Distortion and Distance between the Subject and the Background in MFT system

[pranavanarp]

Hello all,

This is my first post here at Mu4/3 and this feels quite exciting! I own a Gh5s and I am slowly starting to love this whole realm of M4/3. I upgraded from Canon 600d which is a aps-c camera so this whole eco system of M4/3 is quite new to me.

I am trying to learn as much as possible about this new system and there are always some questions whose answers are difficult to find on the internet maybe because I can't articulate some topics quite well. So, I will give it a try here:

I have two questions. First, regarding focal lengths. I have learned that a 25mm in mft is equivalent to a 50mm in full frame. I also came to know that if you're a 'bokeh freak' it's better to have full frame but that's not much of a concern for me. My question is related to the distortion you get on portrait faces when they are taken from different focal lengths. For instance, on a full frame body, the 85mm portrait has a more natural shape to the human face, when say compared to a 35mm. So, if I want to get that same look with the minimum distortion as that of 85mm on a full frame, will the 42.5mm focal length work for me for my Gh5s or do I actually need a 85mm lens to have that same exact look like a full frame body?

Second question is related to the difference in distance produced when different focal lengths are used (I don't know the exact term for it). For instance, say in a 50mm lens on a full frame body the background is much further than the same picture on a 85mm body. So bringing this equation to my mft system, to get the same distance as that of 50mm on a full frame body, will my 25mm do the same thing, or will the 25mm on my camera push the background much further than the actual 50mm on a full frame? I hope my question makes sense.

Also, it would be really helpful if you could post some links to pictures doing these comparisons.

Thank you!

 

[exakta]

Just choose focal lengths that are 1/2 that of FF. So, yes 42.5mm for 85mm. You will get the same foreground-background separation and the same perspective distortion.

Because the m43 frame is a 4:3 ratio where FF is 3:2, the actual angle of view on the diagonal is a bit different but it's close enough for choosing focal lengths by the 1/2 rule.

 

[oldracer]

Read this: https://en.wikipedia.org/wiki/Perspective_distortion_(photography)

 

[DeeJayK]

In every way that matters (angle of view, perspective distortion, reach, etc.) a 42.5mm lens on will perform the same as an 85mm lens on a "full frame" camera.

These lenses would be equivalent to a roughly 52.5mm lens (if such a thing existed) on an APS-C camera like your old 600D.

- K

 

[pranavanarp]

Thank you all for the answers. Thanks for clearing up!

 

[oldracer]

... My question is related to the distortion you get on portrait faces when they are taken from different focal lengths.

This is a common misunderstanding. If you think about it though, your subject's face does not distort or even change a little bit depending on what lens you have mounted.

What changes is our perspective on the face based on distance. If you stand very close to the person, their nose is significantly closer to you than their their eyes and their ears are yet further away. If you stand 100meters away, nose, eyes, and ears are all approximately the same distance. So when close, that nose looks huge, but when far away it does not.

The classic "portrait" focal length range of 85-105mm (35mm equivalent) places the photographer at a distance where their perspective on their subject looks normal. You do not normally stand 30cm from your subject or 100m from your subject. So ... the short end of the range works best for longer shots like whole body, head and torso, etc. With 85mm a tight head shot puts the photographer a little too close and the nose may look unacceptably big. So ... tight head shots are the province of lenses like my old Nikon 105 -- my favorite portrait lens because I like tight head shots.

These are the basic rules, then for "normal perspective"portraits: shorter for shots that include more of the subject, longer as the shots get tighter. But there are no photo police enforcing these rules; people break them all the time in search of the "perfect portrait" suiting their standards. But I think it's important to understand where we came from.

 

[sbm]

What changes is our perspective on the face based on distance.

My initial reaction to this was, "well, duh," but that thought dissipated as I read on.

My next one was to think you weren't making sense and roll my eyes, considering, "I mean sure, if I wanted to stand 20 feet away, I could take a shot with 100mm and 35mm and just crop the 35mm to get the same photo."

