Portrait perspective issues investigated

[Ulophot]

Greetings, everyone. I have just registered, having come upon the site in the past week. Although I am not able to be actively photographing at present, I have for decades and look forward to dusting off my camera and D2 eventually.

The following is, I think self-explanatory, and I hope some may find it useful.

In the course of recently thinking through a return, after years, to making some large-format (4x5) portraits, but without the present ability to run film tests with the camera, I began looking up focal length, distance, and proportion information. The reason for my pursuit, was that I have but one lens for my 4x5, a 210mm, which is roughly equivalent to a 63mm on a 35 and a 110mm on my 645 medium-format camera. For a portrait of a single subject in these smaller formats, I would nearly always choose a lens at least 40-60% longer, unless photographing full standing figure in tight quarters. Since the planned portraits will often be ½- and ¾-length seated poses, I needed to get a better idea of how close I should consider getting to keep distortion within rather strict bounds, and how much cropping of the negative might be necessary at the distances required.
By distortion I mean the visible change in relative size of parts of the face or body as the camera moves closer to the subject. Since the 1960s especially, we have become used to constant distortion from TV news cameras and journalists up-close with very wide angle lenses. Most people today are barely even aware of it, though in my youth it was startling to most.
A web search for "portrait focal length comparison" provides numerous examples of facial distortion -- series of portraits of a subject showing frames with consistent head size taken with short to long focal length lenses (i.e., from extremely close to distant), and some others involving the figure, but I wanted a better measure of proportion for hands, knees, an extended foot, for the range of poses I had in mind. The portraits I intend are conceived with Classical painting and late-19th-, early 20th-Century studio photographic portraiture in mind. Ken Rockwell’s interesting dictum of a 15-foot minimum is not an option for me; I have neither the space nor the inclination to crop my 4x5 negative that severely, in addition to which, the degree of facial “flattening” and broadening at this distance may not suit all subjects well.
Not finding what I wanted, I set up a simple test. I cut two pieces of stiff card stock in half and folded each piece, producing four identically sized, self-standing cards. Then, I sat in a chair and roughly measured the following distances along an imaginary axis parallel to the floor, straight out in front of me: from the middle of my ear to the tip of my nose; from my nose to a comfortably extended hand, as on a chair arm or table; from my nose to my shoe tip on my crossing (upper) leg. The distances in inches are about 5.5, 15, and 30. I then placed the four cards those distances apart along the edge of a long table. Then, measuring from the nose card, I marked off 2-, 3-, 6-, and 9-foot distances and photographed the cards from each distance with a borrowed digital SLR. (See attachment for the ear and nose cards at 2 feet.)
Now, please note that my measurements were all approximate; that’s all they needed to be. Nonetheless, as the attached chart shows, the results are useful in helping to quantify the effects of perspective. I measured the relative sizes of the cards in the several images by using Photoshop’s ruler tool on the height, then derived the ratio of heights of the nose to the ear card, the hand to the nose card, and the toe to the nose card, at each distance.
The attached table shows the Photoshop ruler distances and, below in each row, their ratio (rounded off). Thus, for example, from about 2 feet away, the nose card is 1.25 times taller than the ear card—in practical terms, the front of the nose appears 25% larger than its actual size relationship to the ear, and somewhat larger in relationship to the eyes and facial planes, while the frontal planes of the face will also be disproportionately wider than receding ones. From 3 feet, the card is still about 14%larger than the ear, but from 6 feet, only 6%. However, from 6 feet, the hand card is still about 25% taller than the nose card, and even at 9 feet, the toe card is nearly 50% taller than the nose card; thus, my crossed leg would be noticeably disproportionate. Naturally, these all assume a frontal view; the ratios tend to diminish as the face and body turn towards profile, especially at 6 feet and more.
For my purposes, the results suggest that I may be able to photograph 6 feet from my subject’s face if I keep hands and elbows within a one-foot or so radius from the body and avoid straight-on poses that include knees. I will have to crop the negative but not extremely. At 8 feet, the drawing of the figure will improve and allow a bit more freedom. Using radical.org’s angle-of-view calculator, I see that at 8 feet, a vertically oriented film back will take in 5 feet, a bit more height than a typical full-length seated portrait.
Naturally, the gesture of the pose and other compositional—including tonal—factors may also affect the appearance of relative size, the perception of prominence. However, the exercise has been helpful in providing a more informed starting point.

[brucetaylor]

The 210mm on your 4x5 should work great for portraits. That's the easy part. It's everything else that makes it challengingl

[Mark Barendt]

There is an easier way.

A photo naturally "looks normal" when the angle of view the camera saw, matches the angle of view that the viewer sees the print at. (Cropping is the equivalent of using smaller film stock behind the same lens; it changes the angle of view.)

So, set up your camera in the house looking at the wall that you want to put the print on. The lens (when focussed properly on the wall) should be the same distance from the wall that your face would be when normally viewing the print.

Now look through the camera and use 4 post it notes to define the corners of what the camera can see on the wall. With that lens, that is the size the print should look normal.

