Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
03-16-2012, 07:36 AM   #46
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by bossa Quote
Perspective, as I understand it, is defined as the way an image intersects a flat image plane and thus translates 3d into a 2d representation.
That's reasonably general definition, sure, but it doesn't really shed much light here.

The real problem is that there are two different sense of the term perspective at play here. One is the sense of what is usually called linear perspective - how onjects appear to get smaller as they recede, how parallel lines appear to converge to a "vanishing point", etc. This is all dependent on shooting position *only*. It doesn't matter what lens you shoot with; all parallel lines in a scene will always appear to converge in exactly the same way. But there is also the notion of "perspective distortion", which is dependent on angle of view depicted the image and the size and distance from which you view the image. The effect happens *because* the linear perspective within the image doesn't match the expectations formed by the size of the print and our distance from it.

So, when someone says perspective is dependent only on position and not on focal length, they are right, if they refer to linear perspctive. If they say perspective depends on both angle of view and print size / viewing distance, they are also right, if they refer to perspective distortion. It's when people confuse these two different sense of the word "perspective", or mistake focal length for angle of view, that problems arise.

03-16-2012, 07:44 AM   #47
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by Anvh Quote
They say this 4 sentences later...
Note that perspective distortion is caused by distance, not by the lens per se – two shots of the same scene from the same distance will exhibit identical perspective distortion, regardless of lens used.
That is, I agree, a very misleading sentence. It seems at first the author is confusing the details perspective distortion with those of linear perspective. He does attempt to clarify a few sentences later:

However, crops of these three images with the same coverage will yield the same perspective distortion – the nose will look the same in all three.

I think his point, poorly stated, is that the lens doesn't alter anything about the perspective that was visible from the shooting position, and thus if the images are cropped to the same angle of view, they will appear identical.
03-16-2012, 07:56 AM   #48
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by Anvh Quote
Found these, not the perfect example or the way i wanted it to do but it will do.

One is 28mm lens but cropped
other is 135mm lens, do your best to pick the right one.
Or better yet tell me where the perspective is different between these two.
The perspective is of course not different. Which is a good demonstration of the basics of linear perspective, showing that they are indeed independent of lens.

However, it works fails as a demonstration of the concepts we are discussion that relate to of perspective distortion, because you have cropped them to the same angle of view. Remember, perspective distortion happens when the angle of view depicted in an image differs from the angle subtended by the image itself. These images depict exactly the same angle of view, so of course they will exhibit the same type and degree of perspective distortion when viewing the images at the same size and distance. So I guess it works as a demonstration that focal length per se has nothing to do with it - it is all about angle of view.

But if you wanted to see the difference in perspective distortion that was originally present, you'd post the original uncropped images so we could view them at the same size and distance. The image from the 28mm lens (if that was on APS-C) would make all objects appear exactly the same size as they would if that image were actually a window you were looking through on the scene itself, and the linear perspective (the way the horizontals converge) would match our expectations for how objects at that distance should behave. The image from the 135mm lens would make objects look bigger or closer than they would if that image were just a window looking out on the scene, and linear perspective would be "off" from our expectations of objects of that size and distance.
03-16-2012, 09:18 AM   #49
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
QuoteOriginally posted by Marc Sabatella Quote
The perspective is of course not different. Which is a good demonstration of the basics of linear perspective, showing that they are indeed independent of lens.

However, it works fails as a demonstration of the concepts we are discussion that relate to of perspective distortion, because you have cropped them to the same angle of view. Remember, perspective distortion happens when the angle of view depicted in an image differs from the angle subtended by the image itself. These images depict exactly the same angle of view, so of course they will exhibit the same type and degree of perspective distortion when viewing the images at the same size and distance. So I guess it works as a demonstration that focal length per se has nothing to do with it - it is all about angle of view.

But if you wanted to see the difference in perspective distortion that was originally present, you'd post the original uncropped images so we could view them at the same size and distance. The image from the 28mm lens (if that was on APS-C) would make all objects appear exactly the same size as they would if that image were actually a window you were looking through on the scene itself, and the linear perspective (the way the horizontals converge) would match our expectations for how objects at that distance should behave. The image from the 135mm lens would make objects look bigger or closer than they would if that image were just a window looking out on the scene, and linear perspective would be "off" from our expectations of objects of that size and distance.
Here you go, different photo though.

