Discussion in 'AP Magazine Feedback & Suggestions' started by Terrywoodenpic, Jul 8, 2012.
Depends on viewing distance too.
My question is this, are these facts?
Let's take lens construction. Now correct me if I am wrong but some lenses both telephoto and wides and suffer from distortion that surely has a effect. If the term perspective is the same as POV (is it?) then that not a true fact.
Terry only quotes distance from the lens as the a factor. What about angle. You could have a sphere of cameras (bullet time photography) around a object all at the same distance. But the angles are varying. So surely distance is not the only factor on perspective.
Yes, you can draw out images without reference to a lens (or the perfect lens). But in photography that does not happen even the best lenses will have some error. No system is 100%. You can simulate a view using CGI as if the lens was 100% but it might look unnatural. Like say isometric drawing. Even on a very long telephoto you still have some convergence.
But the important thing for photographers is to use focal length probably above 50mm (35mm film equiv) for portraits to give a natural look to the shots. Unless you want a unsettling effect.
I believe I am correct in that wide angle is a invention of photography because the human is not a wide angle. Where as one could argue that our telephoto is more like digital zoom than focal length zoom. But then we don't know if our brain might be doing a photoshop DOF.
Perspective will only be the same if BOTH camera and subject are fixed relative to each other. When there is a ring of cameras around the subject each camera has its own unique perspective.
Lens distortion is IRRELEVANT. As has already been said, every lens distorts, some more so than others but, other than a Fisheye where distortion is deliberate, this is usually negligible. As I said in an earlier post, panoramic stitch ups are only practicable if distortion is negligible otherwise it would be almost impossible to align the 'cuts' of the different pics because the cut across the middle of one pic would not correspond to the cut across the edge of the adjacent shot.
As I said earlier too, the only difference when different lenses or even different lens apertures are used is that the depth of field will change. This could be confused with "perspective" but the relative positions of different artifacts in the field of view will NOT change although might be more blurred than others.
In my opinion, yes. And I can (but won't) cite umpteen references to prove it. I'm writing a book on photography, and I have several example photos to prove it, with pairs of photos of an object 3 feet away and one nearer 300 yards away. One wide angle, one telephoto. And I can assure you that the wide angle photo when cropped is identical to the tele one.
As to depth of field being equally great before and behind, this is only true at 1:1. At more common photographic distances, there's more behind than in front.
The question of whether the depth of field was identical if you enlarged was one that caused Victor Blackman to have (as he memorably phrased it) "circles of confusion about his eyes". His answer (with an illustration if I remember correctly) was in AP a while ago (1963???).
I can't help but feel that you're not using the word "perspective" in the same way that I am.
Oh do stop being silly. You know what's being discussed.
You always do, but the more pseudo-scientific claptrap you introduce, the rarer it is that you are.
You could hardly be more wrong - check out the angle of view of (a) one eye, and (b) both eyes combined, and you'll find it's much wider than anything we consider as photographers to be a "standard" lens, around 120-130 degrees for both eyes combined - that's pretty much equivalent to my Sigma 12-24mm. At 12mm. On full frame.
What on earth is that supposed to mean?
Well we know no actual optical zoom occurs when we concentrate on a part of a scene.
As for wide angle. Most photographers feel that 35mm is around human eye view. That's not wide angle say of 28mm or 24mm.
Yes, I know what is being discussed the alleged mis-use of the term perspective in AskAP which I not read yet.
Or was Terry raising something else.
Well, in that case the magazine must have it wrong. Someone needs a good smacking.
End of the day. Not my fault. I know what focal length works for me shooting portraits. So I can sleep tonight.
My mind is boggling - I really don't know where you get this stuff.
Let's try from the other end.
Take an image using a 50mm lens.
Crop it to one half of the original size.
Expand the cropped image until it occupies the same frame size as the original image.
What you now have is the image that would have resulted had you used a 100mm lens initially.
But it is the same image as you took originally and the perspective can not have changed.
No problem with that digital zoom works in that way. But it only applies to the centre of the image. Not across the whole wide angle shot.
Terry did not make that clear in his post.
Until I read the piece I cannot comment really on the use or mis-use of the term perspective.
Which I believe is the core of the Terry's first post.
Anyway it's the magazine's problem if they get it wrong they are publishing into the public domain and instructing other photographers. Same things as when they got the sensor size wrong for a camera review. I'll let Mr Jacobs sort it out.
I'm trying to keep up, I really am.
What I have always understood (trying layman's terms here) is that from a fixed viewpoint, any given lens will either include a wide angle of view or a narrower one, but that these are identical, provided the viewer doesn't move; it's just that the narrower angle of view (telephoto) means that a smaller part of the scene fills the sensor (hence the 'magnification' of a smaller part of the overall scene).
Then, the point of focus and DOF define how much of the scene is in focus (but will not alter any angles of view etc - the scene is still 'all there', just that, at a wider aperture, less of it will be distinct)
So a lens does not 'manipulate' (not sure about extreme WA or T+S?) a scene, simply include or isolate (wide and telephoto) different parts of it.
