For correcting converging lines...

[C_Remington]

Is it customary to tilt the lens or the back? Are there adavantes one technique has over the other?

thank you

[Jac@stafford.net]

Keep it simple to begin. No tilts.

The converging lines (also called keystoning) are caused by tilting the camera up or down. Make certain that the camera is level and front and rear standards are parallel. When the camera is level, and presuming you have keystoning at the top, use front rise to correct.

If you cannot get enough of the subject into the frame because you run out of rise, then you need a wider lens. If you get profound vignetting but have the image in the frame, you need a lens with greater coverage.

In a separate post we can discuss front tilt vs back tilt, and the distortion differences of each.

[E. von Hoegh]

Is it customary to tilt the lens or the back? Are there adavantes one technique has over the other?

thank you

Lots of answers here: http://www.largeformatphotography.info/

[AJ Edmondson]

Probably the simplest explanation is that the back controls image shape while the front controls the plane of focus... obviously if you tilt the camera, the back (film plane) is no longer plumb and the effect noted by Jac occurs. If you are doing architectural work you will (unless shooting for some sort of special effect) want the film plane plumb to render verticals correctly. Tilting the front simply tilts the plane of focus to coincide with the focus plane required by your subject. Vertical front and rear are great if you have a lens of adequate coverage to encompass your subject; if this is not the situation, the front (and/or rear) rise and fall (shift if the shortfall is lateral rather than vertical)... this still does not affect image geometry but does require sufficient lens coverage. If you still can't achieve the desired image you have to resort to tilting the camera and making the necessary adjustments with front and back tilts. Best thing to do is set the camera up and play with the camera adjustments while observing what happens on the ground-glass!

[Brian Ellis]

To eliminate converging verticals (usually when photographing buildings but occasionally other things such as trees) I first level the camera to keep the lens and back parallel to the building. With most buildings I photograph that usually leaves too much foreground and doesn't include enough of the top of the building. So then I keep the camera level and try using front rise. Sometimes that alone does the trick, in which case you don't need to tilt the front or the back, but often it too isn't enough to get the foreground and building top where I want them. So then I aim the camera up and position myself so that I have as much of the top/sky as I want and as little of the foreground as I want. With the camera aimed up (i.e. no longer parallel to the building/trees/etc.) you have to tilt the back to keep it parallel with the building/trees/etc. That eliminates converging verticals. You usually also tilt the front to keep the lens parallel with the building in order to minimize the need to stop way down in order to keep the top and bottom of the building in focus. However, if there are things in front of or behind the building (or trees or whatever) that you want to appear sharp in the image you may need to stop down more to get adequate depth of field.

I hope that answers your question. Basically it isn't a question of whether it's more advantageous to tilt the front or the back to eliminate converging verticals. The only way to do it when when you're pointing the camera up is to tilt the back to keep it parallel with the building. Tilting the lens alone won't do it.

[Jac@stafford.net]

Excellent explanations. Concise, going from the simple to the full answer. You people should be writing the textbooks!


Brian: "With the camera aimed up (i.e. no longer parallel to the building/trees/etc.) you have to tilt the back to keep it parallel with the building/trees/etc. That eliminates converging verticals. You usually also tilt the front to keep the lens parallel with the building"

That's rather like a Super Rise, in effect. :)

[Lenny Eiger]

Parallel is parallel. If you want to take care of converging lines, and you have a tall, flat object in from of you, you can angle the camera back and then tilt the front and back to be parallel to the flat plane. This is the same thing as lifting the front only, while they are both parallel. Many lenses can't handle as far as the camera can physically correct. I find that thinking in planes (as in geometry) helps a lot.

Focus is another matter. If you are looking straight out at a landscape everything within the depth of focus should be in focus. There is a plane parallel to the front and rear, usually perpendicular to the ground, that is the area of critical focus (and will be what is in focus when you are wide open). When people talk about increasing focus, they are often talking about shifting the plane of focus so that its along the ground, or angles from the foreground to the height of the trees or small building, instead of some vertically-aligned plane at a specific distance. Tilting the front lens does not "increase" the focus, it moves it. This can be seen inside the camera with a loupe fairly easily. I tilt the lens on occasion, but its often very little.

There are some people that like to have things out of focus in the image, and have one sharper area, to draw someone's attention. This is a reason to shoot wider open. Diffraction is an almost non-existent factor in the final print. I prefer to have as much in focus as possible, so I chose down to the smaller apertures, f45 or f64. You get to choose your own personal aesthetic.

Hope this helps.

Lenny

[Doremus Scudder]

It's even simpler than the great explanations you've got so far.

1. To eliminate converging parallel lines the camera back must simply be parallel with the lines that you want parallel on the film, either horizontal or vertical. You should be able to see this on the ground glass (use a grid!). That should be step one in setting up a shot like this.

2. This position is usually not the framing you want for your picture, so now frame your image with shifts/rise/fall. If you can't do that (including using "point and swing" techniques to increase your shift/rise/fall), then you need a camera with more movements, a lens with more coverage. Sometimes a picture is just not possible.

3. If you need tilt/swing for any reason after you have the back parallel, use front tilt/swing for adjusting the plane of sharp focus. Any moving of the back will destroy your nice parallel lines.

That's it.

Shoot away and have fun.

Doremus

[Leigh]

f the subject into the frame because you run out of rise, then you need a wider lens. If you get profound vignetting but have the image in the frame, you need a lens with greater coverage.
Another option if you run out of rise is to tilt the camera base or rail up at the front, then reset the front and rear standards vertical. This has the effect of increasing the available front rise.

Obviously this or any lens movement technique requires that the lens have an adequate image circle.
Depending on the exact bellows configuration there might be a vignetting problem with extreme rise.

