What defines camera format? [Archive] - Canon Digital Photography Forums

rdenney

7th of April 2005 (Thu), 09:11

What defines the “format” of a camera system? Film/sensor size? Size of the image circle created by the system on the image plane? What?

I realize, historically, the film size was matched to the image circle. But what if the image circle doesn't match the film/sensor size?

Like the 10D. Is it a 35mm DSLR or an APS-C DSLR? And do EF lenses on a 10D have the properties of 35mm-format lenses or APS-C system lenses?

Frame size. That's all.

You mentioned elsewhere that putting an APS-sized sensor in an 8x10 camera would not make it an APS camera. But, in fact, it would. The lenses would be really long, of course.

I do it all the time, actually. I have a 6x9 rollfilm back for my 4x5 view camera. When I put that rollfilm back on my view camera, I've changed it to a 6x9 view camera. I still use the same lenses, but they mean different things. The 90mm/5.6 Super Angulon that is such a stunning wide-angle lens in 4x5 is merely a normal lens when used with the 6x9 back. Sure, it lights up a lot of the inside of the camera that I'm not using, but that's true for all view-camera lenses in order to allow movements.

People used to put 35mm adaptors in their 6x6 Rolleiflexes all the time (i.e. the Rolleikin). Is it then a 35mm camera? Yes. The 80mm lens on the Rolleiflex is now a short tele rather than a normal lens, because we've changed the format.

We will also enlarge the images more. I rarely made prints from 4x5 more than a 4x enlargement, and the usual degree of enlargement was more like 3x. But the 11x14 print that requires a 3x enlargement from 4x5 requires a 5x enlargement from 6x9. Thus, any artifact in the image gets enlarged nearly twice as much. That 11x14 print requires a 19x enlargement from an APS-sized sensor. That's why format is king.

People use lenses intended for 8x10 cameras on 4x5 cameras all the time. A 165mm Super Angulon will cover 8x10 easily, and it is merely a normal lens on 4x5. But that extra large illuminated circle means you can tilt that sucker just about on edge and still light the 4x5 frame. But it's still a 4x5 camera it would be mounted on (even if the bellows, lens board, film back, ground glass, and everything else except the film holder also worked for 8x10). In fact, many view camera systems can exchange many items between the various formats, just to facilitate the use of lenses and other parts across different formats. A "4x5 reducing back" is a standard item for 8x10 view cameras.

Rick "thinking a little view camera experience goes a long way to sorting out format confusion" Denney

Steven M. Anthony

7th of April 2005 (Thu), 09:24

rdenney

7th of April 2005 (Thu), 10:01

So what does it mean that a camera is a "35mm camera" if it's format designation is dependent on the size of the recording medium?

I get that you can rig/adapt a body to accept different film sizes, but as you point out, when you do this, the lenses "mean different things." But while they MEAN different things, they remain the lenses they were before you adapted the size of the recording medium. And they behave the way they did before you adapted the size of the recording medium. So what has changed is your interpretation of how the lenses work--not how the lenses actually work. Right?

I'm curious--does the 4X5 reducing back of an 8X10 camera utilize the same film plane as the 8X10 does? Or does it move the film plane forward in the camera body?

Yes, the 4x5 reducing back uses the same film plane as the 8x10 version. Of course, view camera lenses are symmetrical and mount at the nodal points (with the exception of a few telephoto designs), so the notion of "back focus distance" doesn't mean much--every lens has its own distance from the lens board to the film plane. That's why the cameras use bellows instead of a hard box. But if I focus a 165mm lens on 8x10 ground glass, and then put a reducing back on it with 4x5 film, it will still be focused. In fact, when I use my 6x9 rollfilm back on my 4x5 camera, I focus on the 4x5 ground glass first.

The only difference is that the smaller film frame doesn't include as much of the scene. And that's the only reason the lenses mean different things. If I focused that 165mm lens on 8x10, and made an exposure on 8x10 film and then on 4x5 film, I could take that 8x10 negative and trim off all but 4x5, and have an identical negative to the one shot on 4x5 film. When I did so, that wide-angle lens (in 8x10) would now have the effect of a norma lens (in 4x5).

The lenses therefore produce the same image, but you are using different portions of the image they produce. So, it's not my interpretation that changes, but something more important at play. And that is:

We define lenses as short of long based on their relationship to "normal". A normal lens is defined as the diagonal of the frame. Thus, whether a lens is short or long depends utterly on the format in which they are used. They focus the same, project the same, and illuminate the same. But with a smaller format we take a little out of the middle and enlarge it more to make a given print.

Back in the old days, camera makers used to take their widest rectilinear lens and make a picture of a scene. Then, they would draw rectangles in the scene to show the field of view of longer lenses. I used to think that was wrong, but then I learned that camera position alone determines perspective, and that focal length and format determine how much of the scene we'd actually see in the image. So, if we start with that wide-angle image of the New England town, and then see that the 600mm rectangle only includes the distant church steeple, we can see that we could arrive at that image using one of two methods: We could use a 600mm lens, or we could take that tiny rectangle out of the original image and blow it up. That's why changing format has the same effect as changing focal length, and that's the reason for the 1.6 equivalency factor between a 10D and a 35mm camera.

