Photographing a Person: A test of focal lengths

Started Apr 10, 2013 | Discussions
RoelHendrickx
RoelHendrickx Forum Pro • Posts: 25,949
Re: Photographing a Person: A test of focal lengths

Sudo Nimh wrote:

Matz03 wrote:

thanks for posting these, I was most interested in seeing the 25mm and 45mm perspective and wow, even bigger difference that I thought!  The age old rule is that a 50mm equivalent (your 25mm image) lens is bad for head shots and it obviously shows.  Your 45mm picture finally doesn't show that distortion and going up from there doesn't change things too much.

I see the big difference between 17mm and 25mm. At 17mm and shorter, you feel close to the subject. I actually like that close feel for a lot of head shots, especially of young kids with their flawless complexions and especially when there's something interesting in the background.

Ditto here.

I really don't see that much difference between 25mm and 45mm.  The "jump" in view happens between 17mm and 25mm : if there had been more "background", it would have been even clearer, I feel.

-- hide signature --

Roel Hendrickx
lots of images: www.roelh.zenfolio.com
my E-3 user field report from Tunisian Sahara: http://www.biofos.com/ukpsg/roel.html

OP richj20 Veteran Member • Posts: 6,133
Re: Photographing a Person: A test of focal lengths

Sudo Nimh wrote:

I see the big difference between 17mm and 25mm. At 17mm and shorter, you feel close to the subject. I actually like that close feel for a lot of head shots, especially of young kids with their flawless complexions and especially when there's something interesting in the background.

Good point! There are so many factors that can come into play as we are setting up the composition.

- Richard

-- hide signature --

"Careful photographers run their own tests." - Fred Picker

 richj20's gear list:richj20's gear list
Panasonic Lumix DMC-GX8 Panasonic Leica 100-400mm F4.0-6.3 ASPH Sony Alpha a7R II
OP richj20 Veteran Member • Posts: 6,133
Re: Photographing a Person: Distance

gianstam wrote:

Your test confirms my opinion about portraits.

The photographer must be too close to his subject (for several reasons)

Yes, distance is a factor to consider!

- Richard

-- hide signature --

"Careful photographers run their own tests." - Fred Picker

 richj20's gear list:richj20's gear list
Panasonic Lumix DMC-GX8 Panasonic Leica 100-400mm F4.0-6.3 ASPH Sony Alpha a7R II
OP richj20 Veteran Member • Posts: 6,133
Re: Similar test under controlled conditions

Vlad S wrote:

There we go: How Lens Focal Length Shapes the Face & Controls Perspective

Thanks!

- Richard

-- hide signature --

"Careful photographers run their own tests." - Fred Picker

 richj20's gear list:richj20's gear list
Panasonic Lumix DMC-GX8 Panasonic Leica 100-400mm F4.0-6.3 ASPH Sony Alpha a7R II
OP richj20 Veteran Member • Posts: 6,133
Re: Photographing a Person: A test of focal lengths

.Sam. wrote:

Thanks for the effort.

Obviously the 12mm doesn't look too pleasing. The rest seem to "useful" in a pinch. However, in my view the 45mm (e-90mm) stands out, and looks the best.

That seems to be a popular focal length!

- Richard

-- hide signature --

"Careful photographers run their own tests." - Fred Picker

 richj20's gear list:richj20's gear list
Panasonic Lumix DMC-GX8 Panasonic Leica 100-400mm F4.0-6.3 ASPH Sony Alpha a7R II
OP richj20 Veteran Member • Posts: 6,133
Re: Interesting... - I did the same in a LANDSCAPE setting

RoelHendrickx wrote:

There is some distortion on the face in the wider focal lengths (slightly elongated face, but well controlled, because your angle of view is good : this would be more obvious if you shot from higher to low or from lower to high).

Yes, I wanted to avoid that by using a similar angle of view. Unfortunately, he changed his head position a bit. As I mentioned earlier, I might use a mannequin in another test!

Starting from 25mm, there is not really that much difference anymore between the shots.

Others have made the same comment!

While hiking the Libyan desert, I did a similar little test, when I was explaining to our guide Rhyssa (who became my test subject) the effect of focal length on perspective.