35mm


210mm


35mm cropped.



about perspective distortion, he says nothing about the AOV and i'm inclined more to believe him then Wiki.
Distortion
QuoteQuote:
The designation 'perspective distortion' is frequently encountered to describe converging verticals and other manifestations of the perspective, but this is questionable as it unjustly suggests a misrepresentation of the reality. There is actually nothing that is distorted in Fig. 6, the photograph reveals a natural perspective. A perspective control lens, which is a device that can can be used to 'correct' converging verticals, creates an illusion of a different perspective, but if anything it's the perspective control lens that presents an unnatural perspective, not the normal picture taking lens. (Strictly speaking a PC lens does not correct anything. Its large image circle and mechanical design merely allow a shift of the image to put the subject into the film frame. The prerequisite for nonconverging verticals is that the film plane is kept parallel to the building.)

Converging verticals often encounter psychological resistance, while photographs with converging horizontals such as a road that narrows toward the horizon are readily accepted or even considered as perspective art. Apart from the appreciation by the audience, however, there is no fundamental difference between the two cases.
This from wiki

They wrote it down very strange in the description but what they did is move away and cropped so that the box is the same size.

Now you've two examples one where the distance is the same and one where they change the distance, you clearly see the differences.

03-16-2012, 09:31 AM   #50
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
Wiki at his best.... *sighs*
This is the distortion you talk about and this a better read on Wiki.
Perspective projection distortion - Wikipedia, the free encyclopedia


http://toothwalker.org/optics/distortion.html
QuoteQuote:
Another so-called form of distortion is 'geometric distortion', which arises when a three-dimensional object is projected on a plane such as a digital sensor. For instance, a sphere in the image center is rendered as a round disk on the film, while a sphere in the image periphery of a lens with a large angle of view is elliptically elongated. People in the corners of a wideangle image are deformed (Fig. 7) and may occasionally be saddled with the typical 'egg face'. This geometric effect is a natural consequence of the projection obliquity; the use of the word distortion wrongly suggests an imaging anomaly. In contrast to optical distortion, perspective and geometric distortions are no lens aberrations.
03-16-2012, 05:06 PM   #51
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by Anvh Quote
Here you go, different photo though.
Stand in the middle of the same road with a laptop showing these images and view full screen at arms length. The objects in the 35mm image look the same size as the actual objects in the street. Now switch to the 135mm lens image. the objects look larger than in real life. Or, if you preend the foreground car is simply closer than it actually is, the perspective is off in terms of where the vanishing points would be on a car at that distance.

QuoteQuote:
about perspective distortion, he says nothing about the AOV and i'm inclined more to believe him then Wiki.
He is unfortunaltely talking about a different phemonenon still. Read my own explanations. Is there something about them you find less than than clear? If so, I'd be happy to answer specific questions.
03-17-2012, 06:42 AM   #52
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
I thought we were talking about how the objective will distort the subject...
Yes the grey car looks bigger relative to the "frame" but not relative to for example the blue car in front so the perspective in the image is not changed.

03-17-2012, 07:36 AM   #53
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
As I have been saying, there are several different aspects to perspective. Different focal lengths do not *produce* distortion - linear perspective always remains unchanged (except with tilt/shift lenses, which change the rules a bit). But "perspective distortion" is a very real phenomenon that I have been trying to explain, and it definitely *is* dependent on angle of view.

To illustrate with an example, consider the following from the Wikipedia article on perspective distortion:



If the eye is to accept the front water bottle as being the same size in all three pictures, the illusion is created that the rear water bottle is a different size or a different distance behind the front bottle from one picture to the next. When presumably, that was not the case - it's the same scene, shot with different lenses from different positions. This is perspective distortion. It doesn't mean the perspective is *actually* distorted by the lens - each image is a true representation of the reality at the point where the image was shot. But when we view the resulting images, the eye is fooled into thinking the rear bottle is at a different size or distance from one shot to the next. If the photographer had extended this example and shot another image from much further back with a 200mm lens (keeping the apparent size of the front bottle constant), the two water bottles would appear virtually the same size in the frame, as if they were right next to each other.

The illusion is caused by the fact that we have no way of knowing how close the photographer really was to the scene, so our eyes use the size of the closest object as a gauge to guess. Since the front bottle appears the same size in each frame, we first assume they depict the bottle at the same distance. That is why we perceive the back bottle to have moved closer and/or gotten larger.