A 10D is not a 35mm camera. It is an APS camera to which you can attach lenses intended for 35mm camera. Let's apply a little reductio ad aburdum in the other direction. Let's say we had a digital point-n-shoot with a sensor 2mm by 3mm (I'm not sure such a thing exists, but let's imagine it does). We could mount that sensor in lieu of the APS sensor on a 10D and take pictures all day long. The lenses would focus the same way, and they would project the same image. But they would only provide you with a 2x3mm chunk of that project image. Since a "normal" lens for a 2x3mm sensor would be 3.6 mm, the 50mm lens you might put on that camera would be a 14-times-normal telephoto. That would like putting a 600mm lens on a 35mm camera. Thus, the 50mm lens on that 2x3 sensor would produce an identical image to a 600mm lens on a 35mm camera. You'd get the steeple both ways.

But the reason we use the long lens on a 35mm camera to get that distant church steeple is so that we don't have to enlarge the image more than necessary. I could get that steeple, and it would look just the same, if I took a 2x3 chunk of the 35mm frame and enlarged it. The difference is that I would also be enlarging all my lens faults and also my film grain (or sensor resolution). That's the only reason people bought long lenses--there was a limit to how much they could enlarge the negative before the junk that they are also enlarging causes problems.

So, the lens isn't any different. It focuses the same image on the same film plane. But we use a different portion of it, and the meaning of the lens, vis a vis wide-angle versus telephoto, has only do with the portion of the scene we use.

You asked why we couldn't put a 4x5 film behind a lens intended for 35mm film and get an image earlier. The reason is simple: The lens barrel is too narrow and it will cut off the corners. But if I took a Canon 50mm lens and put it on a MF camera, I'd get a circle in the middle of the film. Within that circle, the image would be absolutely identical to what my 50mm Zeiss medium-format lens would produce. But the glass and the lens barrel isn't designed to light up that much real estate, and it would cut off the corners. That' s why medium-format lenses are so much bigger and more expensive for what you get.

There was always the question of whether it's better to use a teleconverter or enlarge more. When a lens is effectively longer (with respect to normal) because we are using it on a 10D, we are adopting the "enlarge more" strategy. Using a 1Ds with a 1.6 teleconverter (assuming such a thing existed) would be the teleconverter strategy.

Hint: The teleconverter has to be pretty crappy not to be better than enlarging more, heh, heh.

Rick "format is king" Denney

PhotosGuy

7th of April 2005 (Thu), 10:04

does the 4X5 reducing back of an 8X10 camera utilize the same film plane as the 8X10 does? Or does it move the film plane forward in the camera body? The lens will still focus on the same plane as before But while they MEAN different things, they remain the lenses they were before you adapted the size of the recording medium. And they behave the way they did before you adapted the size of the recording medium. Yes. If you put a 35mm lens on a view camera
1. You won't be able to focus unless you mount it on a recessed lens board &
2. you'll get a stinkin' little 1-1/2" diameter pic on the 4X5" film. The angle of coverage is smaller 'cause 35mm cams don't need a bigger pic, except on PC lenses of course.

Steven & rdenney: Thank's a lot for saving me a lot of 2-fingered typing! ;)
Edit: we cross-posted. Thank's again!

Steven M. Anthony

7th of April 2005 (Thu), 11:05

That's why the cameras use bellows instead of a hard box.

Yes--that makes sense. But also brings up an interesting point. 35mm cameras--and most other modern cameras--do not have bellows. They have a hard box. And in most cases, that box is designed to work with a particular size recording medium (acknowledging one can switch backs on some...). The format for these cameras are generally referred to by the size of the recording medium (35mm, 4X5, etc.). And the image circle created on the film plane is optimized for that recording medium. So, a 35mm camera--and lenses, are designed to project an optimized image circle to fill a 35mm film frame.

In my mind, that's the essense of a 35mm camera.

Now, changing the size of the recording medium doesn't change that essence. Nor does using a football to shoot hoops with change the essence of the football--it doesn't BECOME a basketball even though you are shooting baskets with it; you are just shooting hoops with a football.

rdenney

7th of April 2005 (Thu), 11:40

Yes--that makes sense. But also brings up an interesting point. 35mm cameras--and most other modern cameras--do not have bellows. They have a hard box. And in most cases, that box is designed to work with a particular size recording medium (acknowledging one can switch backs on some...). The format for these cameras are generally referred to by the size of the recording medium (35mm, 4X5, etc.). And the image circle created on the film plane is optimized for that recording medium. So, a 35mm camera--and lenses, are designed to project an optimized image circle to fill a 35mm film frame.

In my mind, that's the essense of a 35mm camera.

Now, changing the size of the recording medium doesn't change that essence. Nor does using a football to shoot hoops with change the essence of the football--it doesn't BECOME a basketball even though you are shooting baskets with it; you are just shooting hoops with a football.

You really should think of those boxes as a continuum. For example, the distance from the film plane to the lens mount is not at all constant across 35mm cameras. It ranges from as little as about 28mm for some rangefinders to as much as around 50mm for some SLR's. That's quite a range! The effect it has on the lens is how much barrel is included, and that's about all. One of the real advantages of the EF mount is that is has a fairly short back-focus distance (though, sadly, not as short as the old FD mount). Thus, lenses from cameras with longer backfocus distances are easy to adapt.

