Is there a definitive focal length for tight head shots...

[David Lyga]

Very interesting how, in the 'real world', this focal length determinant is not mired in absolutes. Different lenses create different effects. That attitude, more than anything else, is refreshing, as it segues into the creative element. Some things are worse today but this is an example of thinking that is truly better. Benjiboy's suggestion that one use a 70-150 (assuming for the ability to determine 'on the fly') is potent. Also, agnosticnikon's assertion that even a 500mm (!) portrait was not only acceptable but WANTED by the client (over the others) was interesting.

The feedback here is great and telling, not only from the standpoint of technique, but also from the standpoint of human nature. More opinions are welcome (if not already mentioned). - David Lyga

 

[E. von Hoegh]

"even a 500mm (!) portrait was not only acceptable but ..."

Beyond about 180 - 200mm on 35, there is very little change of perspective ("compression").

If you double the standard focal length, there is a very noticeable difference. Doubling again, to 180 - 200, there is a less-but-still noticeable difference, and so on. One advantage of longer lenses is that the background fades into an abstract mush.

The point is to use what you like, what works for you and your subject.

 

[dynachrome]

My first lens was a 57/1.4. I used a Soligor 2X teleconverter which gave me, effectively, a 114/2.8. Getting close was not a problem but sharpness was not too good until at least f/8 (effective f/16). My next lens was a 28/2.8 and was not used for portraits. The next lens was a 135/2.8. It was a decent lens but only got down to about 5 feet. It was later traded in for a 135/3.2 which focused down to 3 feet. I found that adequate for tight portraits. Years later I got a Vivitar 135/2.8 Close Focusing lens. That lens goes down to 1:2 which is much too close for a portrait but I find it useful for anything from the near macro range to infinity and very good for portraits. I have this lens now in many mounts.

Over time I accumulated many other lenses suitable for portraits. My favorite may be the 100/2.5 Minolta MC Rokkor. I have taken nice portraits with the 85/1.8 Konica Hexanon, 85/1.8 Canon New FD, 85/1.8 Canon FL, 85/2 AI Nikkor, 100/2.8 Canon FD SSC/New FD, 105/2.8 Super Takumar, 100/2.8 Zuiko etc. They are all good and how far you want tp be from your subject when shooting really is a subjective thing. Medium telephoto macro lenses like the 90/2.5 Tamron or 90/2.5 Vivitar Series 1 are also good for portraits but can sometimes be a little too sharp. I find both older and newer 105/2.5 Nikkors good for portraits. There are times when I will try a portrait with a 200/2.8 Canon New FD or a 200/3 Vivitar Series 1. A lens in the 50mm range can work when shooting a small group of people or 2-3 people standing close to each other but I do not like its look for a tightly cropped photo of one person.

 

[Poisson Du Jour]

Nothing stopping you from using a 20mm or even a 17mm if people are whacky enough to want a portrait shot on 500mm....
What did they use 120 years ago? For all intents and purposes, a pinhole will do fine too.

 

[RalphLambrecht]

agreed, my rule of thumb is double the normal focal length or longer.

 

[markbarendt]

I disagree that wide-angle distortion is only noticeable if prints are viewed from too far away. Our minds are very aware that the print is flat so any wide-angle distortions are very noticeable regardless of viewing distance. Even if viewed very closely wide-angle distortion is quite plain to see because, again, our minds know the image is planar. Whether or not any type of distortion is acceptable or preferable is up to the viewer.

All I can say is try it. It's a 3D geometry thing.

If you are viewing the subject matter at all the same angles the camera did, things look pretty normal. That angular relationship is determined by print to viewer distance.

I have 24x36 inch print of a cannon with St. Charles cathedral in the back ground shot very wide. All kinds of lines and angles.

At say 4' the "distortion" is apparent, at 18" to 24" it's gone.

Took this shoot to the local camera club and showed them the change in effect and everybody saw the difference when they got close.

I will say that small prints don't do this effect justice IMO.

 

[removed account4]

hi david

i don't have much to add to your thread, except
use what ever focal length feels right to you.
i shot for a long while for a newspaper head shots
environmental portraits, pretty much everything you can think of ..
and i just used what i had handy. sometimes it was a 28, sometimes it was a 50
sometimes it was a 100 ( even a 135 )... whatever worked ... there was no steadfast rule ..
just general "tips" for positioning good v. bad side general "stuff" i learned
as an apprentice to a portrait photographer a few years earlier.
rules of composition are different now than they were 40 or 50 years ago
for headshots at least, you can cut heads and necks and chins off and no one will say much ... so do what
you like, and if you are doing it for a client or $$ or whatever ... make sure
they like it too

good luck !
john

 

[Old-N-Feeble]

All I can say is try it. It's a 3D geometry thing.