It is true that the change in perspective itself is created by shooting position, not by angle of view. But had there not *also* been a change in angle of view (by changing focal length), then we would not have perceived a discrepancy when we viewed the images. If all three shots had been taken with the same angle of view but from different positions, the front bottle would *appear* closer to us in the shot that was actually taken from closer, since it would appear larger. Hence fact that the back bottle appeared much smaller in comparison would seem perfectly normal. And in the hot from further back, the front bottle would *appear* further away, since it would appear smaller. Hence, the fact that the back bottle appeared almost the same size in comparison would seem perfectly normal. It is the fact that the photographer changed both position *and* angle of view that creates the illusion of the rear bottle moving closer and/or getting larger.

This illusion is also used in film/video. You've probably seen it in any number of movies, usually psychological thrillers in which the director is trying to illustrate a sense of altered reality. You zoom in on the subject while simultaneously moving the camera further away (or zoom out while moving closer), keeping the subject the same size in the frame, and it looks like the background is getting larger / closer. Here is one example I found online. After the first few seconds, the illusion is created that the background is moving in on the subjects, or that they are on the bow of a ship that is moving into the background:

03-17-2012, 10:07 AM   #54
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
I understand that but like you say it's the shooting position that change the perspective and not the AOV but i agree that indirectly the AOV influence the perspective, or to say it better the AOV influence your shooting position.


I believe you're saying the same thing so hopefully we can all agree and go on.
03-17-2012, 10:46 AM   #55
Veteran Member
westmill's Avatar

Join Date: Jan 2012
Location: Stoke on Trent
Photos: Gallery | Albums
Posts: 1,146
QuoteOriginally posted by Marc Sabatella Quote
As I have been saying, there are several different aspects to perspective. Different focal lengths do not *produce* distortion - linear perspective always remains unchanged (except with tilt/shift lenses, which change the rules a bit). But "perspective distortion" is a very real phenomenon that I have been trying to explain, and it definitely *is* dependent on angle of view.

To illustrate with an example, consider the following from the Wikipedia article on perspective distortion:



If the eye is to accept the front water bottle as being the same size in all three pictures, the illusion is created that the rear water bottle is a different size or a different distance behind the front bottle from one picture to the next. When presumably, that was not the case - it's the same scene, shot with different lenses from different positions. This is perspective distortion. It doesn't mean the perspective is *actually* distorted by the lens - each image is a true representation of the reality at the point where the image was shot. But when we view the resulting images, the eye is fooled into thinking the rear bottle is at a different size or distance from one shot to the next. If the photographer had extended this example and shot another image from much further back with a 200mm lens (keeping the apparent size of the front bottle constant), the two water bottles would appear virtually the same size in the frame, as if they were right next to each other.

The illusion is caused by the fact that we have no way of knowing how close the photographer really was to the scene, so our eyes use the size of the closest object as a gauge to guess. Since the front bottle appears the same size in each frame, we first assume they depict the bottle at the same distance. That is why we perceive the back bottle to have moved closer and/or gotten larger.

It is true that the change in perspective itself is created by shooting position, not by angle of view. But had there not *also* been a change in angle of view (by changing focal length), then we would not have perceived a discrepancy when we viewed the images. If all three shots had been taken with the same angle of view but from different positions, the front bottle would *appear* closer to us in the shot that was actually taken from closer, since it would appear larger. Hence fact that the back bottle appeared much smaller in comparison would seem perfectly normal. And in the hot from further back, the front bottle would *appear* further away, since it would appear smaller. Hence, the fact that the back bottle appeared almost the same size in comparison would seem perfectly normal. It is the fact that the photographer changed both position *and* angle of view that creates the illusion of the rear bottle moving closer and/or getting larger.