Medium-format cameras with interchangeable lenses have backfocus distances ranging from nearly as short as the bigger 35mm SLR's to nearly 70mm.

There have been (and still are) many cameras on the market with bellows focusing, so that the lens mount didn't have to have a focusing mechanism built into it. The Mamiya C-series TLR's and the RB and RZ 67's use bellows focusing, as do several Fuji models, nearly every large-format camera ever made, and even a few 35mm cameras with folding lenses.

A camera body with a sensor of a particular size or a film frame of a particular size has a particular format, no matter what else the sheet metal surrounding it might have been used for. It's quite reasonable for an APS camera to have the same distance from film plane to lens mount as a 35mm camera--this allows lenses to be used across formats, as we have with the 10D and 35mm Canons. APS and 35mm are closer to each other than are, say, 35mm and medium format, so it shouldn't be a surprise that the same backfocus distance would be usable by both. But it does put a strain on the wide-angle end, where the lenses require a more pronounced retrofocus design to get the glass out in front of the mount as needed.

I know several people who have taken old telephoto lenses intended for 35mm, cut off part of the barrel, and mounted them on medium-format cameras. One such that is commercially available is the Hartblei 500mm/5.6 mirror lens made in the Ukraine, and available in various medium-format mounts. It was originally a 35mm lens with a longer barrel.

There are two reasons why small cameras don't have bellows focusing, especially these days: 1.) Bellows are tender and these cameras are subjected to abuse, and 2.) modern lenses have internal focusing and moveable elements, not to mention auto focusing, that require the focusing to take place in a complex lens barrel rather than simply at the lens mount.

Rick "seeing a continuum with considerable overlap" Denney

PacAce

7th of April 2005 (Thu), 11:53

robertwgross

7th of April 2005 (Thu), 13:03

This is always what struck me as odd about 35mm. Neither dimension is 35mm. The frame dimensions are 24mm by 36mm. Only the overall width of the film (with sprocket holes) is 35mm.

---Bob Gross---

rdenney

7th of April 2005 (Thu), 14:20

This is always what struck me as odd about 35mm. Neither dimension is 35mm. The frame dimensions are 24mm by 36mm. Only the overall width of the film (with sprocket holes) is 35mm.

This stems from its origins at movie film. The point of the original Leica was to make use of cine film that was measured across the gross width of the film.

Of course, rollfilm film designations have always been wacky. APS, 35mm, 120, 220, 620, etc.? Wacky stuff.

Rick "thinking it must have seemed like a good idea at the time" Denney

robertwgross

7th of April 2005 (Thu), 14:35

I know all too much about gross widths.

---Bob Gross---

Steven M. Anthony

7th of April 2005 (Thu), 15:05

All 35mm cameras, whether they are range finders, P&S, SLR, cinematography camera, or what not, have one thing in common and that's the film(sensor) size they are using. You'll notice that the lenses are all different, depending on the camera and the application but the size of the film/sensor are all the same. All formats are specified by film sizes, not by the type of body nor the lenses they use. Period.

Well, sure--they were made to work with that sized film.

But try this…

I have a 50mm lens for my AE-1. It projects a fixed-sized image circle on the film plane of a 35mm-format SLR such that the 35mm film frame fits inside the circle. As you point out, indirectly, this is a no-brainer: the camera/lens system was built to work with 35mm film.

When I put that 50mm lens on my 10D, it projects the same sized image circle on the sensor plane of that camera.

But the 10D’s sensor does not record the whole of the image the AE-1 records on film. The 10D records what we call a cropped image—a subset of the whole image captured by the AE-1. In fact, the 10D captures the field of view that the AE-1 would capture if it had an 80mm lens attached to it. But that’s where the equivalence between the two stops.

Note: on the AE-1, the film frame recorded by the 80mm lens, which produces an equivalent field of view as recorded by the 10D’s sensor with a 50mm lens attached, would be bigger than the 10D’s sensor—i.e., it would fill the whole 35mm frame. As a result, the perspective in the shot made with the AE-1/80mm combo would be different from that rendered through the 50mm on the 10D.

My point? A 50mm lens on a 10D is not the same as an 80mm lens on a full-frame camera.

The implications? 35mm-format lenses act like 35mm-format lenses no matter how large or small the sensor is on the image plane. In other words, the optical properties and behavior of a 35mm camera/lens combo is independent of sensor/film size.

If, by some feat of magic, placing an APS-C-sized sensor behind my 50mm (35mm-format) lens caused the element of the lens to reconfigure into an 80mm lens, and then manage to generate a smaller image circle to “fit” the 10D's sensor, THEN I would agree that a 50mm on my 10D was equivalent to an 80mm on my AE-1.

But until that happens, I will continue to believe that the 10D just crops the image a 50mm lens throws onto the image plane of the camera—and that while camera/lens systems are initially designed to work with a given resultant image size, their format (i.e., their optical properties and behavior) does not change with changes in recording medium size.

Just like a football stays a football even if you toss it through a basketball hoop, a 35mm camera system stays a 35mm camera system even with an APS-C sized sensor stuck inside.

PacAce

7th of April 2005 (Thu), 15:16

Well, sure--they were made to work with that sized film.

