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

<p>What is the ideal distance from the subject for candid/environmental portrait purely on the "perspective" ground?<br>

In other words, assuming you don't need to interact with the subject (because it is candid, environmental), there is no physical barrier that prohibits you from backing up or closing in, the lens can afford you with shallow DOF at any distance, etc.<br>

If you are too close from the subject, you get distortion - nose too big for example. But if you are too far from the subject, there is also distortion - ears too big, I guess. People say that the ideal distance for candid/environmental portrait is the distance you'd normally keep with the person without a camera. I don't like that approach. I'd normally consider even 5 feet a comfortable distance from the subject, but at 5 feet I think you already see distortion created by the close distance. (The human eyes have a way of deceiving us, as we well know, and we - or at least I - don't perceive the distortion at 5 feet. But a camera will and record it in the photo.)<br>

Since the perspective is purely a function of the distance from the subject, has there been a discussion on what distance gives the most "natural" balance or perspective on the human subject's face and other body parts? </p>

<p>Hi Ken - I don't think there is an ideal distance purely on a "perspective" ground. I think the ideal distance is rather a function of artistic intent, focal length of lens in use, what you want to capture in the portrait, and lastly the issue of comfortable distance. I've found that I generally prefer 7-12 ft with a 90mm focal length lens on a full frame or film camera, and often with a 50mm lens on a cropped sensor, I'll come in as close as 6-8 ft. Here's a shot I took in July at a family gathering which was taken from around 5-6 ft with a 50mm lens, and I don't think introduces unnatural distortion, yet really distills the essence of the subject.</p><div></div>

<p>Uh. This whole "you will get distortion" thing is kinda misleading. You always get distortion, unless you shoot some kind of orthogonal camera (I don't know if such a thing exist). With an orthogonal camera you'd get undistorted portrait. It would be very, very weird. Not distorted, but certainly disturbing.<br>

That said, I don't think there is such a thing as "ideal distance". Depends much on what you want to achieve.</p>

<p>I'll agree with the comments above, which focal length you choose depends on what your trying to achieve. Personal preferences vary widely as to what's considered natural looking. I personally find the portraits I've taken with my 100mm macro lens very nice. But I gotten positive comments on images taken 11mm all the way to images taken at 200mm. I personally think that pictures taken at a distance between 15 and 20 feet yield the most natural looking images. What focal length you choose depends on how much of the scene you want to capture.</p>

<p>The most natural perspective will be where your angle of view matches the angle of view of the viewer. That is, if you'll be making a 24x36cm (9.4x14.1") print and expect it to be viewed from 2m (79") away, then you'd want a 200mm lens in front of your 24x36mm camera.</p>

<p>While camera-to-subject distance is a major factor in perspective, you'll note it's not a factor in the formula above. Thinking about it in terms of angle of view, matching the angle at capture time to the angle at viewing time, means that the viewer will have the same perspective you have, regardless of where you're standing relative to your subject.</p>

<p>If you constrain other factors (print size, viewing distance, etc.) then you may be able to work out the matching camera-to-subject distance, but that feels like the wrong way to go about it. Shoot at the distance that gives you the perspective you want, then print and hang accordingly.</p>

<p>Along with others have said, it helps to have fairly accurate viewfinder with regard to cropping as you adjust both your distance and focal length if you happen to be using a zoom rather than fixed prime lens. Let the composition dictate distance.</p>

<p>Below is an example of some of the subtle distortions of a portrait I shot at 80mm. My viewfinder isn't 100% so I had to adjust distance and chimp on my LCD after the shot. I couldn't really detect until viewing on my computer the slight thinning of the face seen on the left portrait by stepping back compared to what seems a wider overall girth to both the nose and the face by getting closer.</p>

<p>It's almost a parabolic mirror effect where you start noticing a central area of the lens exhibiting this subtle bowl like curved distortion moving in and out. You'll have to locate where this bowl like area resides on your own lens by noticing your subjects face thinning and widening by moving closer and/or adjusting focal length.</p>