If you crop then the print size needs to be reduced accordingly to maintain a normal look. Mask off the ground glass to the cropped size and move the post it notes accordingly, don't move the camera. If you want a bigger print, than what your post it notes define, and want it to look normal, then you need a different focal length lens.

The specific focal length doesn't matter, as-long-as the viewing and taking angles are the same the print should look normal.

[Mark Barendt]

Actually I need to amend my post above, the camera should theoretically be focused the same as you would be shooting.

My destruction above would be about right for a headshot, maybe head and shoulder; a portrait where the lens to subject matches the lens to wall distance I described.

If the subjects will be say, twice that distance from the camera for the portrait, then the camera should be properly focussed for that subject. The lens to wall distance for the exercise though would remain the same.

[jb7]

... By distortion I mean the visible change in relative size of parts of the face or body as the camera moves closer to the subject. Since the 1960s especially, we have become used to constant distortion from TV news cameras and journalists up-close with very wide angle lenses. Most people today are barely even aware of it, though in my youth it was startling to most. ...


I think you're describing 'perspective', though it is not unusual to see it described as 'distortion' when applied to wide angle lenses used close to the subject.

For what it's worth, I think it's possible to use shorter focal lengths on large format than their infinity equivalents in smaller formats, because of the extra draw needed when focused closer.


For a portrait of a single subject in these smaller formats, I would nearly always choose a lens at least 40-60% longer, unless photographing full standing figure in tight quarters.

Sometimes moving to a new format with a new lens forces you to work within your new limitations. Regarding how much you might need to crop, well that's probably a question best answered by your negative, but I would doubt if it will be as much as you think.

[Ulophot]

jb7, yes, perspective is more correct; apparent distortion would be a better term.

Mark, I have read about this, but have difficulty applying it in practice. My prints will be about 10x 12.5 inches; printing regularly on 16x 20 is not really feasible in my darkroom.
Therefore, whether I photograph my subject from about 6 feet, including from about the lower knee up to offer moderate headroom with my 210, or from about 8 or 9 feet, to include the feet, my print size will stay about the same, and I suspect the comfortable viewing distance for most will not exceed 18-24 inches, since the head will be considerably smaller than a head-and-shoulders composition. I don't know how that works out with regard to normal perspective appearance; my experience simply tells me that if, as I indicated, the foot is relatively 40-50% larger than the face in a "classical" composition such as described, it will tend to look disproportionate.

[DrTang]

10 feet is good.. 10-12 feet

[mdarnton]

I think the OP is thinking WAY TOO HARD! Get a camera, use the lens you have, see if you like the results you are getting; if not, change your approach. That is all there is to it. No one can predict with numbers what you will find that you personally prefer the best.

The old theoretical portrait lens was film length + film width. Is there anyone using an 18" lens for 8x10 portraits? Karsh didn't--he seems to have done fine with a 14" Ektar. The studio I worked for in high school used a 210mm lens for 4x5. No one complained that it was a half-inch short. Neither of these combinations result's in anything near even a ten-foot distance.

None of the lenses used by most LF photographers begin to approach the proportions of the normal-for-portraits 90mm lens on 35mm, either--that would be a 27" lens on 8x10, or about 350mm for 4x5! Shooting LF is much different from 35mm. With 8x10 portraits you are approaching what would be macro ratios in 35mm, but at greater distances, still with the very large extensions of macro photography*, resulting in a lens of effectively longer focal length. The way to understand how this works is to dig in and do it.

*(an 8x10 head+ shot might be shot at 1:2, with a 14" lens being used at 21" from the film, where the same shot on 35mm is done with a negligible focus extension, at a very different reproduction ratio.)

[DrTang]

Is there anyone using an 18" lens for 8x10 portraits?



*Raises hand*

[Mark Barendt]

Morning Ulophot,

The disparity between the geometry of the shot and the geometry of the print view, is what introduces the bulging or flattening look/perspective to any photo. The rules of the universe don't care about the limits of our darkrooms or cameras or how big our studios are. ;)

If the eye is in the right place (has the same view the camera did) the print will look normal regardless of what the head vs foot measurements are.

If the foot looks too large when viewing the print I'm going to guess that the viewing distance to the print is simply too long, get your nose closer or print bigger and see if it looks more normal. If that works then to fix the problem you may need to move the subject further from the camera and crop or ...

[Ulophot]

I do appreciate fully that numbers do not a composition make. We are artists, not visual accountants. As I indicated in my OP, I was trying to think through a question, not having the ability to test with the camera, trying to confirm or adjust my impressions from experience in (mostly) other formats using what seemed a simple approach to provide guidelines.

Thanks to all who have taken the time to respond.

[Mark Barendt]

You are welcome, Ulophot. You are working on the right questions whether it makes sense yet or not.

Getting the general idea of how perspective works is an important thing. I remember the first time I was told the perspective was is essentially governed by the distance from the camera to the subject and that it didn't matter what the focal length of the lens was. It just did not compute.

