The heck with the crop factor....

[Scott Root]

Advertisement (gone after registration)

Stephen,

 

Your document was greatly helpful and I stand corrected on the dof argument and I now agree that dof is affected by the size of sensor just as dof is affected by aperture settings. Particularly helpful to me in the process of getting the model right in my mind is to realize that each ray of light entering the lens contributes to the whole picture; it is not a case that a ray of light settles on one point of the sensor which is the line of reasoning I used in my previous post. This then accounts for why a reduced size of sensor has the same effect as reducing the amount of light entering the lens by means of the aperture setting. The dof issue is now resolved for me, thanks.

 

However, there are two perspective issues that I cannot find explained satisfactorily anywhere on the internet and most sites seem to muddle the two phenomena.

 

Plane perspective seems to refer to the perspective a given focal length renders a picture at a standard distance. For this reason individuals claim a 35 mm lens gives the same perspective on a 1.3 cropped sensor as a 50mm lens on a full frame sensor, give or take a few millimeters which is not the point of this example. The two lenses seem to render that plane the same, which is well illustrated on a number of internet sites.

 

Relative perspective does not seem to change for a lens depending upon the size of a sensor. By relative perspective I mean how a particular lens compresses foreground and background elements. A 50 mm lens yields a cropped image on a 1.3 cropped sensor, of course, but its fingerprint of compression does not change whether recording the image on a full frame or cropped sensor. If this is true then it is not true to state a 35 mm lens on a 1.3 cropped sensor is the same as a 50 mm lens on a full frame sensor in respect to relative perspective. Please correct me if I am wrong.

 

 

I see your point as I was claiming exactly the same - but now I am a believer:D

 

As a physicist I first thought that the DoF had something to do with the 'resolution' of the lens, but in fact on a film/digital camera it is not governed by diffraction at all unless you go to smaller apertures like f/11 or so. It really is based on what the eye accepts as being sharp under normal observation i.e. it is a convention.

 

Maybe this document (& spreadsheet) is useful, with my apologies for the rough & ready approximations in some of the equations. Comments & suggestions are welcome.

[BerndReini]

Full-frame is better, end of story. I know it can't be done yet, but it will be in the future and anyone who says it doesn't make a difference what the size of the sensor is will be proven wrong.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[SJP]

If this is true then it is not true to state a 35 mm lens on a 1.3 cropped sensor is the same as a 50 mm lens on a full frame sensor in respect to relative perspective. Please correct me if I am wrong.

Scott,

 

I am pleased that the document helped. Re. your question, figure 12 and equation 12 in my document give the method to keep the same perspective when changing from FF to crop, which boils down to fc=Cf=0.75*50=38mm in your specific case. Next we need to keep the same OOF rendering (bokeh) which requires calculating the 'equivalent aperture', see page13-14 and eq.14. You did not specify what 50mm lens to compare to but for sake of argument let's take a 50/2 summicron at full aperture. Then the crop factor yields that you need 38mm and f/1.463 = f/1.4 so a 35mm f/1.4 summilux is close enough (see the lower half of table A1 on p13).

 

Disclaimer: this is based on an 'ideal' lens, bokeh is more subtle than that & depends on the details of the lens construction, no. of blades in the diafraghm, it's location etc.

 

Note that the effective aperture correction factor in the table is independent of what focal length(s) we are comparing (no 'f' in eq.14), so this is a general rule for all equivalent perspective lens comparisons. Also note that it is not always a factor of 1.33, so the 'form follows format' prescription is not 100% accurate.

 

Stephen

[Raf DeBoeck]

... trying to grab the concept here, haven't quite achieved it yet. I actually think there is no concept at all with respect to this discussion.

 

I use a camera and I have a number of lenses to choose from to use with the camera, I can subsequently change from one lens to the other if and when I find this appropriate. Where's the problem?

[lct]

Full-frame is better, end of story. I know it can't be done yet, but it will be in the future and anyone who says it doesn't make a difference what the size of the sensor is will be proven wrong.

At least FF is better for those favouring shallow DoF. The difference is obvious indeed.

[Guest V64]

Agree on all points, Andy. DoF is in the head and occurs only when we look at the print. Thanks for stressing that.

 

Period.

 

Absolutely COF is all about viewing prints

 

I find that there is a natural 'viewing distance (or viewing angle)' when looking at a print.

