Zuiko 35mm f/2.8 shift VS PC-Nikkor 28mm f/3.5. Which one would you buy?

[LolaColor]

Same thing; two different names. Some lens companies call the shift lenses PC lenses.

In that case, you might explain to me how the suggestion to point the camera at the horizon in order to achiever vertical verticals is "counter to the use of a PC lens". Perhaps one of us is using it incorrectly.

 

[Sirius Glass]

In that case, you might explain to me how the suggestion to point the camera at the horizon in order to achiever vertical verticals is "counter to the use of a PC lens". Perhaps one of us is using it incorrectly.

Fpr an non-PC lens set the offset to zero and use the lens just like any lens of the same focal length as you described. If a PC lens was used and tilted off level, the lens would be offset to achieve a composition with the vertical parallel building lines again appearing parallel. Is that clearer?

 

[LolaColor]

To anyone reading this thread who isn't aware of how a shift lens should be used, but is interested in finding out, I'd suggest reading the link I posted earlier.

 

[LolaColor]

If a PC lens was used and tilted off level, the lens would be offset to achieve a composition with the vertical parallel building lines again appearing parallel

Are you confusing tilt (changing the plane of focus) for shift by any chance?

Or are you talking about some format other than 35mm?

In your example, if we assume you're using shift on a 35mm camera, the verticals won't be vertical. Tilting the camera up means the verticals will converge and shifting the lens is only going to move the image circle. It won't straighten those verticals for you.

All of the shift lenses I've used, Olympus 35, Pentax 28 and Canon 24, work in the way I have described. Or are there some "perspective control" lenses that work in a way I'm not familiar with?

 

[Sirius Glass]

Are you confusing tilt (changing the plane of focus) for shift by any chance?

Or are you talking about some format other than 35mm?

In your example, if we assume you're using shift on a 35mm camera, the verticals won't be vertical. Tilting the camera up means the verticals will converge and shifting the lens is only going to move the image circle. It won't straighten those verticals for you.

All of the shift lenses I've used, Olympus 35, Pentax 28 and Canon 24, work in the way I have described. Or are there some "perspective control" lenses that work in a way I'm not familiar with?

I am talking about 35mm cameras and using PC lenses designed for them. Why would you thing otherwise?

In your example, if we assume you're using shift on a 35mm camera, the verticals won't be vertical. Tilting the camera up means the verticals will converge and shifting the lens is only going to move the image circle. It won't straighten those verticals for you. Exactly. Then the convergence goes away when the lens is appropriately shifted. I never mentioned tilt which is exactly not what a PC does.

 

[LolaColor]

Then the convergence goes away when the lens is appropriately shifted.

That's not the case at all. Try it and you'll see!

 

[Sirius Glass]

That's not the case at all. Try it and you'll see!

I did and I do and it works. I just suppose that you know more about optics than the lens manufacturers. I can't wait to tell them. They will be so so disappointed.

 

[Mike Lopez]

LolaColor, you are correct. Shifting a lens won’t correct the convergence for a camera already pointed upward. Tilting the lens is how you correct the verticals in that case, but I suspect you already know that.

 

[Besk]

Some way or another you both are describing the same thing. Shift and "perspective control" are the same thing -- just different names.

 

[Craig75]

Nikons "pc" and shift are not quite the same thing as a modern nikon pc lens has both tilt and shift.

 

[Mike Lopez]

Some way or another you both are describing the same thing. Shift and "perspective control" are the same thing -- just different names.

Not really. LolaColor is differentiating between shift and tilt, which are two separate methods of “perspective control.”

 

[Craig75]

Shift isn't just perspective correction either. You can use it for perspective distortion too

As Mike says both tilt and shift are controls

 

[Mike Lopez]

Shift isn't just perspective correction either. You can use it for perspective distortion too

As Mike says both tilt and shift are controls

Yes, the "PC" is most aptly thought of as perspective control, and one can really go wild with it.

