Doug Kerr
Well-known member
Today I will discuss an issue that is often misunderstood by "newcomers" to "advanced photography". Experienced photographers are aware of it, but may not have a clear idea of the principles behind it, and thus may not be able to clearly articulate it to others.
Suppose I have a very nice lens with a focal length of 100mm. For a certain task, I say I wish I had a 200 mm lens. What would the advantage be?
Well a civilian might say, "Well, that way you can take a picture of something farther away."
Well, of course, that's not it. My 35 mm lens will let me take a picture of the moon, about 250,000 miles away.
We "advanced photographers" recognize that in fact the purpose is really to have the desired subject region fill more of the frame. But why is that advantageous?
After all, if we take a picture of a bear with a 200 mm lens, and want only the bear (and a certain amount of surround) in the "delivered" image (perhaps a print), we crop to suit. If we took the shot from the same location with a 100 mm lens, we can deliver an image with the same "framing". We just "crop more tightly". So why buy a longer focal length lens?
But, if in fact, the resolution of the entire camera (lens plus sensor and image processing system), at the focal plane, is the same in either case, then the image with the 200 mm lens will be better resolved ("sharper").
If the resolution is 100 line pairs per millimeter, and the bear's length, in the 100 mm shot, occupies 5 mm on the sensor, and in the 200 mm shot, occupies 10 mm, then:
• In the 100 mm shot, the bear (longitudinally) has the benefit of 500 lines of resolution
•In the 200 mm shot, the bear has the benefit of 1000 lines of resolution
Thus clearly, especially if we crop the two frames so the "delivered print" shows the bear the same size, the 200 mm shot will have a substantial advantage from a resolution standpoint.
But now let's consider the case in which the camera with a different 200 mm lens only affords a focal plane resolution of 50 line pairs per millimeters. Then a corresponding "analysis" will show that there is no advantage, regarding resolution of the final "bear portrait", to that lens over our 100 mm lens.
A colleague here (I wish I could remember who, but I've slept since then) speaks of the concept of the "reach" of a telephoto lens, based on this outlook. What that turns out to mean, in explicit mathematical terms, is the "angular resolution" of the lens. In the sense where a smaller number means better resolution (as when we describe the spacing of the resolved line pairs), it is the angular separation between two points that can be resolved (for the lens of interest on a specific camera).
We are not really that familiar with thinking of object point spacing in terms of angle. But if we only plan to use this parameter to compare different lenses as applied to a certain camera, and wish to work with a definition of resolution in which a larger number is "finer" (as when we cite line pairs per mm, or line pairs per picture height), the the "reach factor" can be thought of as the product of the focal length times the focal plane resolution in line pairs per mm (for the lens on a particular camera). (This gives a large number numerically, but so what.)
In the first example, the 100 mm lens, on our camera, has a "reach factor" of 10,000 (100 mm x 100 lp/mm); the "good" 200 mm lens, on our camera, has a "reach factor" of 20,000 (200 mm x 100 lp/mm).
In the second example, the "alternate" 200 mm lens, on our camera, has a "reach factor" of 10,000 (200 mm x 50 lp/mm).
And so, there is no advantage in using the alternate 200 mm lens, over the 100 mm lens, from a standpoint of final resolution of the delivered bear. (There could be other considerations of focal length alone.)
It is fascinating to take measured resolutions of zoom lenses, over their focal length range, and plotting the "reach factor".
This recently came up where when we contemplated superseding our Sigma 18-200mm lens with a lens having a minimum focal length of about 18 mm but a maximum focal length of 250 or 270 mm. The question is, "will we benefit from the available greater focal length?"
The answer was (from practical testing, not based on such a "plot"), for every candidate we were able to test, "no". In some cases, the reach at the maximum focal length setting was less than at, say, 200 mm.
Note that the impact of this on the choice between two lenses depends on the contribution of the camera to the joint resolution, and thus on the camera to be used.
For example, in a camera with a large pixel pitch, all of the lenses we posited may lead to nearly the same focal plane resolution (basically dictated by the sensor). Then, the longer focal length would always be advantageous from the standpoint of the resolution of the delivered bear.
But today, with sensor system resolution steadily increasing, the contribution of lens resolution to overall resolution is more significant.
So if you really want to get that bear in "HD", be prepared to carry something pretty heavy. But you will have help: your wallet will be considerably lighter.
On the tactical front, we sent out Sigma 18-250 and our Tamron 18-270 back! We gave big hugs to our:
• Sigma 18-200 mm f/3.5-6.3 DC OS
• Canon 24-105 mm f/4.0L IS USM
• Canon 70-200 mm f/2.8 IS USM
and put them back in the bag.