This illusion is also used in film/video. You've probably seen it in any number of movies, usually psychological thrillers in which the director is trying to illustrate a sense of altered reality. You zoom in on the subject while simultaneously moving the camera further away (or zoom out while moving closer), keeping the subject the same size in the frame, and it looks like the background is getting larger / closer. Here is one example I found online. After the first few seconds, the illusion is created that the background is moving in on the subjects, or that they are on the bow of a ship that is moving into the background:

Dolly Zoom - YouTube
You have done a first class job in explaining it
but its my guess it will all have been a bit of a waste.
You have the patience of a saint lol
Much respect !
03-17-2012, 04:04 PM   #56
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by Anvh Quote
I understand that but like you say it's the shooting position that change the perspective and not the AOV but i agree that indirectly the AOV influence the perspective, or to say it better the AOV influence your shooting position.
Then you didn't udnerstand what I wrote at all. Please read it again, *carefully*. You are *complewly* missing the point about what perspective distortion is. The effect of perspctive distortion is completwly dependent on angle of view, and doesn't depend at all on shooting position. It's just easier to demeonstrate the effect if we change shooting position as well as angle of view.

Said another way: there is no possible shooting position from which an uncropped image from a wide angle lens will be devoid of perspective distortion. Just as there is no possible shooting position from which an image from a telephoto lens will lack perspective distortion. And there is no possible shooting position from which you can induce perspective dostortion in a normal lens.

Again, linear perspective is dependent on shooting position, yes. but perspective distortion is not - it is dependent on angle of view.
03-17-2012, 04:49 PM   #57
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
You can jump up and down or stand on your head but i don't get it then.

What i see with the photos and the video in your link is change in distance that change how large the subjects in the background appear relative to the other subjects and with the AOV you keep the subject of interest the same size relative to the frame.


I can even calculate it for you that the apparent height of the two bottles becomes bigger the closer you move.
Better yet here is a link.
http://y-intercept.com/rich/rediscovery.html

More photo examples.
www.mhohner.de/essays/myths.php#focalper

Ps. thought i didn't understand the definition maybe of "perspective distortion" but that isn't the case, all source i found besides Wiki talk about something else...
www.photozone.de/lens-terminology
www.cambridgeincolour.com/tutorials/lens-corrections.htm
And toothwalker, the one i've already often linked to.

Last edited by Anvh; 03-17-2012 at 05:49 PM.
03-19-2012, 10:13 AM   #58
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by Anvh Quote
What i see with the photos and the video in your link is change in distance that change how large the subjects in the background appear relative to the other subjects and with the AOV you keep the subject of interest the same size relative to the frame.
Yes, but consider the video - If they had only changed distance but not changed AOV, the effect would be extremely different. It would create the illusion of the background closing in, it would simply look like the viewer was getting further away. While in this particular optical illusion, changing distance was part of it, chanigng APV was absolutely necessary to get the effect as wel.

But that's just this particular optical illusion. The basic concept does *not* depend on distance. I think you are overthinking this, as the effect is actually quite simple, and ao it is hard to explain in words because a long explanation makes it seem like something complicaed qhen it is anything but. Plus, the fact that are many slightly different concepts being discussed doesn't help. But the specific phenonomenon I am talking about most definitely is dependent only on AOV, not on distance. Really and truly. I have been trying to think about how to construct a more convincing demonstration of this, and here's what I have come up with - note it requires you to do the legwork, though.

Take a series of three pictures at different focal lengths (18, 35, and 135, say) of two objects of the same size that are at different distances from your shooting position (one meter apart, say). In other words, the exact same setup as the Wikipedia example I referenced previously, but take all three pictures from the same distance rather than changing distance as they did. Stand anywhere you like when you take the pictures, but take them all from the same position - you need to understand that the effect has *nothing* to with distance to subject and everything to do with AOV.

Now, have those objects at hand as you display the pictures full screen, starting with the "normal" version. Set up the objects in the same relative positions they were when you shot the picture. Adjust the size of the image and/or the position of the monitor so that the sizes and relative positions of the objects in the image exactly match the sizes and positions of the objects in real life, so that if you shift your gaze from one to the other, everything seems to be the same. The idea of "normal" is that this should be possible with a typcial image size viewed from.a typical distance, although personally I find it requires me to view an image larger or from closer than is actually natural. No matter, the idea is to show the diffeence that is caused by AOV, not to nit pick about which specific AOV corresponds to "normal" in this sense.

Now, display the wide angle picture on your monitor at the same size and view it from the same distance and try to position the objects to match what you see in the image. Since the AOV is wider, the objects will appear smaller, thus it will require you to place the objects further away from you in order to get them to match what you see on the screen. That much is so obvious it is barely worth mentioning. The perhaps less obvious part is that once you get the objects positioned so they appear to match the image on screen perfectly, you will that not only are they farther *away*, but also farther *apart* than they were orginally. And if you replace the objects with smaller objects that you can place in the same positions as the oirignal, you will fins you need objects of *different* sizes to get them to match up, whereas the original objects were the same size.