But contrary to the dismissive and condescending remarks which LolaColor was receiving, LolaColor was correct. Shift can't control everything, nor can tilt. And in the example given, shift won't correct the converging verticals.

 

[MattKing]

in the example given, shift won't correct the converging verticals.

Shift can, however, sometimes be used to get the right image on the film, when positioning the camera in a way that prevents converging verticals, also results in the top (or bottom) of the subject being outside the frame.

 

[Mike Lopez]

Shift can, however, sometimes be used to get the right image on the film, when positioning the camera in a way that prevents converging verticals, also results in the top (or bottom) of the subject being outside the frame.

Yes, of course. But that wasn’t the hypothetical example given. The example problem was to point the camera upward and to try to correct the verticals in that case. You’re gonna need some tilt…

 

[Sirius Glass]

Shift can, however, sometimes be used to get the right image on the film, when positioning the camera in a way that prevents converging verticals, also results in the top (or bottom) of the subject being outside the frame.

The PC lens like every other lens has limitations. One can choose to deal with the limitations or seek another solution. For most photographers, the shift alone in the PC lens is enough to get the photograph. Is it the same or as good as with a camera will all the movements, no, and the manufacturers never made that claim.

 

[LolaColor]

LolaColor, you are correct. Shifting a lens won’t correct the convergence for a camera already pointed upward. Tilting the lens is how you correct the verticals in that case, but I suspect you already know that.

Yes indeed, I do!

My original reason for commenting here was to offer some advice to the OP trying to figure out how he would ensure straight vertical lines on buildings when shooting handheld with a shift lens (neither lenses mentioned by the OP have a tilt function). Anyway, it looks like he knows what he's doing and is happy with the results so that's all good.

To recap, what I would do in that situation is, point the lens at the horizon when in an unshifted position, then shift up to achieve the desired composition. A hot shoe bubble level would be perfect for this, if it weren't for the slight difficulty that you can't see the bubble while looking through the viewfinder! Regardless, when you first use a shift lens it can be a bit confusing, especially if the lens is already shifted off-center. So my approach has always been to get the camera aligned correctly first while unshifted, as this is what determines whether the vertical (and horizontal) lines will converge or not, and then to shift. Here's an illustration of that. Note the drainpipe on the left of the image is vertical in both shots:

Next consider the usage that Sirius Glass suggests. Now, bear in mind that it may be the case that I have misunderstood him so far in this thread. If so, I apologise in advance for that. But to the best of my knowledge this is an accurate illustration of what he claims he "did", "does", and that "it works".

Again, note the orientation of the drainpipe on the left. It is a converging vertical in all images, regardless of the degree or direction of shift applied. This is because the film or sensor plane is not parallel with the plane of the wall, and the reason it's not parallel is, of course, because I pointed the camera up. Shifting the lens moves the image circle but won't correct for converging perspective.

But if it's a lens that also has a tilt function, now that we're on the subject, I would still rely only on shift while keeping the camera level for architectural work. My understanding is that tilt is best reserved for manipulations of the focal plane. To give two examples, ensuring that one side of an angled box is in focus for product photography, or the opposite effect could be to create a miniature world effect in landscape work. These manipulations of the focal plane are probably not desirable outcomes in classic architectural photography, but might be of interest as a novelty effect, although I guess that at small apertures the effect won't be pronounced. In any case the amount of perspective correction that can be achieved via tilt only is small in comparison what can be done with shift, at least on the tilt-shift lens that I have used, the Canon TS-24. But it's certainly interesting to consider that if I'm at the limit of the shift movement and I'm still just a little bit short of bringing the top of the building into view, one option could be to point the camera up slightly and dial in a bit of tilt on the lens. Although on that Canon the tilt is limited to one axis only so I don't think it's possible to shift and tilt in the same direction.