Shortly after that your point started to sink in. It's about choosing the focal length for how far away you want to be from your subject, which is usually near the top of my list when picking a lens for a subject anyway – but that's more about what's in the frame, less so about the distance between their nose and ears relative to me.

Fascinating.

Now, back to the OP's topic... I'm by no means an expert on optics and glass, so I'll be asking more questions than answering. But, re: oldracer's post... there is edge distortion, right? I mean, look at images made with a fish-eye. Is the center of the frame the same regardless? Is that why the crop factor doesn't matter?

My guess is the distortion isn't nonexistent, but so vanishingly impactful that it's unworthy of consideration 99% of the time. If someone can confirm or deny this suspicion, I feel we can avoid one of those lengthy, confusing threads about the minutia of glass optics (which, I'll bet already exists elsewhere on this forum).

 

[oldracer]

... Now, back to the OP's topic... I'm by no means an expert on optics and glass, so I'll be asking more questions than answering. But, re: oldracer's post... there is edge distortion, right? I mean, look at images made with a fish-eye. Is the center of the frame the same regardless? Is that why the crop factor doesn't matter?

My guess is the distortion isn't nonexistent, but so vanishingly impactful that it's unworthy of consideration 99% of the time. If someone can confirm or deny this suspicion, I feel we can avoid one of those lengthy, confusing threads about the minutia of glass optics (which, I'll bet already exists elsewhere on this forum).

Well any lens has some distortion and other errors. Pincushion, barrel, chromatic aberration, etc. Distortion is probably weakly correlated to focal length since it is harder to make a good ultrawide vs a more normal focal length. It's probably weakly correlated with zoom and superzoom range too, as those are hard to do. But none of that has anything to do with big nose effects. If you read that Wikipedia article I linked a couple of times that might help.

Here's an experiment: Take a set of lenses, say 24mm, 50mm, 100mm and 200mm. (35 mm equivalent). Place your subject so that they are suitably framed in the 200 and lock down the tripod. Now shoot a set of portraits at each focal length, all from the same distance. Crop out the subject in each portrait to the exact composition of the 200mm shot. What you will find is that all of the noses are the same size.

Now, I'm not against big noses. I have a big nose shot of a photographer friend that makes me smile every time I see it. But she hates it with a passion! So I'd suggest that understanding the basics is important but going outside the conventional envelope is A Good Thing from time to time, too.

 

[sbm]

Here's an experiment: Take a set of lenses, say 24mm, 50mm, 100mm and 200mm. (35 mm equivalent). Place your subject so that they are suitably framed in the 200 and lock down the tripod. Now shoot a set of portraits at each focal length, all from the same distance. Crop out the subject in each portrait to the exact composition of the 200mm shot. What you will find is that all of the noses are the same size.

Or, check out the wikipedia article you shared. It shows this, plus the different focal lengths taken from different distances, encompassing the same frame.

Though there is usually value in doing it (and seeing it) yourself.

So I'd suggest that understanding the basics is important but going outside the conventional envelope is A Good Thing from time to time, too.

Bingo! Learn the rules so you know how best to break them.

 

[pranavanarp]

This is a common misunderstanding. If you think about it though, your subject's face does not distort or even change a little bit depending on what lens you have mounted.

What changes is our perspective on the face based on distance. If you stand very close to the person, their nose is significantly closer to you than their their eyes and their ears are yet further away. If you stand 100meters away, nose, eyes, and ears are all approximately the same distance. So when close, that nose looks huge, but when far away it does not.

The classic "portrait" focal length range of 85-105mm (35mm equivalent) places the photographer at a distance where their perspective on their subject looks normal. You do not normally stand 30cm from your subject or 100m from your subject. So ... the short end of the range works best for longer shots like whole body, head and torso, etc. With 85mm a tight head shot puts the photographer a little too close and the nose may look unacceptably big. So ... tight head shots are the province of lenses like my old Nikon 105 -- my favorite portrait lens because I like tight head shots.