Similarly, try displaying the telepho to image on screen and then setting up the onjects to match. Because the objects appear larger on screen, you will need to place them closer to you on order to get the sizes to match - but also *closer to each other* than they originally were.

That is the essence of perspective distortion - the diea that no matter what ditance you shoot from, when you go to actually view the image at a typical size and from a typical disance, objects shot will a wide angle lens will appear farther apart than they actually are, and objects shot with a telephoto lens will apear closer together than they actually are. It's an illusion caused by our "need" to interpret any viewed image as a "window" on reality, causing us to misjudge all distances in images shot with anything but a "normal" AOV.
03-21-2012, 12:08 PM   #59
Veteran Member
Anvh's Avatar

Join Date: Sep 2011
Posts: 4,616
I see what you mean and i already saw it but it won't change my view on this.
Let me explain why.

The magic number is 53 degrees, that's the diagonal of the angel of view from a typical viewing distance and so lenses with the same AOV will have no distrotion because the distance at which the photo is taken would be the same as the viewing distance.

With a wide angle lens you need to stand closer to get the same framing of the subject so the perspective from the typical viewing distance will be distorted and vice versa for tele-photo lenses.

But it's the distance of either the view or the photographer that change the perspective not directly the AOV.
If you can make a big print and force the people to stand closer to it then you need to use a wider lens.

In the examples posted the relative scale of the objects in the photo stay the same when you stay in the same position but change the AOV.
Because the scale between the objects stays the same the perspective is not changed like in your examples of the bottles, that's all done because the distances is different not the AOV.
If they shot those bottles at the same distance then the relative distance between the two bottles will be the same and so the depth/distance between the bottles will be preserved as the same as well.
Just looked at the photo i posted with the cars, look at the uncropped photos, sure the size of the cars look different but you won't say that the distance between the two jeeps is different and so the perspective is the same.
This link explains the rules and the math behind it y-intercept - Rediscovery


To explain it in one sentence.
Perspective is changed with position and the AOV influence your position.

Last edited by Anvh; 03-21-2012 at 12:22 PM.
03-22-2012, 03:43 PM   #60
Veteran Member
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,685
QuoteOriginally posted by Anvh Quote
I see what you mean and i already saw it but it won't change my view on this.
This is not a matter of opinion. It is science. You might have the "view" that world is flat, but that will not make it so - the world is what it, regardless of your "view". Similarly, the phenomenon known as perspective distortion - as I have described it - is *not* dependent on distance to subject and will never be dependent on distance to subject, regardless of what your "view" is. It is dependent on angle of view and will always be dependent on angle of view. You can choose to read *and understand* my explanation to see why you are wrong, or you can simply be wrong and not understand why, but either way, you are wrong. I don't know what else I can possibly say at this point except to suggest you read the explanation of the phenomenon again.

QuoteQuote:
But it's the distance of either the view or the photographer that change the perspective not directly the AOV.
Yes, and once again, I have never ever ever ever ever ever ever said anything different. *Perspective* is dependent on viewing position and viewing position only, yes, sicnece is very clear on that. But science is also eqully clear on the fact that the phenomenon known as perspective distortion is not and never will be dependent on shooting distance. No matter what distance you shoot from, a wide angle lens will *always* produce an image with that exhibits perspective distortion, and so will a telephoto lens. Stand six inches away and this true, stand six feet away and this is true, stand six meters away and this is true, stand six miles away and this is true, standard six light years away and this is true. Perspective distortion depends not on shooting position, but only on the relative angles of view of the scene versus the print itself. You can cause it to be otherwise by wishing or by choosing not to understand.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
50mm, camera, cameras, film, k-x, lens, pentax help, photography, question, sensor
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
focal length on crop sensor question? littledrawe Troubleshooting and Beginner Help 14 09-28-2011 03:06 AM
focal length pentax k-x Photographic Technique 9 11-22-2010 12:41 PM
focal length ewig Troubleshooting and Beginner Help 11 07-23-2010 09:32 PM
focal length landscaped1 Troubleshooting and Beginner Help 48 05-05-2010 03:04 PM



All times are GMT -7. The time now is 05:02 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top