Ideal distance from subject for candid/environmental portrait in terms of perspective?

<p>I think what some of us are getting at, is that what lens to use doing an environmental portrait is totally dependent on what effect you want to achieve. Just google a page of Arnold Newman portraits. He is one of the most well known environmental photographers. You'll see a variety of approaches from close up wide angle with distortion, to more traditional, to even torn up and re-assembled images. From my own body of work:<br /> Here's one where I used a 24mm lens on 35mm and close up, for a dramatic effect and to include a lot of background. http://www.photo.net/photo/4283534&size=lg<br /> 50mm lens (35mm format) very close, again with distortion on purpose http://www.photo.net/photo/2451837<br /> 24mm lens (35mm format) full length to get more in: http://www.photo.net/photo/2451876<br /> 105mm for soft background http://www.photo.net/photo/17478348&size=lg<br /> 200mm for ultra detail and flatter effect http://www.photo.net/photo/6378693<br /> 35mm on DX close up--still no distinguishable distortion but a feeling of closeness http://www.photo.net/photo/6046475<br /> The point is we're talking about an artistic medium. The effect you want dictates the tools you use, <em>not the other way around! </em></p>

<p> </p>

 

 

<blockquote>

<p>Jeff, isn't what your calling a graphics camera simply a large format camera? I'm sincere in my question. I have no experience what so ever with anything other than SLR style and compact cameras, but what your describing sounds a lot like what I imagine a normal LF camera to be. Sorry, I know it's off subject.</p>

</blockquote>

<p>Are you talking to Jeff or me, Siegfried? My name is Tim.</p>

<p>Did you see the linked picture of the Argyl copy camera I posted? There are room sized versions of similar camera design and yes, it's an overgrown LF camera only it comes with its own lights, is more flat field (distortionless) and can't be taken out to do landscapes. </p>

<p>Some versions of that camera were used to create 4 color separations for commercial presses by using a filtering system much like a camera's Bayer sensor. The results of the seps were 4 grayscale pieces of film for each CMYK color used to burn press plates. This was before laser scanner color separators came along. For the seps to line up on press required a flat field lens to be used to do the separations.</p>

 

 

 

 

<p>Steve, can you explain why your images of the little girls don't look distorted as much as my shots of the old man? Is it because I used an 80mm zoom lens? Did you use primes on those shots and do you believe that has an affect on the lack of distortion?</p>

<p>Yeah, sorry Tim, I was addressing you.<br>

I seen the pictures after I had posted my question. I could swear I've seen pictures of what seems to be in principle the same thing as in your links except that everything is mounted to a horizontal beam and the lens and film were both tiltable in both directions.<br>

What do you mean the camera has it's own lights? As in flash?</p>

<blockquote>

<p>If you've got visual proof this level of specificity isn't necessary please post so we can all be informed.</p>

</blockquote>

<p>Ok, so let me uderstand. What you're saying is, rectlilinear nature of the projected image is somehow less rectilinear on the edges of the image? And it has nothing to do with perspective? I'm sorry to ask, but were you paying enough attention at those perspective drawing lessons you had to take? Things get stretched, that would be your basic projection of a threedimensional space onto a flat plane at work here, wouldn't it? What you demonstrated with your photos is exactly perspective distortion at work. Yeah, nose gets bigger in relation to the farther parts of face, because guess what? You got closer to your subject. I.e. your perspective of the subject just got more distorted. Here, for your viewing pleasure an image shot with two lenses of the same focal length (50mm). One of them got full benefits of Lightroom profile, the other is dumb, connected using lens adapter thingy:<br>

<img src="http://imageshack.us/a/img6/176/4w33.jpg" alt="" /> <img src="http://imageshack.us/a/img42/2471/doin.jpg" alt="" /></p>

<p>Good luck trying to discern them based on the degree of lens made distortion.</p>

<blockquote>

<p>I seen the pictures after I had posted my question. I could swear I've seen pictures of what seems to be in principle the same thing as in your links except that everything is mounted to a horizontal beam and the lens and film were both tiltable in both directions. What do you mean the camera has it's own lights? As in flash?</p>