Think about this way. When you look at the scene you have the same perspective as your camera but you can see a lot more of the World around you, then the camera can; the camera is just cropping the scene down.

So if you imagine straight lines running from the corners of the film, then converging in the lens, and then continuing out into the scene you can get a reasonable idea of what the angle of view actually might be. When the corners of your print matches those lines, the angle of view when holding the print in your hand, then the perspective is normal, the subject matter will look normal, as if you were there. It is my opinion that when you hear someone describe a photo as looking almost 3-D, they have found exactly the right viewing distance.

[Ulophot]

Not to belabor my point, but I just came across the Southworth and Hawes portrait of John Quincy Adams seated, frontal, leg crossed, http://images.metmuseum.org/CRDImages/ph/original/DT1666.jpg. His pose tends to match my description of the crossed leg/foot issue. This image appears to have been taken from perhaps 15 feet, guessing from the perspective. I don't know what focal length the gentlemen had for their Daguerreotype camera(s).

[Alan Gales]

Not to belabor my point, but I just came across the Southworth and Hawes portrait of John Quincy Adams seated, frontal, leg crossed, http://images.metmuseum.org/CRDImages/ph/original/DT1666.jpg. His pose tends to match my description of the crossed leg/foot issue. This image appears to have been taken from perhaps 15 feet, guessing from the perspective. I don't know what focal length the gentlemen had for their Daguerreotype camera(s).

Probably a "normal" lens for that format.

If you look at Karsh's work, as mentioned earlier he used a 14" or 360mm lens on his 8x10. Avedon used a 360mm on his 8x10 for his American West series. A 360 is equal to a 180 on 4x5 so a 210mm would be even better. A lot of 4x5 photographers use a 210mm for portraits.

Try your 210mm and if you are not happy with it then swap it for a 240 or 250mm. A 300mm is nice for head shots or head and shoulder shots.

http://www.karsh.org/#/the_work/portraits

http://www.americansuburbx.com/2011/01/richard-avedon-richard-avedons-in.html

[appletree]

I know this might not be the place, but it is sort of relevant. 18" lenses, 14" lenses, etc? Is this the same as 210mm, 360mm lenses, etc? Just metric vs standard? If it is a long answer, then feel free to just link me somewhere.

And also, will 210mm on 4x5 be fine for full body portrait work as well as chest up? If I have someone slouched over a chair would I be better off picking up my yet-to-purchase 150mm?

[jp]

If you've got room, a 210 is fine for full body, it'll have thinner DOF and less background than the 150.

Newcomers; get shooting is my advice to analysis questions.

[Mark Barendt]

I know this might not be the place, but it is sort of relevant. 18" lenses, 14" lenses, etc? Is this the same as 210mm, 360mm lenses, etc? Just metric vs standard? If it is a long answer, then feel free to just link me somewhere.

And also, will 210mm on 4x5 be fine for full body portrait work as well as chest up? If I have someone slouched over a chair would I be better off picking up my yet-to-purchase 150mm?

Yes just metric vs ...

The biggest difference between using 150, 210, and a 360 for a full length portrait is how far the subject will be from the camera. The longer the lens, the further the subject needs to be from the camera just to fit in the frame.

If you're working in a short room, you need a short lens.

If you're using a short lens, a larger print will typically make the print look more normal and vice versa for any given print viewing distance.

[appletree]

First off, thanks for the information.

Second off, sorry, I should know this after shooting for 5+ years. But TBH my 80mm stays on my Hassie 95% of the time, as does my 50mm Summicron on my Leica. I guess first I need to fully understand "short lens". Not merely the size of the lens, but the relationship of the focal length of the lens vs the film being used...correct?

One additional question is, if I understand you correctly, then generally if using a short lens (ie 75mm, 90mm, etc in 4x5) a print will look more normal (I assume to the human eye/viewer of the print?) when printed on a large scale. Whereas when using a longer lens (210mm, 360mm, etc in 4x5) a print will look more normal when printed smaller. But this also has to do with distance between the viewer and the print?

Thanks again for the help and explanation.

[Mark Barendt]

You bet apple tree.

Normal perspective comes when the camera's angle-of-view matches the print viewers angle-of-view.

Take your Hasselblad with the 80 and point it at a print on your wall. Match the edges of the View finder frame to the edges of the print. When your eyes view the print from the same distance as the camera does the perspective will be normal, the angles of view will match. If you put a wider angle lens on the camera and leave the camera in exactly the same spot, it would take a larger print to fill the cameras view. The distance from your nose to the print would be the same in either case and both prints would look normal.

[appletree]

Mark, got it! Thanks.

And that leads me to the conclusion on how I should start using the "right tool for the right job"...hence swapping out lenses depending on subject matter and it's distance. I need to do some research on rules of thumb to generally get normal perspectives. The distance matters in real life because that is how you make the "print" larger. I think. Or perhaps it merely means I now have more "coverage" (a word in LF I should use loosely) of my scene. Drats, its been a long week. I will research it. Thanks for the help.