But I may not be climbing the steps to the top row of Cowboys Stadium ever again!
Suppose I have a very nice lens with a focal length of 100mm. For a certain task, I say I wish I had a 200 mm lens. What would the advantage be?
Well a civilian might say, "Well, that way you can take a picture of something farther away."
Well, of course, that's not it. My 35 mm lens will let me take a picture of the moon, about 250,000 miles away.
We "advanced photographers" recognize that in fact the purpose is really to have the desired subject region fill more of the frame. But why is that advantageous?
After all, if we take a picture of a bear with a 200 mm lens, and want only the bear (and a certain amount of surround) in the "delivered" image (perhaps a print), we crop to suit. If we took the shot from the same location with a 100 mm lens, we can deliver an image with the same "framing". We just "crop more tightly". So why buy a longer focal length lens?
But, if in fact, the resolution of the entire camera (lens plus sensor and image processing system), at the focal plane, is the same in either case, then the image with the 200 mm lens will be better resolved ("sharper").
If the resolution is 100 line pairs per millimeter, and the bear's length, in the 100 mm shot, occupies 5 mm on the sensor, and in the 200 mm shot, occupies 10 mm, then:
• In the 100 mm shot, the bear (longitudinally) has the benefit of 500 lines of resolution
•In the 200 mm shot, the bear has the benefit of 1000 lines of resolution
Thus clearly, especially if we crop the two frames so the "delivered print" shows the bear the same size, the 200 mm shot will have a substantial advantage from a resolution standpoint.
But now let's consider the case in which the camera with a different 200 mm lens only affords a focal plane resolution of 50 line pairs per millimeters. Then a corresponding "analysis" will show that there is no advantage, regarding resolution of the final "bear portrait", to that lens over our 100 mm lens.
A colleague here (I wish I could remember who, but I've slept since then) speaks of the concept of the "reach" of a telephoto lens, based on this outlook. What that turns out to mean, in explicit mathematical terms, is the "angular resolution" of the lens. In the sense where a smaller number means better resolution (as when we describe the spacing of the resolved line pairs), it is the angular separation between two points that can be resolved (for the lens of interest on a specific camera).
We are not really that familiar with thinking of object point spacing in terms of angle. But if we only plan to use this parameter to compare different lenses as applied to a certain camera, and wish to work with a definition of resolution in which a larger number is "finer" (as when we cite line pairs per mm, or line pairs per picture height), the the "reach factor" can be thought of as the product of the focal length times the focal plane resolution in line pairs per mm (for the lens on a particular camera). (This gives a large number numerically, but so what.)
In the first example, the 100 mm lens, on our camera, has a "reach factor" of 10,000 (100 mm x 100 lp/mm); the "good" 200 mm lens, on our camera, has a "reach factor" of 20,000 (200 mm x 100 lp/mm).
In the second example, the "alternate" 200 mm lens, on our camera, has a "reach factor" of 10,000 (200 mm x 50 lp/mm).
And so, there is no advantage in using the alternate 200 mm lens, over the 100 mm lens, from a standpoint of final resolution of the delivered bear. (There could be other considerations of focal length alone.)
It is fascinating to take measured resolutions of zoom lenses, over their focal length range, and plotting the "reach factor".
This recently came up where when we contemplated superseding our Sigma 18-200mm lens with a lens having a minimum focal length of about 18 mm but a maximum focal length of 250 or 270 mm. The question is, "will we benefit from the available greater focal length?"
The answer was (from practical testing, not based on such a "plot"), for every candidate we were able to test, "no". In some cases, the reach at the maximum focal length setting was less than at, say, 200 mm.
Note that the impact of this on the choice between two lenses depends on the contribution of the camera to the joint resolution, and thus on the camera to be used.
For example, in a camera with a large pixel pitch, all of the lenses we posited may lead to nearly the same focal plane resolution (basically dictated by the sensor). Then, the longer focal length would always be advantageous from the standpoint of the resolution of the delivered bear.
But today, with sensor system resolution steadily increasing, the contribution of lens resolution to overall resolution is more significant.
So if you really want to get that bear in "HD", be prepared to carry something pretty heavy. But you will have help: your wallet will be considerably lighter.
On the tactical front, we sent out Sigma 18-250 and our Tamron 18-270 back! We gave big hugs to our:
• Sigma 18-200 mm f/3.5-6.3 DC OS
• Canon 24-105 mm f/4.0L IS USM
• Canon 70-200 mm f/2.8 IS USM
and put them back in the bag.
But I may not be climbing the steps to the top row of Cowboys Stadium ever again!