On another aside, and to give a little love to the Olympus 35. Yes, if you were to have only one shift lens then wider would certainly be preferable to 35mm. However, I've often found 24mm and 28mm to be too wide for some applications and I use the 35mm focal length a lot. The Olympus 35 is a lovely lens, I think, and for a few different reasons. The image quality is good when stopped down. It's tiny. But best of all is that there are no knobs or twiddling to be done. You just grab the lens and slide it freely in whatever direction you wish. Yes, the shift mechanism can get a bit loose but I seem to remember there are screws that can be tightened somewhere to firm it up. Failing that, a piece of insulation tape applied between the sliding faces adds enough friction to hold the position, and that's what I did with mine.

 

[Bronson Dugnutt]

Yes indeed, I do!

Great exploration and exhibition of the lens's movements. A great scene for b&w film you have there but I can understand why you'd be compelled to include some greenery

Is there any advantage to rise compared to lifting the camera upwards by the same distance?

 

[Mike Lopez]

Yes indeed, I do!

My original reason for commenting here was to offer some advice to the OP trying to figure out how he would ensure straight vertical lines on buildings when shooting handheld with a shift lens (neither lenses mentioned by the OP have a tilt function). Anyway, it looks like he knows what he's doing and is happy with the results so that's all good.

To recap, what I would do in that situation is, point the lens at the horizon when in an unshifted position, then shift up to achieve the desired composition. A hot shoe bubble level would be perfect for this, if it weren't for the slight difficulty that you can't see the bubble while looking through the viewfinder! Regardless, when you first use a shift lens it can be a bit confusing, especially if the lens is already shifted off-center. So my approach has always been to get the camera aligned correctly first while unshifted, as this is what determines whether the vertical (and horizontal) lines will converge or not, and then to shift. Here's an illustration of that. Note the drainpipe on the left of the image is vertical in both shots:

View attachment 292285


Next consider the usage that Sirius Glass suggests. Now, bear in mind that it may be the case that I have misunderstood him so far in this thread. If so, I apologise in advance for that. But to the best of my knowledge this is an accurate illustration of what he claims he "did", "does", and that "it works".

View attachment 292286

Again, note the orientation of the drainpipe on the left. It is a converging vertical in all images, regardless of the degree or direction of shift applied. This is because the film or sensor plane is not parallel with the plane of the wall, and the reason it's not parallel is, of course, because I pointed the camera up. Shifting the lens moves the image circle but won't correct for converging perspective.

But if it's a lens that also has a tilt function, now that we're on the subject, I would still rely only on shift while keeping the camera level for architectural work. My understanding is that tilt is best reserved for manipulations of the focal plane. To give two examples, ensuring that one side of an angled box is in focus for product photography, or the opposite effect could be to create a miniature world effect in landscape work. These manipulations of the focal plane are probably not desirable outcomes in classic architectural photography, but might be of interest as a novelty effect, although I guess that at small apertures the effect won't be pronounced. In any case the amount of perspective correction that can be achieved via tilt only is small in comparison what can be done with shift, at least on the tilt-shift lens that I have used, the Canon TS-24. But it's certainly interesting to consider that if I'm at the limit of the shift movement and I'm still just a little bit short of bringing the top of the building into view, one option could be to point the camera up slightly and dial in a bit of tilt on the lens. Although on that Canon the tilt is limited to one axis only so I don't think it's possible to shift and tilt in the same direction.

On another aside, and to give a little love to the Olympus 35. Yes, if you were to have only one shift lens then wider would certainly be preferable to 35mm. However, I've often found 24mm and 28mm to be too wide for some applications and I use the 35mm focal length a lot. The Olympus 35 is a lovely lens, I think, and for a few different reasons. The image quality is good when stopped down. It's tiny. But best of all is that there are no knobs or twiddling to be done. You just grab the lens and slide it freely in whatever direction you wish. Yes, the shift mechanism can get a bit loose but I seem to remember there are screws that can be tightened somewhere to firm it up. Failing that, a piece of insulation tape applied between the sliding faces adds enough friction to hold the position, and that's what I did with mine.

Your photographs illustrate the error in Sirius’ claims as well as any of the numerous books, diagrams, etc. can. Don’t be dissuaded by all the bluster!