</blockquote>

<p>If you examine closely the linked images there are four small box like aluminum flash housings (two top/two bottom) close to the copy board that function as continuous (recommended 5 sec. max duration) lighting. Each of those house (the model I used, not the one pictured) a 4in. high wattage halogen glass tube. All four combined put out a ton of light which is required due to f22 aperture as instructed. It uses a fixed (non-zoom) prime lens.</p>

<p>Lens and exposure platen can't be tilted which if it did would ruin flat field geometric accuracy.</p>

<blockquote>

<p>What you're saying is, rectlilinear nature of the projected image is somehow less rectilinear on the edges of the image?</p>

</blockquote>

<p>No, I'm saying different lens design builds within their glass groupings and spacing across a variety of lenses will/may distort with a bowl like shape when the subject's head falls within a central area of the lens (not the sides) with an effect similar to parabolic security mirrors in convenient stores given a particular focal length & distance to subject combo.</p>

<p>What else explains the distortion differences of the two shots of the old man I took at 80mm just moving back a couple of feet and yours and Steve's don't exhibit any distortion? This may have something to do with cheap zoom kit lenses is what I'm getting at, not perspective induced distortion.</p>

<blockquote>

<p>Things get stretched, that would be your basic projection of a threedimensional space onto a flat plane at work here, wouldn't it?</p>

</blockquote>

<p>Not that pronounced when your moving only a couple feet near/far in front of the subject. Compare that to noticing why the horizon looks flat when we're on Earth but starts to look curved ascending 1000's of miles in an air plane. That's a big move to just stepping back a couple of feet to get more of the subject in frame.</p>

<p>As for your posted image sample, yeah, I don't see any differences, but my point of WHY lens profiles are used and how I've seen them change the preview in adjusting parabolic like distortion when applied is till valid. Have you seen how they build lens profiles? They use a grid pattern as a target.</p>

<p>Here's two "selfies" I shot that show the differences of parabolic distortion with my 18-55mm kit lens below. The one on the bottom makes me look like a child.</p><div></div>

<p>Here's another more obvious example of parabolic distortion from my playing around shooting "selfies" with a magnifying glass I picked up at a thrift store.</p><div></div>

<p>Tim, the shot up close of the little girl with chocolate on her face was done with a 35mm prime lens on a DX camera (Nikon D80) which translates to something like a 50mm on a full frame camera. I don't know why there doesn't seem to be much distortion. </p>

<blockquote>

<p>What else explains the distortion differences of the two shots of the old man I took at 80mm just moving back a couple of feet and yours and Steve's don't exhibit any distortion? This may have something to do with cheap zoom kit lenses is what I'm getting at, not perspective induced distortion.</p>

</blockquote>

<p>What do you mean, what else? Nothing else, that's just it, perspective distortion. The two feet difference equals probably 15-20% of the distance. You have to think about distortion in relative terms, not absolute. Take a look at these two:</p>

<p><img src="http://imageshack.us/a/img854/81/f9qt.jpg" alt="" width="800" height="450" /> </p>

<p><img src="http://imageshack.us/a/img69/3558/l9ns.jpg" alt="" /><br>

These are two shots from the same sitting (obviously), the top one being a kind of digital polaroid for the bottom one. Lighting setup was slightly changed inbetween, as was the crop.<br>

They were shot with two different cameras and lenses, from almost the same spot - there's about one foot difference. Both lenses cover exactly the same horizontal angle of view, so in terms of perspective they are same thing. Both are very high quality lenses, the first one exhibits very slight barrel distortion which was anyway corrected with lens profile in LR, the second has no distortion whatsoever (or if it does, then it's immeasurable, so it's like there is none at all). Notice how dramatic difference in perspective distortion is introduced by twelve inches of distance. Same thing in your selfies. 45mm and 50mm that's ten percent difference right away. Plus the 45mm was shot closer up (it had to, there's no other way for your head to end up bigger when shot with shorter lens).<br>

Anyway, this leads nowhere. In my portraitist's experience, and I'd call my experience substantial, lens distortion as a factor in portrait distance is negligible. Perspective distortion is the thing you should think about if you're going to care about distortion at all. But even then it's not a big deal if you're not going to use fisheyes or superteles. What you need to think about is your subject and your light. All the rest is fluff.</p>