These are the basic rules, then for "normal perspective"portraits: shorter for shots that include more of the subject, longer as the shots get tighter. But there are no photo police enforcing these rules; people break them all the time in search of the "perfect portrait" suiting their standards. But I think it's important to understand where we came from.

This was incredibly helpful. Thank you very much. Also, I came across this video on youtube (

) that helped me clear the understand that the distortion and the background/foreground distance comes out the same in a 25mm lens in a mft body and a 50mm in a full frame, although the video uses a full frame and an aps-c format camera.

 

[pranavanarp]

One another concern I had regarding similar topic was the argument that Full Frame users have over mft lenses that a 1.4 advertised lens for a mft body is actually 2.8. Now, with my limited understanding, that makes me feel that a 2.8 aperture would also affect the exposure meaning that the 1.4 lens of the mft actually allows light equivalent to a 2.8 lens of a full frame. But I looked over the internet and that seems to be wrong, the difference in aperture does not affect the exposure but only the 'bokeh' effect. Is this a correct understanding?

 

[DeeJayK]

One another concern I had regarding similar topic was the argument that Full Frame users have over mft lenses that a 1.4 advertised lens for a mft body is actually 2.8. Now, with my limited understanding, that makes me feel that a 2.8 aperture would also affect the exposure meaning that the 1.4 lens of the mft actually allows light equivalent to a 2.8 lens of a full frame. But I looked over the internet and that seems to be wrong, the difference in aperture does not affect the exposure but only the 'bokeh' effect. Is this a correct understanding?

Yes, at the risk of wading into a discussion that tends to always go sideways, "aperture equivalence" pertains to depth of field only, not to exposure.

- K

 

[pranavanarp]

"aperture equivalence" pertains to depth of field only, not to exposure.

cool, thanks for pointing out thats clears up the many misunderstandings I had regarding m4/3 systems.

 

[ralf-11]

Suppose I put the model at a certain distance from the background, and want her face to fill the frame...

will m43 allow me to blur the background as much as I can with a fast aperture FF lens?

Location of camera is a variable, but distance of background from model is fixed.

 

[G3user]

Try http://m43photo.blogspot.com/2010/04/portrait-lens.html for some examples.

 

[oldracer]

Suppose I put the model at a certain distance from the background, and want her face to fill the frame...

will m43 allow me to blur the background as much as I can with a fast aperture FF lens?

Location of camera is a variable, but distance of background from model is fixed.

Probably not fair to hijack the OP's thread and move it into the realm of depth of field. IMO DOF is far less important than the OP getting an understanding of perspective "distortion."

But the answer to your question is "it depends." Enough said. Other threads have hashed and rehashed this point.

 

[pranavanarp]

Probably not fair to hijack the OP's thread

Sorry for a really silly question but what does OP stand for? Optical Perspective?

 

[pranavanarp]

"I mean sure, if I wanted to stand 20 feet away, I could take a shot with 100mm and 35mm and just crop the 35mm to get the same photo."

This really got me thinking. I mean of course this is not a practical solution and specially for my camera Gh5s which is a 10 mp model, something like this would be near to impossible. However, I was curious to understand the workings of the glasses- does the same distortion and compression take place if I take two pictures one with 100mm lens and the other with 35mm, say 20 ft away, and crop the 35mm picture to get the same picture?

 

[John King]

OP = Original Post/er.

 

[PakkyT]

However, I was curious to understand the workings of the glasses- does the same distortion and compression take place if I take two pictures one with 100mm lens and the other with 35mm, say 20 ft away, and crop the 35mm picture to get the same picture?

Assuming you are talking about perspective distortion and use the same aperture for both shots, then changing the focal length only changes the "crop" on the photo. So yes they would be the same photo other than how wide of an angle you see of the scene you are shooting.