I mean I obviously understand the amount of "print" you will see depending on a lens chosen. That said I am unsure if the point of view, aspect ratio, normal perspective, etc is also affected by this lens selection, in addition to the basic premise that you can fit more scene from the same tripod distance (nose to print) with a longer lens.

Well, I guess I don't get it. Don't bother wasting time on me, I will do some homework and try to grasp it fully. It is difficult sometimes though, because details can be overwhelming/too in depth, even for my analytical/engineer mind.

I will try and read through this! http://en.wikipedia.org/wiki/Perspective_distortion_%28photography%29

[appletree]

Now I got it! I think using the example of print messed me up, because at first I was thinking actual print from a negative. Enlarging to an 8x10 or 16x20, etc.

from wikipedia:
"Note that linear perspective changes are caused by distance, not by the lens per se – two shots of the same scene from the same distance will exhibit identical perspective geometry, regardless of lens used. However, since wide-angle lenses have a wider field of view, they are generally used from closer, while telephoto lenses have a narrower field of view and are generally used from farther away. For example, if standing at a distance so that a normal lens captures someone's face, a shot with a wide-angle lens or telephoto lens from the same distance will have exactly the same linear perspective geometry on the face, though the wide-angle lens may fit the entire body into the shot, while the telephoto lens captures only the nose. However, crops of these three images with the same coverage will yield the same perspective distortion – the nose will look the same in all three. Conversely, if all three lenses are used from distances such that the face fills the field, the wide-angle will be used from closer, making the nose larger compared to the rest of the photo, and the telephoto will be used from farther, making the nose smaller compared to the rest of the photo."

[Alan Gales]

You can get in close with a normal lens for a portrait. You just have to be careful with the pose to avoid distortion. A longer lens makes it a lot easier. The longer the lens the more you flatten the subject. What focal length looks best? Well, you learn from experience what you like. It really is a personal decision which is why I recommended to the OP to try his 210mm and if he wasn't happy then try something longer.

[Mark Barendt]

If you want the print (yes the 8x10 or whatever) to have normal perspective the print viewing angle needs to match the taking angle.

[Dan Quan]

Morning Ulophot,

The disparity between the geometry of the shot and the geometry of the print view, is what introduces the bulging or flattening look/perspective to any photo. The rules of the universe don't care about the limits of our darkrooms or cameras or how big our studios are. ;)

If the eye is in the right place (has the same view the camera did) the print will look normal regardless of what the head vs foot measurements are.

If the foot looks too large when viewing the print I'm going to guess that the viewing distance to the print is simply too long, get your nose closer or print bigger and see if it looks more normal. If that works then to fix the problem you may need to move the subject further from the camera and crop or ...

I have never heard this but it makes sense in my ignorance. Can you explain it again more concisely and possibly more simply please?

[Mark Barendt]

I have never heard this but it makes sense in my ignorance. Can you explain it again more concisely and possibly more simply please?

So I'm going to blow up the scale here little bit to see if I can clarify the idea.

Take any camera & lens and take a photo. No cropping. Blow the photo up really big, say printed on the garage door of a two car garage.

Set up your camera in front of the garage door the move the camera forward and back until you have the print filling the viewfinder edge to edge.

When you're standing at the camera the print will look normal.

If you go straight forward closer to the print, the subject matter in front of you, will start to look flattened.

Back up behind the camera ways and the print will start to distort and look "wide angle".

This technically works for any lens. In a practical sense it is the most visible when a true wide-angle lens has been used.

In our example a camera with a wide angle lens might end up 20 feet from the garage door. At that point it's easy to get in front of or behind the camera.

When a long narrow angle Lens has been used to take a photo it is tough to get far enough back away from the print to see the effect change much, it always looks flattened.

Think about shooting a 300 mm lens on a 35mm camera and printing on the garage door, now use that same camera and lens and backup until you get that print framed perfectly in the viewfinder. The camera might be back a hundred feet or more from the print.

[Sheldon N]

Shooting LF is much different from 35mm. With 8x10 portraits you are approaching what would be macro ratios in 35mm, but at greater distances, still with the very large extensions of macro photography*, resulting in a lens of effectively longer focal length. The way to understand how this works is to dig in and do it.

*(an 8x10 head+ shot might be shot at 1:2, with a 14" lens being used at 21" from the film, where the same shot on 35mm is done with a negligible focus extension, at a very different reproduction ratio.)

+1

This is a significant issue for shooting close up portraits, even with 4x5 but especially with 8x10. When you focus close, you extend the bellows. When your bellows extend, the field of view of your lens narrows. By the time you get to 1:1 magnification (ie. 4x5' or 8x10" subject size at plane of focus for 4x5 or 8x10 film) you have doubled your bellows extension and cut your field of view in half. On 4x5 with a 210mm lens that would mean that you would have 420mm of bellows extension and would have an effective field of view (effective focal length) of 420mm. It's obviously a sliding scale for any magnification between infinity and 1:1 macro. Here's a quick rule of thumb to help estimate the effect.