 

Try looking at various sized prints (preferably of the same proportions) holding each print at the distance which allows you to view the whole image comfortably.

 

I think you will find that there is a natural (comfortable) viewing angle, and that each image (regardless of size) is held so that the images produced on your retina are all much the same, and are at the size where your eye can cab comfortably 'see' the whole image. (i.e. you dont have to scan the image to see it all)

 

These observations clearly do not apply to viewing 1:1 on monitors or looking at prints through loupes. They do apply to viewing transparencies on screens, large prints in exhibitions, advertising hoardings etc.

[SJP]

Advertisement (gone after registration)

I use a camera and I have a number of lenses to choose from to use with the camera, I can subsequently change from one lens to the other if and when I find this appropriate. Where's the problem?

As far as I am aware there is no problem except that if for some reason you want to figure out how a lens on the M8 compares to a different lens on FF or film then you run into some not so intuitive issues.

 

That's all.

 

Just use what you have and be happy - that is what I am doing.

[Raf DeBoeck]

As far as I am aware there is no problem except that if for some reason you want to figure out how a lens on the M8 compares to a different lens on FF or film then you run into some not so intuitive issues.

 

That's all.

 

Just use what you have and be happy - that is what I am doing.

 

I will take you word for it.... I realize that on my M6 my 35mm will represent differently than on my M8 (which I don't have yet). But both will be based on different standards won't they, known standards so you might use a different focal length on the M6 than on the M8 for several reasons, or not... or am I totally beside the point here?

[bybrett]

Full-frame is better, end of story. I know it can't be done yet, but it will be in the future and anyone who says it doesn't make a difference what the size of the sensor is will be proven wrong.

 

Full-frame is in fact double-frame, the M8 sensor is almost identical to the original motion picture size of 18x24 Anyone shooting M8 movies?

[SJP]

I think the "suck its and see" method is fine - there is no reason to start agonising about all of this stuff. I think the perspective conversion factor is the prime result (35mm on M8 is about the same as a 50mm lens on film M & FF). You will find that out soon enough yourself as it is easy to see.

 

The DOF stuff should not really be of major concern. The M8 is close enough to FF for the time being (at least for me). But what this sort of calculation does show is that there is no way that a P&S or mobile phone will get anywhere near the quality of an M8 in terms of 'isolation' and 'bokeh', it is not allowed by physics.

[Scott Root]

Stephen,

 

This is not quite answering my question. To put it another way...

 

The moon taken with my M8 and 35mm lens looks smaller than the moon in pictures I took with my M6 and 50mm lens even though the buildings in this comparison look proportionately the same. Unless I am missing something, the reason for this is the principle of relative perspective. I must use my 50mm lens on my M8 to get the same magnification of the moon I got with my M6 and 50 mm lens, but using the 50mm lens on the M8 will also render the building larger than they would appear if the picture was taken with the M6 and the 50mm lens, of course. I am suggesting that the relative perspective of a 35 mm lens on a M8 is different than a 50mm lens on an M6 and there is no way to mimic this perspective with any lens on an M8. Please correct me if I am imagining something.

 

 

 

Scott

 

I am pleased that the document helped. Re. your question, figure 12 and equation 12 in my document give the method to keep the same perspective when changing from FF to crop, which boils down to fc=Cf=0.75*50=38mm in your specific case. Next we need to keep the same OOF rendering (bokeh) which requires calculating the 'equivalent aperture', see page13-14 and eq.14. You did not specify what 50mm lens to compare to but for sake of argument let's take a 50/2 summicron at full aperture. Then the crop factor yields that you need 38mm and f/1.463 = f/1.4 so a 35mm f/1.4 summilux is close enough (see the lower half of table A1 on p13).

 

Disclaimer: this is based on an 'ideal' lens, bokeh is more subtle than that & depends on the details of the lens construction, no. of blades in the diafraghm, it's location etc.

 

Note that the effective aperture correction factor in the table is independent of what focal length(s) we are comparing (no 'f' in eq.14), so this is a general rule for all equivalent perspective lens comparisons. Also note that it is not always a factor of 1.33, so the 'form follows format' prescription is not 100% accurate.

 

Stephen

[SJP]

Scott,

 

Next attempt

 

Is this based on an actual comparison? The 35mm on M8 is slightly wider than a 50mm on film. The equivalent focal length would be 38mm so a 35mm lens on the M8 is about 10% wider. Would that explain what you are saying?