<blockquote>

<p>All the rest is fluff.</p>

</blockquote>

<p>So you're saying we wasted our time in this discussion and learned nothing of value in knowing how to get good looking portraits and finding out for sure what can ensue in preventing that from happening? I don't agree with that.</p>

<p>I think in reading your responses and how you explain yourself in reaching your point you're more concerned about coming across as an authority on the subject rather than a problem solver.</p>

<p>The difference between adjusting the focal length on my 18-55mm zoom kit lens from 45mm to 50mm is barely noticeable from the change in how the subject covers the frame.</p>

<p>I was a photorealist portrait artist back in the '70's. I know the difference between perspective distortion and what I see through a camera's viewfinder which has nothing to do with perspective. I stand by what I say and I know what I'm talking about.</p>

<p>Kinda lost for words here...<br>

The OP's problem is not with lens distortion, but with "ideal distance" whatever this may be. You brought up the lens distortion as a factor in portrait - I believe it's a contrived problem. Moot. And if you feel it is a factor in your portraiture, then a solution would be pretty simple wouldn't it? Don't use lenses that suffer from distortion. No one forces you to shoot lame kits. Plop on one of Pentax's cute little Limiteds and you're set to go. Problem solved.<br>

I only participated in this because you started with the statement about being able to shoot up close without distortion with a full frame camera. I'm a curious man, therefore I asked for clarification. And I disagreed with you therefore I verbalized (or keyboardalized as it might be) my thoughts on importance of lens distortion in portraiture. Which I believe is none altogether. I am not suggesting that I have more experience than you - what I said is simply that I have a lot of it. Quite enough actually, to be able to decide what is important in photographic portraiture, and what's not. Important for me that is. Your mileage may vary, but making it personal is not the right way of leading any kind of discussion. Going Schopenhauer on me will not help to drive your point home, believe me.</p>

<p>Ken:<br>

Are you closer to an answer for the <em><strong>first</strong></em> question you asked?</p>

<p>WW</p>

<blockquote>

<p>Ken:<br />Are you closer to an answer for the <em><strong>first</strong></em> question you asked?</p>

</blockquote>

<p><br /> Actually, yes. :-) I've kept quiet in order to let people express their thoughts, which have been quite vigorous to say the least, but I think I kind of got the answer from the first few responses. Marcin's statement that perspective distortion, in the strict sense, is unavoidable as long as we put 3D reality into a 2D flat photo makes a lot of sense. Even a 3D view from the same spot may be perceived differently from one person to another, because our brains deceive us. <br /> So, I think I'll go back to what I've been doing, I guess. Try to shoot from a position where I can capture the best look of the subject in a well-blended surrounding. I shouldn't worry about a little distortion caused by proximity or distance, unless it's extreme.<br /> In a nutshell: there is no "ideal" distance even if you were to place priority on avoiding "perspective distortion", which shouldn't be a priority to begin with.</p>

<p>I've made related comments on a recent thread, and was surprised at the resistance to some well-known ideas about perspective in photography. Rather than enter the quagmire myself, here is an excerpt from <em>Optics in Photography</em>, by Rudolf Kingslake. Few industry experts would quarrel with Kingslake, but perhaps some here could teach him a lesson.</p>

<blockquote>

<p>By far the most important rule for correct perspective in photography is that the final print must be viewed from approximately its correct center of perspective, so that the angles subtended at the eye by the various images in the picture will be the same as the subtense angles of the original objects at the camera lens....<br>

...<br>

[note: Kingslake then describes the situation for prints, including an example of enlargement from a 35mm negative]<br>

...<br>

The gain in realism obtained by enlarging small negatives in this way is quite marked and often astonishing.</p>

</blockquote>

<p>This is from the first chapter of his book, and seems to be readable from Google books.</p>

<p>In other photographic literature, "perspective distortion" is generally considered to be a result of viewing the image from the "wrong" distance. I'm not planning to defend this myself (although I might), but my prior offer to supply literature references still stands.</p>