In that case, I was not aiming to show distortions on a body or face, but rather the relationship between a person and his surroundings.

Yes, that is quite another situation. Good examples you show.

- Richard

 richj20's gear list:richj20's gear list
Panasonic Lumix DMC-GX8 Panasonic Leica 100-400mm F4.0-6.3 ASPH Sony Alpha a7R II
OP richj20 Veteran Member • Posts: 6,133
Re: Photographing a Person: A test of focal lengths

RoelHendrickx wrote:

I really don't see that much difference between 25mm and 45mm.  The "jump" in view happens between 17mm and 25mm : if there had been more "background", it would have been even clearer, I feel.

There were numerous scenarios to choose from. I may do another test with a different background.

Thanks for your observations!

- Richard

-- hide signature --

"Careful photographers run their own tests." - Fred Picker

 richj20's gear list:richj20's gear list
Panasonic Lumix DMC-GX8 Panasonic Leica 100-400mm F4.0-6.3 ASPH Sony Alpha a7R II
farrukh Senior Member • Posts: 2,720
Re: Photographing a Person: A test of focal lengths

Would be good to see this with an open background

Sergey Borachev Veteran Member • Posts: 4,477
Re: Photographing a Person: A test of focal lengths
2

Ha ha, I think a few people are either a bit insensitive and cannot detect the distortion in the face so well or they simply missed the point of this exercise - that is to see how good the facial features look, and not how the photo looks in terms of colour saturation, resolution, or any other measurements.  Many got it, and it is not that hard if they can compare directly the first and the last photo side by side and see which photo make the man better looking.

This is a good series of photos but they could be better.  If you had used a more direct frontal pose, the man's nose would be closer to the lens and get bigger and more distorted even more than the chin, ears etc at close range.  If you had a lady model instead, people might look for what you are trying to show easier, when they then would tend to look at how good she looks.  If you had taken a really tight close-up, like filling the frame starting from the hairline (just a little hair) on top and only go to the chin, which I like to do,  it would show the distortion more clearly, or the difference.  The human face (round with things sticking out) is flattered with a perspective of a picture taken at a longer distance, at least 5 ft for acceptable result, and that means a 100mm lens or more if you fill the whole frame with a tight head shot.

Even with this series, it should be clear that the man looks better in the last two photos.  For many portraits and head shots, the purpose is to get the subject looking their best.  I know some people do not see the distortion much, but they should be able to tell and select the photo when the person looks better.

Sergey Borachev Veteran Member • Posts: 4,477
Re: Photographing a Person: A test of focal lengths

Sergey Borachev wrote:

Ha ha, I think a few people are either a bit insensitive and cannot detect the distortion in the face so well or they simply missed the point of this exercise - that is to see how good the facial features look, and not how the photo looks in terms of colour saturation, resolution, or any other measurements.  Many got it, and it is not that hard if they can compare directly the first and the last photo side by side and see which photo make the man better looking.

This is a good series of photos but they could be better.  If you had used a more direct frontal pose, the man's nose would be closer to the lens and get bigger and more distorted even more than the chin, ears etc at close range.  If you had a lady model instead, people might look for what you are trying to show easier, when they then would tend to look at how good she looks.  If you had taken a really tight close-up, like filling the frame starting from the hairline (just a little hair) on top and only go to the chin, which I like to do,  it would show the distortion more clearly, or the difference.  The human face (round with things sticking out) is flattered with a perspective of a picture taken at a longer distance, at least 5 ft for acceptable result, and that means a 100mm lens or more if you fill the whole frame with a tight head shot.

Even with this series, it should be clear that the man looks better in the last two photos.  For many portraits and head shots, the purpose is to get the subject looking their best.  I know some people do not see the distortion much, but they should be able to tell and select the photo when the person looks better.

Just clarifying what I meant, by 100mm lens above, I meant a lens with 100mm equivalent FOV, or about 50mm for M43.