But try this…

I have a 50mm lens for my AE-1. It projects a fixed-sized image circle on the film plane of a 35mm-format SLR such that the 35mm film frame fits inside the circle. As you point out, indirectly, this is a no-brainer: the camera/lens system was built to work with 35mm film.

When I put that 50mm lens on my 10D, it projects the same sized image circle on the sensor plane of that camera.

But the 10D’s sensor does not record the whole of the image the AE-1 records on film. The 10D records what we call a cropped image—a subset of the whole image captured by the AE-1. In fact, the 10D captures the field of view that the AE-1 would capture if it had an 80mm lens attached to it. But that’s where the equivalence between the two stops.

Note: on the AE-1, the film frame recorded by the 80mm lens, which produces an equivalent field of view as recorded by the 10D’s sensor with a 50mm lens attached, would be bigger than the 10D’s sensor—i.e., it would fill the whole 35mm frame. As a result, the perspective in the shot made with the AE-1/80mm combo would be different from that rendered through the 50mm on the 10D.

My point? A 50mm lens on a 10D is not the same as an 80mm lens on a full-frame camera.

The implications? 35mm-format lenses act like 35mm-format lenses no matter how large or small the sensor is on the image plane. In other words, the optical properties and behavior of a 35mm camera/lens combo is independent of sensor/film size.

If, by some feat of magic, placing an APS-C-sized sensor behind my 50mm (35mm-format) lens caused the element of the lens to reconfigure into an 80mm lens, and then manage to generate a smaller image circle to “fit” the 10D's sensor, THEN I would agree that a 50mm on my 10D was equivalent to an 80mm on my AE-1.

But until that happens, I will continue to believe that the 10D just crops the image a 50mm lens throws onto the image plane of the camera—and that while camera/lens systems are initially designed to work with a given resultant image size, their format (i.e., their optical properties and behavior) does not change with changes in recording medium size.

Just like a football stays a football even if you toss it through a basketball hoop, a 35mm camera system stays a 35mm camera system even with an APS-C sized sensor stuck inside.
From past (and current) experiences, I know better than to try to debate anything with you that you have already made up your mind about. All I will say is that facts are facts and no matter how you try to rationalize your differring views of them, it won't change the facts one iota.

Steven M. Anthony

7th of April 2005 (Thu), 15:54

From past (and current) experiences, I know better than to try to debate anything with you that you have already made up your mind about. All I will say is that facts are facts and no matter how you try to rationalize your differring views of them, it won't change the facts one iota.

I haven't made up my mind. I just haven't seen any evidence strong enough to change my current pov. True, facts are facts. But when the facts presented are unrelated to the issue, they don't convince me of anything. How people use camera bodies and lenses does not define the essence of the format. An optical system should have its essence defined in terms of optical properties and characteristics, no?

If you don't enjoy such intellectual exercises, fine. But adding negativity to the discussion doesn't help.

slin100

7th of April 2005 (Thu), 16:11

Note: on the AE-1, the film frame recorded by the 80mm lens, which produces an equivalent field of view as recorded by the 10D’s sensor with a 50mm lens attached, would be bigger than the 10D’s sensor—i.e., it would fill the whole 35mm frame. As a result, the perspective in the shot made with the AE-1/80mm combo would be different from that rendered through the 50mm on the 10D.

My point? A 50mm lens on a 10D is not the same as an 80mm lens on a full-frame camera.

Leo may have given up, but I'll give it a shot.

Steven, I don't know what your definition of perspective is, but using the standard definition of photographic perspective, it will be the same.

You can easily see for yourself. Take your 28-300 and set it to approximately 31.25mm. The image you see in the 10D viewfinder will look exactly the same as what you will see in the AE-1 w/ the 50mm, modulo viewfinder crop and viewfinder magnification.

robertwgross

7th of April 2005 (Thu), 16:59

The whole concept makes good sense if you think that the distance of the film/sensor behind the lens is not fixed.

---Bob Gross---

DocFrankenstein

7th of April 2005 (Thu), 17:14

Or you can try thinking of the lenses not in mm but in angles.

IE A normal lens covers 40 degree view... regardless of the amount of film that's put behind it.

But a canon 50/1.8 prime will only make a circle 25mm in diameter, while that large format 165 mm super somethigon will make a circle of around 20 inches with a really wide angle of view.

The size of the film you choose to put behind it, is up to you.

Cheers

rdenney

7th of April 2005 (Thu), 17:18

I know all too much about gross widths.

---Bob Gross---

I thought you might enjoy that.

Rick "a little on the gross side himself" Denney

rdenney

7th of April 2005 (Thu), 17:34

My point? A 50mm lens on a 10D is not the same as an 80mm lens on a full-frame camera.

You're right, it's not the same. It's a 50mm lens in front of an APS sensor, instead of an 80mm lens in front of a 35mm frame. And the nodal point of the 80mm lens will be further from the film to get it into focus. Sure.

But it will produce the same image from the same camera position. And it's camera position alone that defines the look of the image (i.e., the perspective). The focal length determines the magnification, and the format determines how much of the scene is visible in the frame at that magnification.

The magnification of the 80 is greater by a large enough margin to offset the larger frame, such that the image produced is the same. That's why it's an equivalency factor, and not a "crop" factor.

But you will have to enlarge the image from the APS sensor 1.6 times more to get a print of the same size.