I must use my 50mm lens on my M8 to get the same magnification of the moon I got with my M6 and 50 mm lens, but using the 50mm lens on the M8 will also render the building larger than they would appear if the picture was taken with the M6 and the 50mm lens, of course.

The image of the moon on the film and sensor plane is the same size if you use a 50mm on both cameras, but to get the same print size the M8 image needs more enlarging.

I think it is even fair to say that the relative perspective is independent of which lens and/or sensor you are using. It actually only depends on your "location in space". Clearly a wide angle lens with a small focal length will show you more of the surroundings than when using a tele-lens. But the parts of the image that coincide are actually identical in terms of perspective.

 

Hope this helps, I am sure we will get there somehow.

[Roberto Watson]

After the cold shower there is nothing new but a nice explanation to the fact that CoC are values given by the photographic industry to fix a standard resolution for diferent photographic formats, although a nice reading. Unfortunately my degree is as photographer no as math or physic.

[nugat]

This thread is based in 90% on misinformation. One of the few posters that tries to spread knowledge is ho-co. The whole mischung of crops, focal lengths , DOF, FOV comes down to one fallacy that is propagated over and over and over...

" 50mm focal length on full frame 36x24mm film is the closest natural/normal approximation of human eye vision". Well, it's not. Same as 24 frames per second (fps) for film (moving pictures, my world) is not the best frequency of a series of still frames to replicate the "real world movement". We grew to accept the wheels of a forward moving vehicle TURNING BACK on film (moving picture) as normal. It took 100 years to get used to. 24fps was simply cheaper than the more faithful 60 or 120 or 180 fps.

Fast 50mm lens on 36x24mm film camera was simply cheaper to manufacture when 25 ISO film were the norm.

The angle of human vision is not even close to that of a 50mm lens on FF. The latter is equal to 46 degrees. Human vision is 150 degrees. Just spread your hands until you stop seeing them.

For those that claim that 50mm is equivalent to "sharp human vision angle"--well , not really. Human eye sees sharply in 20 degrees, or 120mm focal length on FF.

Human eye's focal length is 22 mm actually, corrected for the fluids factor. Without that it is 17mm. The pupil goes from 2 to 8mm, which corresponds to f/2.75-11. (...and the "megasensel" count is 6 million)

Also the argument that 50mm has "the same perspective" as the human eye is a fallacy. A ho_co wrote perspective is a geometrical characterstic, does not depend on optics.

So what lens is "natural"? Use your eye to find out. Cut out in A4 sheet of paper 10x15cm (4x6 inches). At close 10 cm (4 inches) it corresponds to a 24mm lens (122 degrees). Extend your hands to 70cm (over two feet) and you get a 170mm lens. 50mm equivalent you get at 21 cm (9 inches). Comfortable? Bet you NO.

The only reason 50mm on FF is called NORMAL is the 80 year habit.

Take from the above anything you want.

Just don't claim what focal lenghts are "good", "natural" or "equivalent" to whatever standard you pick.

[lct]

..." 50mm focal length on full frame 36x24mm film is the closest natural/normal approximation of human eye vision". Well, it's not...

It's not for you perhaps but it is for me. When i use a 40 i have the feeling that is a long 35, and when i use a 60 it's like a short 75. One of the reasons why i don't like the M8 BTW. Makes 50mm lenses either to short or too long for me. YMMV.

[nugat]

I see your point as I was claiming exactly the same - but now I am a believer:D

 

As a physicist I first thought that the DoF had something to do with the 'resolution' of the lens, but in fact on a film/digital camera it is not governed by diffraction at all unless you go to smaller apertures like f/11 or so. It really is based on what the eye accepts as being sharp under normal observation i.e. it is a convention.

 

Maybe this document (& spreadsheet) is useful, with my apologies for the rough & ready approximations in some of the equations. Comments & suggestions are welcome.

 

Can you explain the little miracle described on page 10 of "this document" linked above? In my part of the world when I change focal length WITHOUT MOVING PLACE the subjects grow as I zoom in. Here the scientist managed to keep the front bear the same size both at 28mm and 90mm? He claims that "lens change perspective".

