait for an affordable and portable 4x5 Digital view camera :) That'll be the day...

 

A TILT move in a 35mm lens has NOTHING to do with its ROTATION as suggested above (which is a movement ONLY peculiar to some T/S lenses, not to all - some may not offer rotation at all). The ROTATION is a different move altogether.

 

The rotation can simulate the swing on a view camera. For example: if the lens TILTs only downward, rotating it 90 degrees will now produce a swing in either direction, L or R. Because some lenses like the Super Rotator can have multiple lock points within a 360 degree you can achieve some very effective corrections/effects.

Giampi

<P>

<I>I can't wait for an affordable and portable 4x5 Digital view camera :) </I>

<P>

indeed, I've been watching the development of polymer LED sheets with the

hope that one day we'll see 1200dpi arrays of photo diodes on these sheets.

Still modern memory will be challenged by recording 100Meg or so of data

(even assuming 14 bit and bayer array style sensor)

<P>

 

since it was me who mentioned rotation, I think you've misunderstood what

I'm describing (and perhaps the lens). If you orient the lens such that

the <i>lens</i> pivots around <I>its axis</I> left and right that would be swings

would it not?

<P>

Then if you rotate the lens on its mount by 90° so that the lens

now pivots about a horizontal axis (up and down),

that would be a <I>tilt</I> would it not?

<P>

The Canon TS-E 24 has the restriction that the tilt and shift

movements are at 90° to eachother (though this is modifiable to

be parallel, but still not as flexable as the view camera.

<P>

if I'm getting my nomenclature for these movements confused I'm

happy to be corrected.

Chris,

 

Are you saying that you always keep the rail or bed perfectly parallel to the ground, and use nothing but shifts to keep lines from converging? I, of course, start off like you with everything zeroed out.

 

The difference seems to be that my first move is always to get my rough composition by angling the rail whichever way it needs to go; to either side and/or up or down. Then I "correct" as necessary to square things up the desired amount with tilts and swings. Then I reframe with horizontal or vertical shifts of the standards. This way, I don't have to set up the camera way off from the intended composition, and imagine where my eventual movements will put me.

 

The main problem with doing everything with a rail that is always parallel to the deck and only using shifts is that you can easily run into the physical shift limitations of your camera, especially when it comes to lateral shifts.

 

In short, this "shift-only" approach gives you the same eventual result (assuming that the necessary shift is actually physically possible), but the main drawbacks are 1) more difficulty in composition, 2) slower to set up, and 3) less available net shift.

 

Keith

Hi Keith

 

yep, that's what I'm saying.

 

I generally run out of lens coverage first (before I run out of camera movement).

 

If that happens and I still want (say) more rise, then I'll bring the camera back a little and play with the back. But I seldom need more than I've got. (perhaps that's my own imaginative restrictions at work here). I have an old copy of a book called (iirc) "view camera" that I thumb through from time to time which has some handy (simple) reference images of all the movements (and some ABC blocks and other objects) to show perspective.

 

I don't know why the rotation of the lens on the TS-E confuses so many people, its not like rotation of the lens itself has any effect on the image.

 

Sometimes I don't mind the 'rigid camera' parameter and make no movements at all, but not commonly.