If, having read the above posts, I have managed to fail to grasp the concept, apologies!
You've got it right. Forgive me if the following seems simplistic - it is.
The easy way to think about it is to imagine that you are looking at a scene and that you have two frames in your hands, one large and one small. Look though those frames at the scene. Apart from the overall field of view, nothing changes between them. The relative size and position of specific objects in those frames remains identical.
Now, instead of two frames, you have a camera firmly mounted on a tripod and two lenses, one a modest wide angle and the other a modest telephoto. Take a photo with each. Apart from the image scale (each filling the same sensor area), the perspective is identical and the relative size and position of two present objects in both photos remains the same.
Yes, things would look a little different if the WA was actually a Fish Eye lens where distortion is the norm. But otherwise the lens makers have gone to endless trouble to minimise any distortion. That is one reason why they use aspheric lens elements just to reduce distortion and improve overall IQ.
"Superzooms" have acquired a bad reputation for distortion, yet even that is all relative. The fact is that it is almost impossible to make a zoom lens that is effectively free from distortion at both ends of its range. A prime lens is usually better than any zoom in this respect, yet even then, the tiny amount of distortion remaining can vary as the lens is focused. Yet, most photographers don't even notice it and it is only revealed in a bench test. I have a Sony 18 - 250, yet I have never been aware of the effect of distortion in any of the thousands of photos I have taken with it. Yes, it might be obvious if I take a seascape with the horizon very close to the top edge of the frame. Being aware of this I would normally avoid that type of photo! However, as far as perspective goes, that would be a non problem.
It has been argued that if you take a crop from the middle of a frame and one from a corner having moved the camera so that the two are about the same, then they can never be identical. I would say that unless it is an extreme WA lens, then they will be, for all practical purposes, identical. Yes, there may be worse IQ from the corner crop, but two objects in those crops will have the same relative size and position!
thank you...I was losing the will to live....
OK, let's start again from the basics:
Perspective is no more nor less than the relative magnification of subject elements within a scene. For any given viewpoint, the relative magnifications of each element will be the same regardless of the lens used - if a foreground object is twice the size of a background object on a standard lens, it will be twice the size of that background object on a wideangle or a telephoto lens. What does change in these circumstances is the field of view and the absolute magnification - a wideangle will give a wider view and less magnification, a telephoto a narrower view and greater magnification, but just to repeat, neither changes the relative magnification of various subjects in the frame.
Now it's true that lenses are used to manipulate perspective, but that's by using the angle of view and absolute magnification properties to change the viewpoint: to fill the frame with a wideangle, you get closer to the foreground object, so it's magnification relative to a background object is greater - that's why portraits taken with a wideangle are generally unflattering, as the nose appears much larger in proportion to the rest of the face. With a telephoto, you stand further away to take a shot so that it's all in frame, and by so doing, the relative magnification of foreground to background objects is reduced - it appears to compress space. But what's changing the perspective - that is the relative magnification - is just the viewpoint. What the lens does is to force you to change the viewpoint to get the main subject the size in the frame that you want.
That's all there is to it. But, Mr Stoddart, what you've written is just a lot of incorrect, confusing mumbo-jumbo. Which might not matter if you weren't "publishing into the public domain and instructing other photographers". Fair enough for you to say that what was published in the magazine was "Not my fault", but the rubbish you have written in here clearly is. "Someone needs a good smacking" - quite, and for the garbage you've posted in this thread, it's you.
"Until I read the piece I cannot comment really on the use or mis-use of the term perspective."
Given your total lack of knowledge on the subject, it would be a very good thing indeed for everyone if you refrained from comment on this subject permanently.
It's more likely to be true at greater than 1:1! Most of the time it just looks like there's more behind.....it's an illusion depending upon how well defined the depth of focus is.
Just to add my 2p worth and also because another article in this weeks AP found me in the garden with a camera I took two images from the same tripod position at the extreme ends of a 12-60 zoom lens. I cropped the centre portion of one:
and just resized the other without cropping
Of course the relative positions of the plants are identical - as Nick and others have explained in words - also the DOF of the 60mm version is clearly smaller than the WA crop
Well done.... and thank you.
Though I expect some will claim trickery was involved??????
Nothing would surprise me
I stand by what I wrote. I could cite authorities (and even equations) to prove my point but I simply don't think it's worth the effort. Did you mean "depth of focus", by the way? And what did you mean by it, because I fail to understand that point at all.
Actually stitched Panoramas are a special case and the advanced software used has no difficulty stitching even Fisheye shots.
"Stitching" in panorama speak, does not mean the joining together but it means the reprojecting of the image on to a spherical surface so that the images can be repositioned. These are then reprojected on to a flat surface.
These examples show how the close part of a shot is enlarged (distorted) in a normal cylindrical projection.
And how a rectiperspective projection can show all radial lines as straight rather than curved . These were all multiple shots with a cropped 17mm lens and cover at least 140 degrees.
Normal cylindrical projection.
Normal cylindrical projection
Same stiched shot but output as a rectiperspctive projection.
I always thought it was 1/3 in front and 2/3 behind the point of focus?
Separate names with a comma.