Your "effective" focal length is roughly the same as the amount of bellow extension used.

[Mark Barendt]

Your "effective" focal length is roughly the same as the amount of bellow extension used.

Yes.

The focal length marked on a lens is only right when focused at infinity.

Part of the struggle in explaining "my" concept above is this variation in focal length to focus. When framing the print on the garage door in my example I have to ignore focus. The camera needs to be focussed like it was focussed when shot, not focussed on the garage door.

[cowanw]

Funny, I had a dream about this last night. It's a clever argument and often repeated as the justification for using shorter lenses in ULF, but, I wondered if the laws of optics of a 420 mm lens are the same as the laws of optics of an "effective 420mm lens" i.e. a close focussed 210mm lens. I should like to see some evidence of this, before I necessarily buy into this oft repeated statement. Especially, since, as the field of view is cut off, the lens must move closer to the subject to fill the frame (as in close up portraits)


+1

This is a significant issue for shooting close up portraits, even with 4x5 but especially with 8x10. When you focus close, you extend the bellows. When your bellows extend, the field of view of your lens narrows. By the time you get to 1:1 magnification (ie. 4x5' or 8x10" subject size at plane of focus for 4x5 or 8x10 film) you have doubled your bellows extension and cut your field of view in half. On 4x5 with a 210mm lens that would mean that you would have 420mm of bellows extension and would have an effective field of view (effective focal length) of 420mm. It's obviously a sliding scale for any magnification between infinity and 1:1 macro. Here's a quick rule of thumb to help estimate the effect.

Your "effective" focal length is roughly the same as the amount of bellow extension used.

[Mark Barendt]

Especially, since, as the field of view is cut off, the lens must move closer to the subject to fill the frame (as in close up portraits)

This sentence doesn't make sense.

The angle of view for specific lens remains constant, as the lens moves farther from the film the projection at the film gets larger, more coverage. This is the reason that coverage is also measured normally at infinity Focus.

If a 420 and 210 lens both had the same angle of view, then both would cover the same area when focused at the same bellows draw, what would not be the same is the focus.

What Sheldon was getting at though is the limitation of the camera if all I can hold is 4 x 5 film moving from 210 to 420 change is the angle of view that the camera can see, it does not change the optics of the lens but it does change where the plain of sharp focus falls.

[mdarnton]

Why so much talking? This is really easy to demonstrate:
A 12" lens is a bit shorter than 8x10 diagonal, which is about 12.5". 35mm diagonal is about 42mm, so a 42mm lens would be a hair longer equivalent on 35mm film to the 12" on 8x10, but it's close enough. If the two function equally at portrait distances, they should cover the same field when the lenses are at the same distance from the subject, right?

I just set up my 8x10 with a 12" lens pointed at my book case at portrait distance so that the vertical field of view was two shelves---30 inches. The total bellows extention is 15.5 inches, which is conveniently close to equivalent to 50mm on 35- approximately diagonal plus 20%, in both cases. The extension of the 50mm at this distance is, of course, negligible. If the case under discussion is true--that it's not the lens but the extension--then the 50mm (which is a long lens on 35mm by the normal rules) at the same distance (same lens to subject distance) should be cover the same as the 12".

And it does. A 35mm lens, which isn't too far from being a 12" equivalent, is wildly wider--another shelf and then some more. That makes a 12" on 8x10 about equal to 50mm on 35mm film. I suspect that with longer lenses on 8x10 the situation becomes even more in favor of these lenses acting like their extension, not their focal length.

The important point is that perspective = field of view compared with eye to subject distance. All the rest is extraneous.

You can do this at home, if you have a camera. :-) In fact, someone who's more fussy than I am should do it and report back.

[cowanw]

Quite right. Poorly worded.
What I was trying to indicate was that to fill the picture at a larger format size with the same size lens requires the lens to be positioned closer to the subject. Despite this, it is commonly said that ULF images do not need longer (calculated as normal) lenses, that the virtual extension makes it a longer lens.



This sentence doesn't make sense.

The angle of view for specific lens remains constant, as the lens moves farther from the film the projection at the film gets larger, more coverage. This is the reason that coverage is also measured normally at infinity Focus.

If a 420 and 210 lens both had the same angle of view, then both would cover the same area when focused at the same bellows draw, what would not be the same is the focus.

What Sheldon was getting at though is the limitation of the camera if all I can hold is 4 x 5 film moving from 210 to 420 change is the angle of view that the camera can see, it does not change the optics of the lens but it does change where the plain of sharp focus falls.