<p>It took me a while to pick up on this subtle distortion even shooting at 80mm on an APS-C DSLR Pentax K100D.</p><div></div>

<p>Now to rule out distortion differences produced by having the hat and wider crop in the left shot affecting perception, I cropped the one on the left to be the same dimensionally in order to show there really is a distortion produced by the lens. </p><div></div>

<blockquote>

<p>What is the ideal distance from the subject for candid/environmental portrait purely on the "perspective" ground?</p>

</blockquote>

<p>For mostly all Environmental Portraiture, I like to work between about 8 Feet (2.5Meters) and 20 Feet (6 Meters).<br /> Within that range of distances, the two primary determinates as to the exact distance I would work would be:</p>

<ul>

<li><a href="/photo/12352872&size=lg">The intimacy I wanted with the Subject - this example at a closer SD and using a Wide Angle Lens.</a></li>

<li><a href="/photo/9204992&size=lg">The space (of the Subject) that I did not want to invade - this example at a longer SD and using a Short Telephoto Lens </a></li>

</ul>

<p>***<br>

Candid Portraiture is a different Genre to Environmental Portraiture.<br>

For Candid Portraiture I tend to want to work quite close: more in the 6 Feet to 12 Feet range this is mainly to achieve the intimacy that I usually require.<br>

<a href="/photo/9197192&size=lg">In this example using a Standard Lens</a><br>

<a href="/photo/10738709&size=lg">And in this example using a Short Telephoto Lens.</a><br>

<a href="/photo/12352873&size=lg">And this using a Wide Angle Lens: and in this image, I don't mind the "distortion", at all.</a></p>

<p>***<br>

<br>

Perspective is not purely a function of the distance from the Subject: Perspective is a function of the Camera Viewpoint relative to the Subject - that is Distance and also Elevation. Though Elevation is often not a consideration in many Portrait shots: for the question that you are asking, I think that this point is quite important to recognize.</p>

<p>WW</p>

<p>+1. Personally, I prefer anything over 70mm for portraits. Also note that the focal length will interact with the camera-to-subject distance as well as the subject-to-background distance.</p>

<p>Ain't geometry fun?</p>

<p>There's an assumption here that portraits should meet some "lowest common denominator" when it comes to appearance. That ignores a lot of things, including history - most portraits from the "classic" period of photography were made with "normal" lenses. It ignores that photographers might have something they are looking for. "Natural" is an imposition on a photographer's creativity. For commercial work, like what William does (and I sometimes do), there are expectations. However, for anything else, it seems that these threads aim at looking at stunted photography, stuck in some arbitrary definitions of "natural." </p>

<p>I shoot most portraits around "normal" focal length, like so many of the people I grew up respecting and learning from - Arbus, Avedon, Bravo, Rogovich. The familiarity with the focal length allows me to make portraits look the way I want them to look. When I do commercial work, I use a longer lens if I have the room, but that's because there are expectations for a look that meets some specific requirements. However, I have done commercial portraits, including for magazine covers, at very short focal lengths when situations dictate. If you're not doing commercial work, following this will only help to produce generic portraits.</p>

<p>Another short f.l. portrait...</p>

<p><center><img src="http://spirer.com/images/reggaesinger1.jpg" alt="" width="467" height="700" /></center></p>

<p>Another one, about 52mm (full frame equivalent)...what would distance and f.l. have done to this if I followed the guidelines here?</p>

<p><center><img src="http://spirer.com/images/skateportrait1.jpg" alt="" width="467" height="700" /></center></p>

<p> </p>

<p>I think any photographer is doing themself a disservice if they prescribe the shot by the mathematics of it.</p>

<p>Equally, I think that it is important to know the mathematics, but that knowledge should be used as part of the method to achieve the Photographer’s vision. (For example using your question) the Shooting Distance and the Focal Length of the Lens is chosen on what you have learned, so that there is very little fiddling around or guessing at the time when the shot has to be made.</p>

<p>There are many threads about Portraiture which seem to be predicated on, or assume ‘rules’ concerning lenses and distances and other technical procedures apply as a ‘one solution’. I think that a skilled Photographer, will seek to know the mathematics or the ‘rules’ and apply them as it suits each individual situation: but not be a slave to them.</p>