As I have said before, if you think in terms of focal length's relationship to normal, which is solely, singly, and by definition determined by format, then equivalency factors are lot easier to understand. (Doc's notion that normal is defined as 40 degrees is determined by the format. That's the angle subtended by the frame, not by the image circle, the latter of which is controlled by lens design).

You really should spend a month taking pictures with a view camera. I think your view of the world would really expand.

Rick "noting that all Super Angulons of whatever focal length have the same angle of view with respect to their illuminated circle, but only the 165 is a normal lens for 4x5" Denney

rdenney

7th of April 2005 (Thu), 17:48

But a canon 50/1.8 prime will only make a circle 25mm in diameter, while that large format 165 mm super somethigon will make a circle of around 20 inches with a really wide angle of view.

But my 50mm Zeiss Flektogon will make a circle 70mm in diameter. And my 47mm Super Angulon will make an image circle 120mm in diameter. But they are all normal lenses if you put them in front of a 35mm frame, and within that frame, they all produce exactly the same image.

The nomenclature problem in photography goes back further than small sensors used by 35mm photographers. We have always, and incorrectly, used terms like "wide angle" and "telephoto" lenses to describe lenses that are shorter and longer than normal. That's a mistake, but we all do it, including me.

A wide-angle lens has a wide illuminated field. A retrofocus wide-angle lens has a wide illuminated field couple with a design that puts most of the glass in front of the nodal point. Most view-camera wide-angle lenses are not retrofocus designs, but most SLR wide-angles, that must sit in front of a mirror box, are. Telephotos are lens designs that put most or all of the glass behind the nodal point, and optically they are the opposite of retrofocus wide-angle lenses.

A Super Angulon is a wide-angle lens, no matter what focal length. That's why it's called a Super Angulon, and not a Symmar or Tele-Zenar, even though their focal length ranges overlap considerably.

In between are lenses that are usually designed to be used as normal lenses for their format, and usually have a symmetrical design with about a 45-degree illuminated circle. Typical designs include the old Tessar and the newer Planar designs. The Planar design is a plasmat, which is the general plan used for normal lenses by just about everybody.

But a wide-angle lens of normal focal length will still cover the frame and will provide the same image. I like wide-angle lenses even when used as normals on view cameras because it gives me more lens movements.

Rick "who knows he's just adding to the confusion, heh, heh" Denney

PacAce

7th of April 2005 (Thu), 17:57

I haven't made up my mind. I just haven't seen any evidence strong enough to change my current pov. True, facts are facts. But when the facts presented are unrelated to the issue, they don't convince me of anything. How people use camera bodies and lenses does not define the essence of the format. An optical system should have its essence defined in terms of optical properties and characteristics, no?

If you don't enjoy such intellectual exercises, fine. But adding negativity to the discussion doesn't help.
I wasn't adding negativity anywhere, was I? All I said was that I didn't want to debate the topic with you. I do enjoy intellectual exercises but only to a point. To use your own words from a previous discuss we had, " it's like teaching pigs to sing--it just frustrates you and the pig never learns to sing...". :)

Steven M. Anthony

7th of April 2005 (Thu), 17:58

Leo may have given up, but I'll give it a shot.

Steven, I don't know what your definition of perspective is, but using the standard definition of photographic perspective, it will be the same.

You can easily see for yourself. Take your 28-300 and set it to approximately 31.25mm. The image you see in the 10D viewfinder will look exactly the same as what you will see in the AE-1 w/ the 50mm, modulo viewfinder crop and viewfinder magnification.

By perspective (maybe the wrong word) I mean that in the shot with the 80mm on the AE-1, the image will show foreground-to-background compression that the 50mm/10D combo won't exhibit. It might be slight at 50mm vs. 80mm, but it's there.

Steven M. Anthony

7th of April 2005 (Thu), 18:04

You're right, it's not the same. It's a 50mm lens in front of an APS sensor, instead of an 80mm lens in front of a 35mm frame. And the nodal point of the 80mm lens will be further from the film to get it into focus. Sure.

But it will produce the same image from the same camera position. And it's camera position alone that defines the look of the image (i.e., the perspective). The focal length determines the magnification, and the format determines how much of the scene is visible in the frame at that magnification.

Not true--you must know of the "compression" you get (fore-to-background) as lens fl gets larger...

As I have said before, if you think in terms of focal length's relationship to normal, which is solely, singly, and by definition determined by format, then equivalency factors are lot easier to understand. (Doc's notion that normal is defined as 40 degrees is determined by the format. That's the angle subtended by the frame, not by the image circle, the latter of which is controlled by lens design).

But if you took a 35mm-format 40mm (normal) lens and put it on a 10D, you wouldn't get the same shot as you would with the AE-1--the 10D image would be a "crop" of the full frame. THAT'S why the "factor" is referred to by those in the know as a "crop factor." :)

Belmondo

7th of April 2005 (Thu), 18:09

By perspective (maybe the wrong word) I mean that in the shot with the 80mm on the AE-1, the image will show foreground-to-background compression that the 50mm/10D combo won't exhibit. It might be slight at 50mm vs. 80mm, but it's there.

Yes, but to get the same framing between the AE-1 and the 10D using the same lens (any lens), you will have to stand farther away from the subject with the 10D. Once you do that, the foreground/background compression changes.