[cowanw]

Yes to this (I think). I apologize if I am too dense to fully understand.
but again what I was addressing was the equivalent of the question of taking your 50mm lens and using it as a portrait lens for 8x10. Perspective must change (as the lens moves) yet upsizing this question to 20x24 changes things, according to LHF lore (apparently because the putative length of the lens changes.
It would be much easier for me to see and comparison example.
What I would really like to see is two same full face subject photos done with say a 360mm lens on 8x10 and 20x24. Then both printed to 20x24.
I am trying to think how I could test this with max 8x10 gear.
The 8x10 full face is no problem.
If I measure a 1 inch object and then I move in to make it 2.4 inches, for a second photograph, I will have a 8x10 section of a 20x24 print.
Then I enlarge the 8x10 full face to make the 1 inch, become 2.4 inches and compare the two.


Why so much talking? This is really easy to demonstrate:
A 12" lens is a bit shorter than 8x10 diagonal, which is about 12.5". 35mm diagonal is about 42mm, so a 42mm lens would be a hair longer equivalent on 35mm film to the 12" on 8x10, but it's close enough. If the two function equally at portrait distances, they should cover the same field when the lenses are at the same distance from the subject, right?

I just set up my 8x10 with a 12" lens pointed at my book case at portrait distance so that the vertical field of view was two shelves---30 inches. The total bellows extention is 15.5 inches, which is conveniently close to equivalent to 50mm on 35- approximately diagonal plus 20%, in both cases. The extension of the 50mm at this distance is, of course, negligible. If the case under discussion is true--that it's not the lens but the extension--then the 50mm (which is a long lens on 35mm by the normal rules) at the same distance (same lens to subject distance) should be cover the same as the 12".

And it does. A 35mm lens, which isn't too far from being a 12" equivalent, is wildly wider--another shelf and then some more. That makes a 12" on 8x10 about equal to 50mm on 35mm film. I suspect that with longer lenses on 8x10 the situation becomes even more in favor of these lenses acting like their extension, not their focal length.

The important point is that perspective = field of view compared with eye to subject distance. All the rest is extraneous.

You can do this at home, if you have a camera. :-) In fact, someone who's more fussy than I am should do it and report back.

[jb7]

Yes to this (I think). I apologize if I am too dense to fully understand.
but again what I was addressing was the equivalent of the question of taking your 50mm lens and using it as a portrait lens for 8x10. Perspective must change (as the lens moves ...

I think you have to imagine a camera that focuses by moving the rear standard.

Putting the lens for each format in the same position relative to the subject is what controls perspective; moving the ground glass until it is co-incident with the in-focus image is a good way to focus without altering perspective in that close up zone.

[jb7]

ok, maybe I didn't read far enough...

yes, perspective will change as the lens moves- the purpose of this comparison is not to compare the same lens on two different formats, but to compare lenses that will give the same perspective on different formats. Since the convention for measuring the length of lenses is at infinity, and not at close up distances, it's impossible to directly correlate the equivalence of two optical systems except at infinity, or close to it, if there is significant magnification involved.

Those who say they would like to use the equivalent of an 85mm lens on 35mm for close up portraiture on an 8x10 camera might find themselves wondering why they don't have enough bellows, if they get a lens that produces an equivalent angle of view at infinity...

[cowanw]

Quite right, regarding your comments, jb7.
It is the other that I can't get my wee brain around.

[Mark Barendt]

I think there are a few important side notes that should be made here.

First, the perspective of the taking lens is for the most part defined by where it is and where it's pointed. It does not matter what film size is behind the lens. A 4 x 5 is just a crop of an 8 x 10 which is just a crop of a 16 x 20. The prospective relationships stay the same, the angle of view changes (the crop).

Second, is that the instant that you turn a three dimensional scene into a two dimensional representation, the prospective relationships in the two-dimensional representation is fixed. The variable that controls whether or not when you see distortion is our relationship to the two-dimensional print or negative. What we call print viewing distance, although I'm not sure that properly defines the concept.

[cowanw]

Yes and yes.
Sheldon N brought up and I am discussing the close up portrait, wherein the position of the lens is changed to occupy the film size with the same image.
The flip side of perspective coin as it were.

[Sheldon N]

Yes, another term for it is "focus breathing". Filmmakers complain about it, since the field of view of their shot changes as they pull focus. Some complicated lenses with internal focusing can avoid it, but since we're using a simple lens design and bellows focusing in large format we get to experience the full effect. It's also why macro photographers tend to use focusing rails instead of just focusing the lens. Each time you refocus you change your framing and angle of view, it's simpler to just move the whole camera on a focusing rail than it is to focus/reposition/focus/reposition/focus.

To get into some of the lens comparisons mention above (50mm on 8x10, 360mm on 8x10 vs 20x24, crops, etc) , I'd suggest thinking about it this way....

1) How big is your negative?
2) How big is your subject?
3) How much magnification is this going to result in?

Once you know the magnification (and the focal length of your chosen lens), you can figure out how much bellows extension the shot will need. Then this will give you a rough idea of your effective focal length. You can then compare that to your format size to see what your field of view is.

An example....

My negative is 8x10 inches
My subject is 8x10 inches (tight headshot)
My resulting magnification is 1:1
I'm using a 50mm lens, so I need 100mm of extension to get to 1:1.