<p> </p>

<p>I'll say it for the last time there's no such thing as perspective distortion between a lens and a head shot that fits within a foot of space and just feet away. You have to get way farther back than that to start seeing perspective induced distortion.</p>

<p>NOBODY TOOK DRAFTING CLASS IN HIGH SCHOOL IN THIS THREAD?!</p>

<p>Heck I learned this from Walter Foster drawing books I picked up at hobby store when I was ten years old. They even provided the draftsman formula for plotting perspective using top view geometrics from a blueprint. It's called "projection" and it only works on houses way bigger than your head.</p>

<p>Would the visual concept of a centralized hub (your head/portrait) in relation to the fanning out of the spokes of an enormous wheel help you understand you can't get perspective distortion with a lens at that close range? Does that help?</p>

<p>Tim, I think you must be talking about something different than the rest of us. Perhaps it too is called perspective distortion, but it's not the same perspective distortion that the rest of us are talking about.</p>

<p>With the kind we're talking about, the sentence you've written doesn't even make sense. The distance from camera to subject, or viewer to print, or any other distance is not enough to say whether there will be distortion -- there's only distortion when one form (angle of view at capture time) differs from another (angle of view at viewing time).</p>

<p>If you're a foot away from your subject with a 50mm lens on a full-frame camera, you've got roughly a 40 degree horizontal angle of view. If the print (or other rendering, e.g. JPEG shown on an electronic device) subtends the same angle for the viewer, no distortion. If the print subtends a larger or smaller angle there is distortion, with the amount of distortion varying with the difference from the original.</p>

<p>No lens-induced distortions (barrel, etc.) have anything to do with the perspective distortion that the rest of us are talking about, and you can certainly get distortion at *any* camera-to-subject distance, if the print size and/or viewing distance doesn't result in the right angle of view.</p>

<p>I'm having trouble visualizing the sort of distortion you're talking about, but it's clearly not the same sort of distortion I'm talking about, and I don't think it's the sort of distortion the original poster asked about.</p>

<p>I'm really confused now. There seems to be three schools of thought: (1) there is correct perspective, (2) perspective distortion cannot be cured and is not that important, and (3) no such thing as perspective distortion at least in the context of portraiture.<br>

Of these three, I'm really puzzled over Tim Lookingbill's argument along the lines of (3). That really goes against what little knowledge I have in photography.<br>

@ Tim, how do you explain something like <a href="http://www.mcpactions.com/blog/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/">http://www.mcpactions.com/blog/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/</a>? You can certainly see perspective distortion specifically caused by proximity to the subject.</p>

<p>Ken, if you took that first 24mm shot and printed it big and held it close to your face so that it subtended 53x74 degrees, it would look undistorted.</p>

<p>@ Bill C,<br /> I think I'm finally beginning to understand the logic in your quote. Ok, it makes logical sense. If the camera had certain angle of view, then having the viewer to stand in front of the photo at such point that the viewer will have the same angle of view as the camera will result in correct perspective --- for the viewer!<br /> That is not exactly what I'm asking in the OP. In the sample photos below, the girl definitely has an unnaturally large nose in the 24mm shot. If the viewer were to stand so close to the photo that the viewer has a 24mm focal length's angle of view (which is 85'), then, yes, the viewer is seeing what the camera was seeing. If viewer was where the camera was, the nose would look relatively larger than the rest of the face. <br /> <a href="http://www.mcpactions.com/blog/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/" rel="nofollow" target="_blank">http://www.mcpactions.com/blog/2010/07/21/the-ideal-focal-length-for-portraiture-a-photographers-experiment/</a><br /> But, that doesn't change the fact that the 24mm sample looks unnatural. When I stand in front of my subject such that I have an 85' angle of view, I don't think I'll perceive my subject's nose to be that enlarged. Because our brain conveniently deceives us. (So, to be exact, maybe it's not the 24mm sample that has distortion. Maybe it's the human brain that creates distortion in order to conform our view with our pre-conceived world. Just like our brain deceives us about white balance and makes us perceive skin tone as "natural" under tungsten, sunlight or fluorescent, although the camera will record them "correctly" in weird colors, such as yellow, bluish or greenish.)<br /> <br /> Going back to my OP and also the sample shots in the above link: I don't know the girl in the photo personally, so I can't say which one looks like the most natural depiction of her. Somebody that knows her will surely say "Her nose is not that big", looking at 24mm or 35mm shot. (They are not going to say "Ok, that's the perspective I'd get if I stood so close to her to get 85' view of angle. Nothing weird.") So which one of the normal to telephoto is most natural? That was the gist of my OP. When someone who knows her personally sees these photos, which would be picked as the most natural or correct depiction?<br /> But since I've decided that it's not that important (i.e., 50mm - 165mm shots all look good), I'd like to put this thread to a rest.</p>