I don't think it matters if OP chooses a different background, of if there is more or less blurred background.  I think the exercise is to not exactly to test whether you can tell which FL was used.  Look at the person, like a paying portrait client.  There may be just a slight differeence after a certain FL, but every little bit helps as the objective is to get the person to look his/her best.  It could make a difference when it matters. It takes a good eye to see the difference, but most people can tell whether they like one picture more than another even though it is not apparent that it's due to distortion.  Of course if it shooting a funny character, a very old and wrinkled face or movie villain character, or other non-glamour type of photo, special effects, then it is different.  Pretty basic stuff, actually.

ryan2007 Forum Pro • Posts: 12,001
Re: Photographing a Person: A test of focal lengths

This "test" concludes really nothing.

What I would do is have the subject six feet from the camera. Be sure to use a tripod and set it to the eye level of the subject.  If you shoot manual focus be sure to zoom all the way in and focus that way all focal lengths will be in focus as long as you do Not move the camera.

If focal length is the only variable do not change exposure just focal length.

So, tripod at eye level, lens is six feet from subject.  When you get to the equivalent focal length of about 85 mm you should have a nice head & shoulder portrait.

You do not want to use different focal lengths and just fill the frame with them.

You can repeat from 12 feet back.

Dheorl Veteran Member • Posts: 4,112
Re: Photographing a Person: A test of focal lengths

forpetessake wrote:

richj20 wrote:

BingoCharlie wrote:

So what have you concluded from this experiment?

That there isn't too much difference. If I couldn't see them side by side, I wouldn't think that any were completely unacceptable!

- Richard

You confirmed what traditionally was considered a good FL for people portraits 80-135mm in FF equivalent (most portrait lenses used to be 90-105mm). Below that the image becomes distorted and cartoonish, above that it becomes flat.

Tbh I think this is all down to personal preference. If I hired a portrait photographer and they took the shots with the longer focal lengths I sure as hell wouldn't use them again. In this case I think it makes his neck to fatter and his chin no-existent, whereas I much prefer the shorter focal lengths (N.B. Better lighting with the longer focal lengths might have helped this).

Yes shorted focal lengths can distort stuff, but sometimes that's a good thing if used carefully and moderated. People spend hours in photoshop distoring pictures of ppl, why not just give a girl bigger boobs with the choise of lens/pose?

Sergey Borachev Veteran Member • Posts: 4,477
Re: Photographing a Person: A test of focal lengths

ryan2007 wrote:

This "test" concludes really nothing.

What I would do is have the subject six feet from the camera. Be sure to use a tripod and set it to the eye level of the subject.  If you shoot manual focus be sure to zoom all the way in and focus that way all focal lengths will be in focus as long as you do Not move the camera.

If focal length is the only variable do not change exposure just focal length.

So, tripod at eye level, lens is six feet from subject.  When you get to the equivalent focal length of about 85 mm you should have a nice head & shoulder portrait.

You do not want to use different focal lengths and just fill the frame with them.

He is doing that to show the different perspective distortion at the different distances (moving is necessary to fill the frame).  I think it amounts to the same thing as what you are describing.  Yes, we all understand that perspectives are determined by distance from the subject, not exactly by FL.

You can repeat from 12 feet back.

Ulfric M Douglas Veteran Member • Posts: 4,818
Re: Photographing a Person: A test of focal lengths

A good idea, but spoiled by the first two being lit totally unlike the rest.

Surely you could have done those two again with the later position and lighting?

Also, short focal lengths have more apparent ballooning of the face when it is front-on, not at the angle you chose.

So I'd like to see the same test, with a woman, face-on and with the same lighting for all pictures...

Whenever

Micromegas777 Regular Member • Posts: 191
for me, 75mm rules
Dheorl Veteran Member • Posts: 4,112
Re: Photographing a Person: A test of focal lengths

Sergey Borachev wrote:

Sergey Borachev wrote:

Ha ha, I think a few people are either a bit insensitive and cannot detect the distortion in the face so well or they simply missed the point of this exercise - that is to see how good the facial features look, and not how the photo looks in terms of colour saturation, resolution, or any other measurements.  Many got it, and it is not that hard if they can compare directly the first and the last photo side by side and see which photo make the man better looking.