<p>It’s a all good to ask how other Photographers do things, but their goals, might not be your goals, so I think it is a good idea, when you ask; “what distance?” to also ask; “why do you choose that distance?”</p>

<p>WW<br>

</p>

<p>Further reading:</p>

<p>http://www.journalofvision.org/content/12/5/8.full</p>

<blockquote>

<p>There's an assumption here that portraits should meet some "lowest common denominator" when it comes to appearance.</p>

</blockquote>

<p>Jeff, is avoidance of facial distortion as the OP is seeking in a finished image considered the lowest common denominator? Is the OP in your opinion not being creative by choosing to want "a look" that is void of facial distortion? The OP just wants to record the person exactly as the eyes see the subject. </p>

<p>Don't get me wrong, I understand the use of lens/distance to create a unique and compelling portrait, but the OP isn't asking for that.</p>

<p>Came across an interesting article on page 68 of my September 2013 copy of "Popular Photography" of an interview of George Lange portrait techniques where he was asked...</p>

 

<blockquote>

<p><em>Do you have a favorite lens?</em>...Where he replies..."mostly 24-70mm...nothing less than 50mm...I don't like being much farther than arm's length"</p>

 

</blockquote>

<p><strong>OMG he's advocating a selfie!</strong></p>

<p>Of course he's using a full frame Canon EOS 1Dx which makes getting closer doable with little distortion. His portraits printed in the article look quite flat field with regards to a lack of facial feature distortions. </p>

<p>Which brings me to what I should've added to my little demo in that I was using a Sigma 28-80mm f/4 FA lens designed for full frame 35mm film cameras. That I'm guessing is going to change where the bowl of the lens covers the subjects face in relation to where it falls on an APS-C crop sensor. Someone correct me on this if I'm wrong. I don't have a full frame sensor camera to test this.</p>

<p>Something else that didn't get mentioned is the use of lens profiles and whether it can correct for lens/distance induced facial distortion.</p>

<p> </p>

<blockquote>

<p>Of course he's using a full frame Canon EOS 1Dx which makes getting closer doable with little distortion.</p>

</blockquote>

<p>This piqued my interest. Would you care to explain how bigger sensor format allows you to get closer to your subject with little distortion?</p>

<blockquote>

<p>The OP just wants to record the person exactly as the eyes see the subject.</p>

</blockquote>

<p> <br>

1. That's impossible.<br>

2. That's not stated. It sounds like a scientific exercise. The "eye" doesn't see the subject. The combination of the eye and the brain see the subject. And, as numerous scientific studies have shown, that creates a certain amount of fuzziness about what something looks like.</p>

<p>Jeff, did you just go into "Spock" mode on that last answer?</p>

<p>I would think given a decently accurate viewfinder the photographer could easily see and compare against the actual subject he's looking at to check for any head shape, size and facial distortions.</p>

 

<blockquote>

<p>This piqued my interest. Would you care to explain how bigger sensor format allows you to get closer to your subject with little distortion?</p>

</blockquote>

<p>Modern digital lenses are designed/engineered both within the build of the glass as well as any electronics communicated to the camera's onboard processor to focus the best results according to the size of the sensor and communicate this to aid in either jpeg rendering or Raw format for the camera manufacturer's Raw converter.</p>

<p>This is the reason I asked about lens profiles for third party converters. There's got to be some distortion compensation going on according to what I've seen with my own Pentax DSLR and my 18-55mm digital kit lens from the changes I've seen to ACR previews applying a lens profile. There's no lens profile for barrel distortion correction for that film legacy Sigma I used.</p>

<blockquote>

<p>Modern digital lenses are designed/engineered both within the build of the glass as well as any electronics communicated to the camera's onboard processor to focus the best results according to the size of the sensor and communicate this to aid in either jpeg rendering or Raw format for the camera manufacturer's Raw converter.</p>