- Leigh

[Doremus Scudder]

Another option if you run out of rise is to tilt the camera base or rail up at the front, then reset the front and rear standards vertical. This has the effect of increasing the available front rise. ... - Leigh

... and the same if you need more shift; just point the camera to the desired side and reset the standards parallel using the swing movement.

These to maneuvers I call "point and swing/tilt" and are really useful on cameras with limited shift/rise/fall, such as field cameras. I use these movements a lot.

Another trick to get a bit more rise, if you have a camera that allows it and takes off-center Technika boards, is to mount the board turned 180°; the off-center mounting now gives a bit more rise. Very helpful when you need that little extra. I use this technique with my Horseman Woodman often.

Best,

Doremus

[ic-racer]

Are there adavantes one technique has over the other?

thank you

Other than one works and the other does not?

[C_Remington]

Other than one works and the other does not?

So, which is which?

[Leonard Evens]

As other have said, if you want verticals to be parallel, make sure the back is not tilted. Tilting the front has no effect on whether or not verticals are parallel, but it can effect what is in focus. Finally, use rise fall and sideways shift to control waht is in the frame.

[Heroique]

Sometimes, an old photo is worth 1,000 words. :D

Initially, I set-up in front of the camel – but that concealed its shadow.

So I moved my tripod to the left, and turned the camera to the right to face the camel.

This composed the shadow the way I liked – but now the wall’s masonry lines (and the front edge of the camel’s base) converged toward the right in a distracting way.

So I applied back swing (left), bringing my film plane parallel to the wall.

Presto, all the converging lines “corrected” themselves – that is, my back swing “changed” the perspective (that is, if you choose to define perspective this way). DOF wasn’t a problematic issue.

(Apologies for my shadow – the setting sun sneaked it into the scene.)

Toyo 45c
Polaroid Type 55
Fuji A 240mm/9
Epson 4990/Epson Scan

[C_Remington]

As other have said, if you want verticals to be parallel, make sure the back is not tilted. Tilting the front has no effect on whether or not verticals are parallel, but it can effect what is in focus. Finally, use rise fall and sideways shift to control waht is in the frame.

That's what I understood from reading the previous posts. That last post was distracting. What a nasty man.

[Leigh]

These to maneuvers I call "point and swing/tilt" and are really useful on cameras with limited shift/rise/fall, such as field cameras.
They're not limited to field cameras.

My 8x10 Sinar F2 has limited rise/fall. Apparently they expect you to tilt the base rail as needed.


...if you have a camera that allows it and takes off-center Technika boards, is to mount the board turned 180°
That works fine as long as the mount does not require the board to be bevel-down.

- Leigh

[Doremus Scudder]

... That works fine as long as the mount does not require the board to be bevel-down. - Leigh

Yeah, I know. I have Wistas and a couple other field cameras that don't allow upside-down mounting of the lensboard and with which I can't get that bit of extra rise using the offset.

I've always wanted to design a camera that took perfectly square lensboards with an offset so that one could mount the lens in any orientation and use the offset to augment the rise/fall/shift movements. Maybe someday I'll design or modify a front standard anyway...

Best,

Doremus

[Jim Jones]

Both the head and the lens stage on my ancient DeJur Versatile Professional 4x5 enlarger can be tilted, permitting distortion correction during printing. On other enlargers, I've shimmed one side of the negative carrier and tilted the easel to accomplish this.

[Andrew O'Neill]

I've always prefered a little key-stoning. Building tend to look odd to my eye when the convergence has been corrected.

[ic-racer]

As other have said, if you want verticals to be parallel, make sure the back is not tilted. Tilting the front has no effect on whether or not verticals are parallel, but it can effect what is in focus. Finally, use rise fall and sideways shift to control waht is in the frame.

Everything you need to know to set up a view camera in that post. Really! I'm amazed at how people can write a whole book on it :)

[Bob Salomon]

"Probably the simplest explanation is that the back controls image shape while the front controls the plane of focus... "

Sorry but this is not totally correct.
Both the front AND the rear tilts and swings control the plane of sharp focus (Scheimpflug). Only the back controls converging lines. The shape of the object is controlled by the back as well as foreshortening caused by the lens focal length.

[Leigh]

I've always prefered a little key-stoning. Building tend to look odd to my eye when the convergence has been corrected.
I'm the same way.

Unless the camera position is opposite the mid-point of the building, full-square looks weird.

- Leigh

[C_Remington]

I'm the same way.

Unless the camera position is opposite the mid-point of the building, full-square looks weird.

- Leigh

How is it weird? That's how you see it in real life?

[Leigh]

How is it weird? That's how you see it in real life?
That's only true if you're standing on the roof of a building half the height of the subject building.

If you're viewing a building from the ground, you see keystoning, worse with taller buildings.

- Leigh

[Doremus Scudder]

How is it weird? That's how you see it in real life?

Actually, the image the lens of the eye projects on the curved surface of the retina is crazy distorted compared to a projection on a flat surface.

90% of what we think we see is a construct of our perception; i.e., our brain stitching together images from our eyes and "correcting" them according to our wiring, experience and understanding (or not... that's when it's an optical illusion). Our perception of perspective and parallelism is largely a construct of our minds, not a literal representation of the image formed by our eyes.

And, there's kind of a spooky feedback loop that happens when we use our perception guide our use of movements to "correct" an image in camera and then make a two-dimensional flat print which our brains have to re-interpret to get any three-dimensional information out of... Our minds are wonderful things.

Best,

Doremus

[andrew gardiner]

How is it weird? That's how you see it in real life?

I'm not sure how we see it in real life is relevant in this instance. Eliminating (or at least reducing) converging lines is part of the process of composing a final,two-dimensional picture. This has all sorts of it's own concerns such as the subjects relationship to the edges of the image etc.