100mm field of view on 8x10 (if the lens could cover 8x10) is still a super wide angle. Your picture will be very distorted because you'll have to be super close to the subject to fill the frame. Replace the 50mm lens with a 360mm lens, and you end up with 720mm of extension and an effective field of view a lot like a 100mm lens on 35mm film. You'll have a pleasing perspective because you had to pull the whole camera back to frame the shot.

As you go smaller in format, less magnification is needed for a given picture so there's less bellows extension and less of an effect on the field of view.

[jb7]

... It's also why macro photographers tend to use focusing rails instead of just focusing the lens. Each time you refocus you change your framing and angle of view, it's simpler to just move the whole camera on a focusing rail than it is to focus/reposition/focus/reposition/focus.

...

This is not a universal truth, it relates only to cameras in which the film plane is fixed, and the lens is moved in order to achieve focus. Using a large format camera which can focus by moving the film plane, these reiteration a are not necessary, though they might form part of the picture taking process anyway. Sometimes it's not possible to achieve focus at all by moving the lens position only, hence the need to move the whole camera as a unit.

[Sheldon N]

Yes, lenses with internal focusing that are free from focus breathing solve that problem. It's the bellow style or true extension style macro lenses that are more difficult.

Focusing with the rear standard solves the problem that focusing with the front standard has, where you are pushing the lens toward your subject and working against yourself when you actually need to back the lens up. It doesn't totally get rid of the iterative process of focus/reposition, because there's only one correct distance of lens to subject for a given magnification/framing and you need to find that somehow.

[Old-N-Feeble]

Perspective distortion is a matter of distance from lens to subject. It doesn't matter what film format, provided the same ratio, e.g. 1:1.25. It doesn't matter what lens focal length other than acquiring the desired angle of view. For head-and-shoulders portraits most folks seem to prefer 2x-3x focal length so for 135 film, cropped to 1:1.25 ratio, thats a 85-135mm lens. For 4x5 format that's a 300-450mm lens. This is if the subject is framed closely and the image isn't cropped during enlargement. Of course, this begins to become impractical fairly quickly so one will probably need to compromise much of the time.

Regarding print viewing distance being a major factor in how one lens focal length appears vs. another: My personal opinion is that's largely bunk. I don't care how closely I view a print (within reason) of a portrait taken with a lens that's far too wide the image looks distorted.

[Mark Barendt]

Perspective distortion is a matter of distance from lens to subject.

No.

That is an urban myth based on a fairly arbitrary tradition of viewing distance.

Essentially I believe this tradition comes from viewing prints hand held.

Pick up a 2x3 print and you'll probably hold it pretty close (given average subject matter), 4x6 not quite as close, 5x7 bit further, 8x10 further again, and so on. I think the root of "normal" came because "normal lenses" made prints looked normal hand held.

That "standard of normal" isn't a technical standard, it's a variable, it's a tradition.

[Old-N-Feeble]

I disagree, Mark. What I stated is "most folks prefer" and this is true for the most part. Nearly anyone will agree that a head-and-shoulders portrait shot with a lens of .5x FL instead of a lens of 2x FL (neither cropped during printing) will look extremely distorted no matter what viewing distance.

[Mark Barendt]

Print that wide angle shot 30"x40" and hand hold the print yourself to view it and you may get surprised.

[Old-N-Feeble]

Print that wide angle shot 30"x40" and hand hold the print yourself to view it and you may get surprised.

Most folks don't view a 30x40 inch print at 18 inches. Some do but 'most' don't.

[Mark Barendt]

I agree that most don't.

It is though the reason that there is visible distortion, the wide angle look is because people don't normally print wide angle shots big enough for the viewing distance that the print will be seen at.

What I'm suggesting though is that we can apply this principle to make our prints look better to the observer. If all we have is a 150mm lens for our 4x5 but we like the look we have seen with 210mm lens on a 4x5 then we might just want to enlarge a bit more when printing. You are right in suggesting that if the distance to the subject is constant the perspective the camera remains. If we keep the same size paper the cropped print from a 150mm lens mimics the un-cropped 210mm lens print almost perfectly. This assumes the prints are viewed from the same distance.

What changes when we change focal length lenses is the magnification and angle of view. The magnification can be adjusted when printing. We can also make a bigger print from a wider angle lens (by not cropping to match the view of the 210 in this case) and have the subject/face remain the exact same size just have more context around the edges.

The distortion only comes when we change the viewing distance to the print.

[Ulophot]

This has been, for me, a fascinating discussion, covering in considerable depth the subject of print viewing distance, which certainly is a factor. Given the many the variables of the process from photographing to displaying, it seems to me to be something one should bear in mind but, with all due respect to Mark's and others' precision, somewhat broadly.

Reviewing Ansel Adams's later series of books a week ago in the course of pondering various LF considerations for future portraits (I am, regrettably, still not able to re-engage in actual film photography), I came across his portrait of Brassai (The Camera, pg 112) taken with the normal, 80mm lens on a Hasselblad. In the caption he states that he had moved in fairly close "but not enough to cause distortion," and that he cropped the negative somewhat in printing. The face fills about 2/3 of the print vertically, so he was indeed pretty close.