</blockquote>

<p>Um. I think you misunderstand something here. The lens specific distortion correction that comes with your camera's jpeg engine or RAW developer or Photoshop has nothing to do with perspective distortion. You can't really do anything about perspective distortion without changing the subject distance, which is, well, physically impossible after the shot was taken, right?</p>

<blockquote>

<p>I would think given a decently accurate viewfinder the photographer could easily see and compare against the actual subject he's looking at to check for any head shape, size and facial distortions.</p>

</blockquote>

<p>There's a huge difference between a 3D object and a 2D photograph. No matter where you stand or what lens you use. So it comes down to how the brain interprets things, and that is not the same for any two people. And the subject will never look like what the eye sees in a photograph.</p>

<p>Slap on a lens and have a look. If it looks good then it's "ideal".</p>

<p>To me "environmental portrait" covers a lot of ground. It could mean a close up of a face with sweat dripping off it, using a 200mm lens, or a wide angle shot of a farmer standing in a wheat field. I've done a lot of candid/environmental portraits (see my folders) and for me its a matter of "what works the best" in this situation. I've never worried about distortion or "correctness" etc. As Jeff points out, a photograph is a 2 dimensional image, and not reality. Its also art, and needs to be understood from its emotional impact and not a formula of physics. That's my 2 cents.</p>

<blockquote>

<p>The lens specific distortion correction that comes with your camera's jpeg engine or RAW developer or Photoshop has nothing to do with perspective distortion.</p>

</blockquote>

<p>I had to study perspective drawing in my drafting class at the design school I attended. I even had to learn a formula for calculating the plotting of rate of converging lines landing near horizon for architectural illustration so I have a pretty good idea what the difference is between perspective induced distortion and what I see from parabolic distortion off a lens which is what I'm talking about and what I demo'ed with images above.</p>

<p>I really don't know the point you're getting at in making a distinction with perspective distortion which is not what I'm talking about and demonstrated above. Lens profiles and what the manufacturer builds into a lens designed for digital sensors has to be taken into account. A film legacy lens doesn't know anything about a cropped APS-C sensor and neither do I, but I don't rule the possibility out that it may have an affect.<br /> What I demonstrated above with the thinning of the face on the left and enlarging of the face and nose on the right <strong>has nothing to do with perspective</strong> from the POV of creating a photographic image that appears as the eyes see the subject.</p>

<p>It has everything to do with where the portion of the image (face) falls in the area of the arc of the bowl shape glass groupings within the lens as it directs the photons to draw the image onto the sensor that creates the distortion described above. Move the subject closer or farther away changes the degree of arc on how it is drawn on the sensor.</p>

<p>Quite a few photographers say to use a long lens and step farther back to avoid wide angle distortion but they never quite indicate exactly what level of distortion is tolerable or if it can be completely eliminated. I showed above what the tolerable level looks like even using an 80mm film legacy lens.</p>

<p>Because of this lack of specificity often communicated in discussions of this sort, I included the possibility that a film legacy lens, which a cropped sensor digital camera knows nothing about, may have an influence when drawing (directing the photons) of the subject onto the sensor.</p>

<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>There's a huge difference between a 3D object and a 2D photograph. No matter where you stand or what lens you use. So it comes down to how the brain interprets things, and that is not the same for any two people. And the subject will never look like what the eye sees in a photograph.</p>

</blockquote>

<p>Jeff, ever use a graphics camera? It's a giant box camera on rails with a (I'm assuming) flat field lens. And what I saw on the viewing platen where I placed the film was exactly what I got after developing. For instance if I shot a 1x1 foot square shape on this camera, I could lay the photographed version right on top and it would line up perfectly.</p>

<p>I used to turn the copy board facing up acting as a table and photograph 3D objects, same results except I couldn't line up a finished piece of film to know exactly for sure but to my eyes comparing both the image and the original subject I didn't see any distortion as I demonstrated above.</p>

<p>Here's the graphics (copy) camera I used...</p>

<p>http://farm6.staticflickr.com/5185/5759590672_152fd22ae4_o.jpg</p>

<p>From this page...</p>

<p>

<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>