Did he move for the second picture in or is he a creationist too? But God loves geometry, he got A+square for that subject. For his sake:

Perspective is the function of the distance between the observer and the observed ! Not of the lens. High school geometry.

If it depended on the focal length what would "the crop factor" mean actually???

 

The "document" on page 10 says:

 

"Fig.11. The difference in perspective of a wide-angle and a long focal lens. The two bears are the same

size, left 28 mm, right 90 mm lens, on Leica M8 (0.75 crop factor).

*"

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[BerndReini]

Well Nugat,

 

you are wrong yourself. I am not trying to be condescending and lecturing, but this may actually be some interesting information. The reason the 50mm on 35 full frame has been considered a normal lens for decades is not that it has the same viewing angle as the human eye, but that it most closely corresponds to the way the human eye sees as far as perspective. Look at the extremes of focal lengths: a 200mm flattens an image, a 12mm exaggerates perspective. this has to do with the relationship of the size of objects close to the lens to objects further away from the lens. And as far as this relationship is concerned, the 50mm on full-frame most closely corresponds to the way the human eye sees. There you have it. It turns out that it has not been chosen randomly, and generations of photographers, physicists, and biologists were not idiots.

 

Let me go one step further. As far as viewing angle is concerned, the 35mm most closely corresponds to what the eye sees without the analytical help of the brain. Yes, your peripheral vision includes much more than that, but what is optically perceived in focus, without moving your eye or your head closely corresponds to the viewing angle of a 35mm lens. this is why this lens is sometimes considered as a normal lens in 35mm photography.

 

Look at the two pictures of the bears: the bear in front is indeed the same size in both pictures, but the bear in the back is not. This is what different perspective means: the size relationship between foreground and background HAS ABSOLUTELY CHANGED!!! The pictures look different, feel different, ARE COMPLETELY DIFFERENT! And this response is not directly aimed at nugat, just trying to explain the topic using the pictures he posted.

[Raf DeBoeck]

wow.... don't understand the arithmetic but thats quite an example. It actually does change "the image"

[Raf DeBoeck]

wow.... don't understand the arithmetic but thats quite an example. It actually does change "the image"

[Advertisement]

Hallo,

Look here: Leica M

[nugat]

Well Nugat,

 

you are wrong yourself. I am not trying to be condescending and lecturing, but this may actually be some interesting information. The reason the 50mm on 35 full frame has been considered a normal lens for decades is not that it has the same viewing angle as the human eye, but that it most closely corresponds to the way the human eye sees as far as perspective. Look at the extremes of focal lengths: a 200mm flattens an image, a 12mm exaggerates perspective. this has to do with the relationship of the size of objects close to the lens to objects further away from the lens. And as far as this relationship is concerned, the 50mm on full-frame most closely corresponds to the way the human eye sees. There you have it. It turns out that it has not been chosen randomly, and generations of photographers, physicists, and biologists were not idiots.

 

Let me go one step further. As far as viewing angle is concerned, the 35mm most closely corresponds to what the eye sees without the analytical help of the brain. Yes, your peripheral vision includes much more than that, but what is optically perceived in focus, without moving your eye or your head closely corresponds to the viewing angle of a 35mm lens. this is why this lens is sometimes considered as a normal lens in 35mm photography.

 

Look at the two pictures of the bears: the bear in front is indeed the same size in both pictures, but the bear in the back is not. This is what different perspective means: the size relationship between foreground and background HAS ABSOLUTELY CHANGED!!! The pictures look different, feel different, ARE COMPLETELY DIFFERENT! And this response is not directly aimed at nugat, just trying to explain the topic using the pictures he posted.

 

I think I'll rest my case.

Last try.

The scientist in the above two pictures MOVED--that's why the perspective changed.

Human eye does not see "like the 50mm lens" as far as perspective is concerned.

Perspective is not the property of a lens. Any lens. Of course all lenses have distortions, the wider the lens the bigger distortion.

Perspetive is a geometrical function, not optical. It does not depend on the focal length. It depends solely on the distance between the observer and the observed.

Eye and brain are a couple of conmen. They cheat all the time.

The brain tries to keep the horizon always where it belongs. Stand in front of a mirror. Look at your left eye, then at right. Your eyeballs moved, but you don't notice it. Look at any well edited film. The "cuts" replicate the way brain edits what we see.

I don't know what else I can bring up here .. Can somebody with better powers of explanation come to rescue?