<blockquote>

<p>Ken, if you took that first 24mm shot and printed it big and held it close to your face so that it subtended 53x74 degrees, it would look undistorted.<br /> <br /><br /></p>

</blockquote>

<p>Mark, I just did that, and the girl's nose in the 24mm still looks big from up close. :-) I think we are using the term "distortion" differently. Even the girl's big nose is not distortion in the sense the it's how it should look like from that angle of view. But as a lay person, I've been using "distorted" as synonym for "unnatural" or "weird".<br>

Caveat: I didn't print the photo really big. just letter size. If I had printed so big that it filled an entire wall of my room, maybe the nose won't look that big. I don't know. But that doesn't sound like reasonable approach. Taking a photo and expecting the viewer to print it so large and stand at a position where "unnaturalness" dissipates.</p>

<blockquote>

<p>Would the visual concept of a centralized hub (your head/portrait) in relation to the fanning out of the spokes of an enormous wheel help you understand you can't get perspective distortion with a lens at that close range? Does that help?</p>

</blockquote>

<p>Tim, I know this is lame but will wikipedia help? This is what I (and I think everyone here except you) mean when using term "perspective distortion": http://en.wikipedia.org/wiki/Perspective_distortion_%28photography%29<br>

When you move, relations between sizes of projected objects changes. Noses get bigger, ears get smaller or the other way around. It's got nothing to do with lens, it's pure geometry of non parallel projection. The closer you get, the more dramatic the change with every move.<br>

And I'll say that again, it's not such a big deal in portraiture as long as you keep your mm's within reasonable limits.</p>

<p>Well, I've been following this thread and thought I'd add my 2¢.</p>

<p>Tim, I think Mark is right about your definition of perspective distortion differing from what is normally understood by photographers. Everything I've read about it seems to point towards it referring to the relative difference in size of objects in the image, such as the eyes, nose, etc. in a portrait. As the difference in distance to the camera relative to each other increases, the more the closer objects grow in relation the parts further back, causing what most photographers refer to as perspective distortion. For example, when I first got my wide angle I took head shots from close-up and the resulting images had very exaggerated features.</p>

<p>Mark, I think the notion that the viewing distance of an image influences the apparent perspective distortion is an idea that sounds good on paper but doesn't seem to work that way in practice, at least not until you start getting into extremes. My impression of whether an image was taken from too close up doesn't change noticeably with my viewing distance. I think Ken hit the nail on the head, our brains are doing a lot of processing on everything our eyes "see". That's why I can look at the little images at the bottom of the page and they appear the same as if they filled the screen, as far as any distortion goes.</p>

<blockquote>

<p>But that doesn't sound like reasonable approach. Taking a photo and expecting the viewer to print it so large and stand at a position where "unnaturalness" dissipates.</p>

</blockquote>

<p>Yep, it doesn't sound reasonable. Yet how many of us avoid the front row at a movie theatre? If you find that the movie is more enjoyable when you sit near the middle of the theatre, you may have recognized perspective distortion, and found a way to minimize it.</p>

<p>Anyone who has gotten a big-screen television for the immersive experience is essentially doing the same thing. When the viewing position is right, to repeat Kingslake's words, "The gain in realism ... is quite marked and often astonishing."</p>

<p>To be clear, all this is a separate issue from Ken's original question about the ideal distance.</p>

<p>ps: agree 100% with Mark Sirota.</p>