Wish I had an AE1 with a 50mm lens. It would be fun to do this experiment. I'm sure someone, somewhere, has done it already.

rdenney

7th of April 2005 (Thu), 19:56

By perspective (maybe the wrong word) I mean that in the shot with the 80mm on the AE-1, the image will show foreground-to-background compression that the 50mm/10D combo won't exhibit. It might be slight at 50mm vs. 80mm, but it's there.

No, it won't. The relationship between foreground and background objects is controlled solely by the camera position. The only way you'll get compression is by moving back and using a longer lens or cropping the image more, but moving back is the required first step.

Remember what I said above about how camera companies in the old days would take a single wide picture and draw rectangles in it to show the effect of longer lenses. All the rectangles are on the same picture--the perspective is the same. Ansel Adams would for the correct perspective without even setting up the camera. When you shoot with a large-format view camera, you don't walk around with it seeing how it looks in the viewfinder! You look with your eyes, and when you see the relationships you want (including the foreground and background perspective), then you set up the camera, and choose the lens that covers the scene you want on that format.

If you have an AE-1, go buy a roll of film and try the experiment. It will enlighten you.

A 4x5 image taken with a 160mm lens will be identical to an AE-1 image taken with a 45mm lens (normal is really 43 on a 35mm camera), and that will be identical to a 10D image taken with a 28mm lens. You won't be able to tell the difference, exept the 4x5 image will be sharper and look better, heh, heh.

And if you put the 10D sensor behind the 160mm lens, the perspective would be identical to what it is on the 4x5 camera, it just wouldn't show nearly as much of the scene. You'll only get telephoto compression by moving back and using a longer lens or cropping.

Rick "who thought he'd cleared this one up already" Denney

rdenney

7th of April 2005 (Thu), 20:30

THAT'S why the "factor" is referred to by those in the know as a "crop factor." :)

I really do admire your self-assuredness! I suppose I'm not "in the know" having done photography in every format from a Minox to a 4x5 camera for the last 30 years.

That 40mm lens doesn't know what you are putting behind it. Within whatever frame you provide you'll get the image magnified to the same degree.

Listen carefully: Camera position controls perspective. Focal length controls magnification. Format controls how much of the scene is included within the frame.

The equivalency factor is a way to know which lens will project the same image within a different frame. To get the same image in a bigger frame, you have to magnify it more. That's why you must have a longer lens--it magnifies more. The frame is 1.6 times bigger with 35mm, so you need 1.6 times more magnification to get the same image on that larger frame. That's why you use an 80 rather than a 50. But unless you move the camera position, the perspective remains the same.

Of course, you usually have control only of the camera position and the focal length. So, you start with position, and the choose the lens that provides the field of view for the given format.

Rick "who is not theorizing" Denney

CyberDyneSystems

7th of April 2005 (Thu), 21:19

Oh My god! http://members.bellatlantic.net/~vze2hdzu/pics/cower.gif

Can't we ever leave you guys alone for a few minutes without it all turning into another X-Factor debate? http://members.bellatlantic.net/~vze2hdzu/pics/banghead.gif

The title seemed innocent enough,... but no!http://members.bellatlantic.net/~vze2hdzu/pics/doh.gif

Just when it all seemed clear,. http://members.bellatlantic.net/~vze2hdzu/pics/duck.gif

rdenney

7th of April 2005 (Thu), 21:24

Can't we ever leave you guys alone for a few minutes without it all turning into another X-Factor debate?

Sheesh, CDS, I leave for a year or two and come back, and it's like I never even left. Do you mean that the x-factor debates left and then came back just for me?

Or are all you guys too worn out to take up the cause?

Rick "who'd be worn out, too, after 11 kiloposts" Denney

CyberDyneSystems

7th of April 2005 (Thu), 21:35

Rick,.. you mean you haven't seen this?.....

-=CROP FACTOR=- 10,000 posts on the X-Factor (http://photography-on-the.net/forum/showthread.php?t=45388)

...so clearly,. no ,. the X-Factor threads have never left,. they are with us...

Allways!

http://members.bellatlantic.net/~vze2hdzu/pics/gangpunch.gif

ilya

7th of April 2005 (Thu), 21:47

Listen carefully: Camera position controls perspective. Focal length controls magnification. Format controls how much of the scene is included within the frame.

Can we make a camera strap with this logo.

Can we make these words the only sticky on top of the forum.

Is this available in a tattoo.

slin100

7th of April 2005 (Thu), 21:50

By perspective (maybe the wrong word) I mean that in the shot with the 80mm on the AE-1, the image will show foreground-to-background compression that the 50mm/10D combo won't exhibit. It might be slight at 50mm vs. 80mm, but it's there.
Yes, perspective is the wrong word, and, no, the compression is not there.

This link is the best thing I could find to illustrate my point.
http://jamesmskipper.tripod.com/jamesmskipper/photo_perspective.html

Look at the architectural photos taken with the 55mm and 80mm lenses. That comes pretty close to duplicating a 50mm on 10D and 80mm on film setup. Making allowances for the fact that 55mm is 5mm too long, there is no difference in perspective and no difference in compression between the cropped 55mm and 80mm shots.

CyberDyneSystems

7th of April 2005 (Thu), 21:57

Can we make a camera strap with this logo.