If you are viewing the subject matter at all the same angles the camera did, things look pretty normal. That angular relationship is determined by print to viewer distance.

I have 24x36 inch print of a cannon with St. Charles cathedral in the back ground shot very wide. All kinds of lines and angles.

At say 4' the "distortion" is apparent, at 18" to 24" it's gone.

Took this shoot to the local camera club and showed them the change in effect and everybody saw the difference when they got close.

I will say that small prints don't do this effect justice IMO.

That's only true to a degree and it's far from infallible. Too, as you already pointed out, this really only works with large prints. Who's going to view an 8x10 inch print from 5 inches away unless it's a photo of Heidi Klum in a bikini?

 

[Old-N-Feeble]

A few folks have mentioned environmental portraits. This type, and other images in which people are placed into an environment that must be shown, partially dictate lens focal length. Typically, wider lenses are chosen for this type of portraiture... sometimes very wide. As others pointed out, there is no right or wrong regarding focal length.

But... the OP did offer quantifiers to his/her question and that is "minimizing distortion of facial features" at a particular subject cropping factor. As other pointed out, shooting distance controls compression distortion, and how much subject you want in the frame dictates focal length. Shooting angle and lighting add to the final result. The environment, intended effect/message, and facial expression all make a difference on how much distortion you might want. Also, as others pointed out, compression distortion is decreasingly apparent as focal length increases. The bottom line is I follow my previously posted rules-of-thumb loosely but they work very well "for me" given the OP's specific requirements.

 

[darinwc]

Wide angle shots only look distorted when viewed from to far away or put another way printed too small for the intended viewing distance.

OK I noticed a reference to this and I had to look back a few pages to find out for sure if someone had actually said this.
I cannot disagree more. The print is flat. Changing the viewing distance of the print cannot change the proportions on the print.

My nose would still be 80% the width of my face, no matter how near or far the print was. Nor what angle it was viewed at. Trust me, I've tried
Or perhaps it may be possible that my nose would actually be appear larger if I put my real nose right up to the print.
But this would be the opposite effect of what you were trying to describe.

 

[markbarendt]

OK I noticed a reference to this and I had to look back a few pages to find out for sure if someone had actually said this.
I cannot disagree more. The print is flat. Changing the viewing distance of the print cannot change the proportions on the print.

My nose would still be 80% the width of my face, no matter how near or far the print was. Nor what angle it was viewed at. Trust me, I've tried
Or perhaps it may be possible that my nose would actually be appear larger if I put my real nose right up to the print.
But this would be the opposite effect of what you were trying to describe.

All I can suggest is trying it, as I described earlier.

 

[MikeTime]

As far as I know, if you view an image at the same distance it was taken, the perspective will appear similar. I could be wrong, and I also think the brain perceives perspective in a print different from real life.

If you were to photograph someone with a 35mm lens at 1 meter, you would have to view the resulting image at the same distance?

 

[Thomas Bertilsson]

As far as I know, if you view an image at the same distance it was taken, the perspective will appear similar. I could be wrong, and I also think the brain perceives perspective in a print different from real life.

If you were to photograph someone with a 35mm lens at 1 meter, you would have to view the resulting image at the same distance?

Incorrect. A 35mm lens on a 35mm camera has a different 'angle of view' compared to the human eye. With a 50mm lens you are roughly correct.

As others have pointed out, it is a matter of taste what is 'correct' and what isn't. Personally, I favor slightly wide to normal lenses for portraiture, mainly because I get closer to the subject, which makes it a more intimate experience. I'm convinced this changes how the images result, because of the intimacy and the reactions it provokes. I also like the much different perspective and view that results from shorter focal length objectives; again, a matter of taste and a preference for a certain way of working.
But in a situation when I wouldn't want to be close to the subject, for whatever reason, I would compromise and use a longer lens obviously. It could be that the sitter finds it difficult to work with me so close, for example.

 

[BetterSense]

If you were to photograph someone with a 35mm lens at 1 meter, you would have to view the resulting image at the same distance?

I think the argument people are making is that, the perspective would look natural IF: You printed the image lifesize (meaning the actual width of the head ON the image was the same as the width of the person's head), AND you viewed it from the same distance the camera was away from the subject when the image was taken.

Since it's unlikely that you would view such a large print from such a close distance, perspective would be perceived as unnatural.