This is a good series of photos but they could be better.  If you had used a more direct frontal pose, the man's nose would be closer to the lens and get bigger and more distorted even more than the chin, ears etc at close range.  If you had a lady model instead, people might look for what you are trying to show easier, when they then would tend to look at how good she looks.  If you had taken a really tight close-up, like filling the frame starting from the hairline (just a little hair) on top and only go to the chin, which I like to do,  it would show the distortion more clearly, or the difference.  The human face (round with things sticking out) is flattered with a perspective of a picture taken at a longer distance, at least 5 ft for acceptable result, and that means a 100mm lens or more if you fill the whole frame with a tight head shot.

Even with this series, it should be clear that the man looks better in the last two photos.  For many portraits and head shots, the purpose is to get the subject looking their best.  I know some people do not see the distortion much, but they should be able to tell and select the photo when the person looks better.

Just clarifying what I meant, by 100mm lens above, I meant a lens with 100mm equivalent FOV, or about 50mm for M43.

I don't think it matters if OP chooses a different background, of if there is more or less blurred background.  I think the exercise is to not exactly to test whether you can tell which FL was used.  Look at the person, like a paying portrait client.  There may be just a slight differeence after a certain FL, but every little bit helps as the objective is to get the person to look his/her best.  It could make a difference when it matters. It takes a good eye to see the difference, but most people can tell whether they like one picture more than another even though it is not apparent that it's due to distortion.  Of course if it shooting a funny character, a very old and wrinkled face or movie villain character, or other non-glamour type of photo, special effects, then it is different.  Pretty basic stuff, actually.

It's not as to whether you have a "good eye" or not that you see the difference. Also as I said in reply to someone else not everyone always thinks less distorted is better. I personally think photos that are too flat can often look horrible, especially at this angle. The later ones don't do it for me at all.

rsmithgi Senior Member • Posts: 2,114
Re: Photographing a Person: A test of focal lengths

Thanks for the effort put into this. I think it was controlled well enough to be useful. Your framing from shot to shot was very well managed (far better that I would have the patience to do). BTW, it shows the effect of camera to subject distance on how the face is represented not the effect of focal length. http://en.wikipedia.org/wiki/Perspective_distortion_(photography)

I found the effect of flattening of the face at the longer focal lengths/greater distances most interesting. Your framing was done so well that placing the photos side-by-side really pointed out the differences.

Thanks again.

 rsmithgi's gear list:rsmithgi's gear list
Olympus PEN E-P5 Olympus E-M1 Olympus M.Zuiko Digital 45mm F1.8 Panasonic Lumix G 14mm F2.5 ASPH Panasonic Lumix G Vario HD 12-32mm F3.5-5.6 Mega OIS +6 more
gak44 Regular Member • Posts: 300
Re: Photographing a Person: A test of focal lengths

forpetessake wrote:

richj20 wrote:

BingoCharlie wrote:

So what have you concluded from this experiment?

That there isn't too much difference. If I couldn't see them side by side, I wouldn't think that any were completely unacceptable!

- Richard

You confirmed what traditionally was considered a good FL for people portraits 80-135mm in FF equivalent (most portrait lenses used to be 90-105mm). Below that the image becomes distorted and cartoonish, above that it becomes flat.

I agree completely.

 gak44's gear list:gak44's gear list
Sony RX100 III Panasonic FZ1000 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-G5 Panasonic Lumix DMC-GX85 +9 more
Calais225 Junior Member • Posts: 34
Re: Similar test under controlled conditions

This link illustrates exactly what I needed to know.  Now I can see the difference.  Thanks for posting

Vlad S wrote:

There we go: How Lens Focal Length Shapes the Face & Controls Perspective

Tom Axford Veteran Member • Posts: 4,806
The viewing size has an effect too
2

Thanks for posting your tests.

There is an interesting discussion of this at http://www.journalofvision.org/content/12/5/8.full and some quite good animations of the effect.

One point that doesn't seem to have been mentioned in previous posts (unless I have missed it) is the effect of the size of the image as seen by the viewer. Your examples images were posted at quite a modest size, but viewing them much larger would change the visual effect. The distortion apparent in portraits taken with a wide-angle lens looks more unnatural in a thumbnail image than in the same image filling a large screen and viewed from fairly close.

Keyboard shortcuts:
FForum MMy threads