This led me to run a quick series of self-portraits with a digital video camera so that I could compare frame-grabs side by side, with the camera at 5, 7, and 9 feet. The results were instructive, so I will indicate my procedure for any who might wish to replicate the experiment in one medium or another. I simply sat at a table with hands folded about a foot in front for a frontal shot series. Leaving one arm on the table, I turned about 45 degrees and looked back toward the camera a t a slightly lesser angle, for a second series. I used the zoom lens to take each pose at each distance twice: one including hands and head, one zoomed in to a head shot.

The differences in the drawing, or perspective, are appreciable in all the frames. For the kinds of portraits I have in mind to make (see original post), I shall tend to stick with the longer distances. However, I have learned from this exercise more about the means of using perspective that may be applied with subtlety on one or the other subject. Hence, my view towards using my one 4x5 lens, the 210mm, for closer framings has widened.

[Alan Gales]

These may help you.

http://mcpactions.com/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/

http://digital-photography-school.com/how-focal-length-changes-the-shape-of-the-face-in-portraiture/


I remember when I started in photography back in the early 1980's with a 35mm camera. Everyone said for a close portrait you needed an 85 to 135mm lens. I was taking a portrait of my then girlfriend in my car. My 100mm lens was too long to use in the cramped space so I used my 50mm. The photograph came out great and everyone commented on how pretty my girlfriend looked in the photograph.

You can get away with a normal lens up close if you are careful but it is easy to get distortion like large noses. A longer lens is better. How long depends upon your taste.

[Old-N-Feeble]

Alan... Did you crop your girlfriend's portrait to 8x10 format? If so, a 'normal' lens for 24x30mm film being 38mm... and assuming you printed the full frame width... and the 50mm lens you used was indeed precisely 50mm FL, then that was like shooting a 201mm lens on a 4x5 camera. If you cropped the width a tiny bit (most likely) then it was probably closer to 210mm on 4x5. Many so-called 'normal' 50mm lenses for 135 cameras are up to 3mm longer than 50mm. So... if you cropped to 8x10... and cropped the short side a tiny bit... and your 50mm lens was really a 53mm lens... then that would have been like shooting with a 225-250mm on 4x5.

Was that a waist up shot, head-and-shoulders or face close-up?

[Alan Gales]

The shot was not much more than head and shoulders. I shot it with a 50mm f/1.7 Zeiss lens on my Contax. The original print was printed at 3x5? Wasn't that the size before 4x6 became popular? I also had a 5x7 enlargement made.

We were in my 1973 Dodge Dart Sport front seat so we were pretty close.

Man, 3x5 prints and a Dodge Dart. That was a long time ago! I remember when 4x6 was called super size and you had to pay extra. The only reason I remember the shot was because I broke the so called rules and learned something.

I will say this. My girlfriend was pretty but she was extremely photogenic.

[Jac@stafford.net]

Alan... Did you crop your girlfriend's portrait to 8x10 format? If so, a 'normal' lens for 24x30mm film being 38mm...[...]

:) It's been a while since I thought in terms of 135 film, but the diagonal (nominal normal) is 43mm.

An issue with normal lenses is that the usual lens is not distortion-corrected for close-ups.


I will say this. My girlfriend was pretty but she was extremely photogenic.

Yeah! Post the picture or it never happened. :) Oh, and a shot of the car, too!

[Old-N-Feeble]

:) It's been a while since I thought in terms of 135 film, but the diagonal (nominal normal) is 43mm.<snip>

...unless 135 is cropped to 4x5/8x10 format which only uses 24x30mm area of the film (at most). :) However, apparently very little cropping was done as 3.5x5 inch prints use most of a 135 format frame.

[Alan Gales]

Yeah! Post the picture or it never happened. :) Oh, and a shot of the car, too!

Boy, you're tough! This was back when I was 21. I'm now 53 and I no longer have the 19 year old girlfriend, the car, the camera or the photographs. ;)

[Jac@stafford.net]

Boy, you're tough! This was back when I was 21. I'm now 53 and I no longer have the 19 year old girlfriend, the car, the camera or the photographs. ;)

Well, I'm 70 and I still have my old '58 VW Bug (http://www.digoliardi.net/58bug.jpg). Still working on the 19 year-old girlfriend thing. It ain't lookin good.

[Alan Gales]

Well, I'm 70 and I still have my old '58 VW Bug (http://www.digoliardi.net/58bug.jpg). Still working on the 19 year-old girlfriend thing. It ain't lookin good.

Yeah, but a 58 VW Bug is pretty neat. A 73 Dodge Dart not so much! I do wish I had held onto my 67 Chevy short bed pickup though.

Good luck with the 19 year old girlfriend. I think you have to be a Rolling Stone or famous movie star to get that. ;)


By the way, that is one nice looking car you have!