Can we make these words the only sticky on top of the forum.....

I like this idea,. but only if the sticky also says this is the only subject where all the members and Mods are all allowed to point and laugh when it comes up again... :p

PhotosGuy

7th of April 2005 (Thu), 22:23

Can we make these words the only sticky on top of the forum.....
And put Ricks words at the top, "Listen carefully: Camera position controls perspective. Focal length controls magnification. Format controls how much of the scene is included within the frame." ;-)

Steven M. Anthony

7th of April 2005 (Thu), 22:59

Yes, but to get the same framing between the AE-1 and the 10D using the same lens (any lens), you will have to stand farther away from the subject with the 10D. Once you do that, the foreground/background compression changes.

Wish I had an AE1 with a 50mm lens. It would be fun to do this experiment. I'm sure someone, somewhere, has done it already.

But if you do that, it's not the same shot! And you have to move because the 2 lenses have different properties and behave differently. So to make the 2 shots look the same, you have to do all this compensation. If a 50mm on a 10D WAS THE SAME as an 80mm on an AE-1, you wouldn't need to compensate!

Steven M. Anthony

7th of April 2005 (Thu), 23:07

No, it won't. The relationship between foreground and background objects is controlled solely by the camera position. The only way you'll get compression is by moving back and using a longer lens or cropping the image more, but moving back is the required first step.

What?! You haven't seen the compression a long lens gives? Ever notice how at a NASCA race, when the cars come off a turn, and the camera is shooting down the straight, the cars look nose-to-tail? Well they are not--not coming out of a turn. It's the compression of the long lens. I'm sure someone here knows the math on that one. And I think you talked about it on your niagara falls shot you posted...

Remember what I said above about how camera companies in the old days would take a single wide picture and draw rectangles in it to show the effect of longer lenses. All the rectangles are on the same picture--the perspective is the same.

That's because it's just rectangles drawn on the same photo--used as a sales tool.

CyberDyneSystems

7th of April 2005 (Thu), 23:18

You use a longer lens,. to get the distance,. But it is the distance that gives the compression,.not the lens. Rick said that allready.

Steven M. Anthony

7th of April 2005 (Thu), 23:21

I really do admire your self-assuredness! I suppose I'm not "in the know" having done photography in every format from a Minox to a 4x5 camera for the last 30 years.

You know, sometimes I think we must just be confusing each other--because there are times I think we are saying the same thing; then others when you seem to not know what you are talking about--which doesn't fit with your obvious depth and breadth of experience and knowledge.

That 40mm lens doesn't know what you are putting behind it. Within whatever frame you provide you'll get the image magnified to the same degree.

That's exactly my point--the size of the film/sensor has no impact on the properties or behavior of the lens! That's why my pov is that the "format" of a lens/camera system is independent of sensor/film size. In other words, the 35mm format of the EOS 10D/EF lens system STAYS a 35mm format systen even with the 10D's APS-C sensor inside. Isn't that, in essence, what you are saying above?

Steven M. Anthony

7th of April 2005 (Thu), 23:28

Yes, perspective is the wrong word, and, no, the compression is not there.

This link is the best thing I could find to illustrate my point.
http://jamesmskipper.tripod.com/jamesmskipper/photo_perspective.html

Look at the architectural photos taken with the 55mm and 80mm lenses. That comes pretty close to duplicating a 50mm on 10D and 80mm on film setup. Making allowances for the fact that 55mm is 5mm too long, there is no difference in perspective and no difference in compression between the cropped 55mm and 80mm shots.

Look at the 2 pictures of the guy--50mm enlarged and cropped (which is what you'd have to do with the 10D's smaller file to get the same sized print) and 80mm. If you can't see the difference in apparent distance between him and the back of the chair, I guess you just can't see it.

Steven M. Anthony

7th of April 2005 (Thu), 23:33

You use a longer lens,. to get the distance,. But it is the distance that gives the compression,.not the lens. Rick said that allready.

How can distance give one compression?

Steven M. Anthony

7th of April 2005 (Thu), 23:38

slin100

7th of April 2005 (Thu), 23:40

Look at the 2 pictures of the guy--50mm enlarged and cropped (which is what you'd have to do with the 10D's smaller file to get the same sized print) and 80mm. If you can't see the difference in apparent distance between him and the back of the chair, I guess you just can't see it.
First off, that's not the picture I asked you to look at. I asked you to look at the architectural photo. Second, the example with the person is a terrible example. You are just imagining a compression effect because the man shifted position between the two shots. The man was the photographer. He had to switch lenses between shots.