I have a FlexBody, a 4x5, and some other bellows-based setups, and now this thread is making me want to pick up some PC lenses for my Nikons! Damn you!

Cheers.

 

[Sirius Glass]

Yes indeed, I do!

My original reason for commenting here was to offer some advice to the OP trying to figure out how he would ensure straight vertical lines on buildings when shooting handheld with a shift lens (neither lenses mentioned by the OP have a tilt function). Anyway, it looks like he knows what he's doing and is happy with the results so that's all good.

To recap, what I would do in that situation is, point the lens at the horizon when in an unshifted position, then shift up to achieve the desired composition. A hot shoe bubble level would be perfect for this, if it weren't for the slight difficulty that you can't see the bubble while looking through the viewfinder! Regardless, when you first use a shift lens it can be a bit confusing, especially if the lens is already shifted off-center. So my approach has always been to get the camera aligned correctly first while unshifted, as this is what determines whether the vertical (and horizontal) lines will converge or not, and then to shift. Here's an illustration of that. Note the drainpipe on the left of the image is vertical in both shots:

View attachment 292285


Next consider the usage that Sirius Glass suggests. Now, bear in mind that it may be the case that I have misunderstood him so far in this thread. If so, I apologise in advance for that. But to the best of my knowledge this is an accurate illustration of what he claims he "did", "does", and that "it works".

View attachment 292286

Again, note the orientation of the drainpipe on the left. It is a converging vertical in all images, regardless of the degree or direction of shift applied. This is because the film or sensor plane is not parallel with the plane of the wall, and the reason it's not parallel is, of course, because I pointed the camera up. Shifting the lens moves the image circle but won't correct for converging perspective.

But if it's a lens that also has a tilt function, now that we're on the subject, I would still rely only on shift while keeping the camera level for architectural work. My understanding is that tilt is best reserved for manipulations of the focal plane. To give two examples, ensuring that one side of an angled box is in focus for product photography, or the opposite effect could be to create a miniature world effect in landscape work. These manipulations of the focal plane are probably not desirable outcomes in classic architectural photography, but might be of interest as a novelty effect, although I guess that at small apertures the effect won't be pronounced. In any case the amount of perspective correction that can be achieved via tilt only is small in comparison what can be done with shift, at least on the tilt-shift lens that I have used, the Canon TS-24. But it's certainly interesting to consider that if I'm at the limit of the shift movement and I'm still just a little bit short of bringing the top of the building into view, one option could be to point the camera up slightly and dial in a bit of tilt on the lens. Although on that Canon the tilt is limited to one axis only so I don't think it's possible to shift and tilt in the same direction.

On another aside, and to give a little love to the Olympus 35. Yes, if you were to have only one shift lens then wider would certainly be preferable to 35mm. However, I've often found 24mm and 28mm to be too wide for some applications and I use the 35mm focal length a lot. The Olympus 35 is a lovely lens, I think, and for a few different reasons. The image quality is good when stopped down. It's tiny. But best of all is that there are no knobs or twiddling to be done. You just grab the lens and slide it freely in whatever direction you wish. Yes, the shift mechanism can get a bit loose but I seem to remember there are screws that can be tightened somewhere to firm it up. Failing that, a piece of insulation tape applied between the sliding faces adds enough friction to hold the position, and that's what I did with mine.

Basically you are first aligning with horizon and then rotating on only one axis. I start by standing vertically and only rotation on the horizontal axis [looking up at a building] or vertical axis [looking lengthwise at a building]. So we are really saying the same thing, six of one, half dozen of the other, three pairs, or two trinities.

 

[Craig75]

Has anyone used the laowa 15mm shift lens?

I look at it and think hmm... interesting but I'm on the fence

 

[LolaColor]

I start by standing vertically

Well, that sounds like a good place to start. Now, I'm going to go and stand horizontally for a while. That's quite enough nonsense for one day!

 

[Neil Grant]

...Nikon PC 35mm f/2.8

, probably @ f/11. Fuji RDP2