 

[David Lyga]

Yes, Thomas Bertilsson, your choice of moving closer and using short lenses is YOUR decision, having to do with YOUR sense of the aesthetic, in the immediate situation. This is refreshing to hear and attenuates the definitiveness of the 'standard' rule regarding proper focal length.

I want to say that I never could understand that the 'normal' focal length, however, should be the diagonal of the full 35mm frame: i.e., 43mm. I have found the Helios 2/58 to match more exactly what the eye really sees. (Using that lens for portraiture is, however, a matter not related to that assessment of 'normal'). - David Lyga

 

[markbarendt]

Incorrect. A 35mm lens on a 35mm camera has a different 'angle of view' compared to the human eye. With a 50mm lens you are roughly correct.

Thomas, what I'm suggesting is that when "the camera's angle of view" matches "the viewers angle of view" (defined by print viewing distance and print size), the the print will look normal.

There are definitely limits to this, a 180 fish eye for example, but if viewing distance remains constant and print size increases a wider lenses can be used.

The flattening effect of the longer lenses that we are all so familiar with is because of the mismatch in camera angle of view vs print viewing angle.

The same math formula applies to both applications.

 

[Thomas Bertilsson]

Thomas, what I'm suggesting is that when "the camera's angle of view" matches "the viewers angle of view" (defined by print viewing distance and print size), the the print will look normal.

There are definitely limits to this, a 180 fish eye for example, but if viewing distance remains constant and print size increases a wider lenses can be used.

The flattening effect of the longer lenses that we are all so familiar with is because of the mismatch in camera angle of view vs print viewing angle.

The same math formula applies to both applications.

Sorry, Mark, I don't see it.
To me wide angle looks wide angle, normal looks normal, and telephoto looks telephoto.
What do you mean by viewer's angle of view? We can only look at a print one single way - with our eyes, and as far as I understand, they do not have the ability to adjust angle of view, like switching lenses on a camera. Please explain in very specific terms what you mean; if you have some sort of drawings or pictures to show what you mean that would help.

 

[markbarendt]

The viewers angle of view the angle defined by print width (or height) vs viewing distance (print to eye distance).

 

[Thomas Bertilsson]

The viewers angle of view the angle defined by print width (or height) vs viewing distance (print to eye distance).

I think your theory would work if it were a 3-dimensional object. But the print is 2-dimensional.

If you take a picture of the same scene, one with a 35mm lens, and the other with an 85mm lens, how exactly would you position the prints so that the perspective would be the same, with relation to the viewer?

 

[markbarendt]

All I can say Thomas is try it.

 

[Thomas Bertilsson]

Oh no, brother, I don't have time. It's your claim; if you want us to believe you - convince me.

http://www.mcpactions.com/blog/2010...h-for-portraiture-a-photographers-experiment/
Dead Link Removed

I don't even know what it is you want me to do exactly. How do I position myself with relation to the examples, in order for them to look the same.

 

[cliveh]

Thomas, Mark Barendt is correct. However, what he describes is ignored by most people viewing any photograph, painting or drawing. Perhaps viewing lines should be marked on gallery floors, but people would ignore them.

 

[Thomas Bertilsson]

Thomas, Mark Barendt is correct. However, what he describes is ignored by most people viewing any photograph, painting or drawing. Perhaps viewing lines should be marked on gallery floors, but people would ignore them.

I'm not saying he isn't correct, but I don't understand it. Can you explain?

 

[MattKing]

Thomas:

Mark is right.

Using the examples linked to as a reference, you no doubt understand that when the photographer took the photos, he/she had to stand at different distances from the subject in order to fill the frame with the head. As an example, he/she may have needed to stand 1 foot away when using the 24mm lens, 1.4 feet away with a 35mm lens, 2 feet away with a 50mm lens, etc. (example distances may be incorrect .

If the photographer then printed each example to lifesize, and put them side by side on a board, for you to look at, the perspective would look natural to you if you viewed each print from the same distance as the original shot was taken from.

You would need to view each one with one eye closed, and the relative size of the subject would decrease as you stepped farther and farther away, but the relationship of the size of the different parts of the subject's face to each other would appear natural in each case.

Binocular vision and the brain's ability to interpret differences without prompting tend to make the perspective distortions less obvious than they might be, but this approach does work.

 

[Thomas Bertilsson]

OK, now that's a great explanation, Matt. Thanks. I understand now, and hopefully many others (or was it just me being dense? )

I'll let others argue about the utility of knowing this.