So, why won't you respond to my original request? Examine the architectural photos and comment on my assertion that there is no perspective nor compression changes as the result of changing focal length.

slin100

7th of April 2005 (Thu), 23:44

From: http://www.cs.mtu.edu/~shene/DigiCam/User-Guide/4500/ON-CAMERA-LENS/optical-zoom.html (http://www.cs.mtu.edu/%7Eshene/DigiCam/User-Guide/4500/ON-CAMERA-LENS/optical-zoom.html)

The following are the images of the same scene taken with focal lengths 24mm (WC-E63 wide angle converter), 38mm, 100mm, 200mm and 310mm (TC-E2 2x tele converter) and 465mm (TC-E3ED 3x tele converter). As you can see, the wide angle side exaggerates foreground objects, while ``pushes'' background objects away. The moderate telephoto shot (i.e., 75mm) looks more natural. The longer the focal length we use, the more compression we will get. ... With this comparison in mind, if you wish to exaggerate the foreground, consider using a wide angle lens. ... If you wish to have a compressed effect that pulls distant objects close together, use a telephoto lens.Sorry, this example is irrelevant. The camera to subject distance was not kept constant. Now, I will quote Rick, "Camera position controls perspective."

slin100

7th of April 2005 (Thu), 23:57

That's exactly my point--the size of the film/sensor has no impact on the properties or behavior of the lens! That's why my pov is that the "format" of a lens/camera system is independent of sensor/film size. In other words, the 35mm format of the EOS 10D/EF lens system STAYS a 35mm format systen even with the 10D's APS-C sensor inside. Isn't that, in essence, what you are saying above?
Pardon me, Rick, if I may attempt to answer using Steven's own words, but corrected to represent Rick's POV.

The "format" of a lens/camera system is independent of the actual size of a lens's image circle. In other words, the APS-C format of the EOS 10D/EF lens system STAYS an APS-C format system even with an EF-S lens possessing an image circle smaller than an EF lens. This is, in essence, what Rick is saying.

Steven, see how this POV is completely flipped around from yours?

Moppie

8th of April 2005 (Fri), 03:26

Definitions are subjective and contextual.

35mm, 120, medium format, large format etc etc, they are all definitions.
But they are not what is being debated, despite what the first page of this thread would have you believe.

What is being debated is the behaviour of light as it passes through a lens and is projected onto a flat surface. That is not subjective or contextual, it is instead governed by the laws of physics, which as they relate to this topic are best summed up by the following:

Camera position controls perspective. Focal length controls magnification. Format controls how much of the scene is included within the frame.

Everything else simply surports the above, very simple, yet all inclusive statement.

There really is no debate, merely a lack of willingness to understand some very basic concepts hidden under clumsy definitions.

Steven M. Anthony

8th of April 2005 (Fri), 08:17

Pardon me, Rick, if I may attempt to answer using Steven's own words, but corrected to represent Rick's POV.

The "format" of a lens/camera system is independent of the actual size of a lens's image circle. In other words, the APS-C format of the EOS 10D/EF lens system STAYS an APS-C format system even with an EF-S lens possessing an image circle smaller than an EF lens. This is, in essence, what Rick is saying.

Steven, see how this POV is completely flipped around from yours?

Not really. Well, the way you put it, yes--but it seemed to me that what Rick was saying is that if you take a 35mm-format "40mm" lens, it magnifies X amount no matter what size sensor/film you stick behind it. Extending that to the image circle, is it not true that when attached to a specific 35mm-format camera body, the image circle from that 40mm lens will always be the same size--regardless of what size sensor is behind it? Because THAT'S my point. I don't know enough about lenses to know if the image circle of every 35mm-format lens is the same size. But I do know enough about optics to know that for a given fixed fl lens, its image circle is always the same on a given camera body--regardless of the sensor size behind it.

So, I guess even if the way I say it is "flipped around," we're still saying the same thing.

Belmondo

8th of April 2005 (Fri), 08:32

So, I guess even if the way I say it is "flipped around," we're still saying the same thing.

A good point at which to leave this topic.

rdenney

8th of April 2005 (Fri), 08:51

But if you do that, it's not the same shot! And you have to move because the 2 lenses have different properties and behave differently. So to make the 2 shots look the same, you have to do all this compensation. If a 50mm on a 10D WAS THE SAME as an 80mm on an AE-1, you wouldn't need to compensate!

You talk about lens behavior as if the lens has two legs and walks.

Here is what a lens does: It projects an image. That is all that it does. What image does it project? The image dictated by its focal length. The scene is magnified according to the focal length.

Lenses designed for bigger formats may cover a greater illuminated circle, but the magnification is the same. You can take the lens, hold it under an office flourescent light, and project the image of the light onto the surface of your desk. Try it now. A lens has no other "behavior". (Lenses do have other characteristics, of course, but none that bear on this discussion.)

I really do wish you would actually test some of your ideas with real stuff. You are theorizing based on what you think you know, instead of describing what you personally know to be true by having seen it with your own eyes.

So, an 80mm lens magnifies the image more. What is "magnification"? It means taking the image, AND MAKING IT BIGGER. How can a lens change the relationship between the foreground and background without making the camera move? The view from the camera can't change unless light rays can bend. THINK!

The image from the 80 on 35mm looks the same as the image from the 50 on a 10D because it magnifies the scene more to fill up the bigger frame with the same scenery information. The frame is 1.6 times bigger, so the magnification has to be 1.6 times greater.

So, when you change from a 10D/50 to a ElanII/80, YOU DON"T HAVE TO MOVE. The reason you don't have to move is because the 80 magnifies the image by just enough more to fill up the larger frame with the same scene. And because you don't move, the perspective doesn't change.

Look out the window, right now. What do you see? You see something in the foreground, and something behind it. What do you have to do to bring that background item into view. You have to move! If you want the background item to be bigger in relationship to the foreground item, what do you have to do? You have to move back. If you want the foreground item to be bigger with respect to the background, you have to move closer. What are your eyes? They are lenses that project onto a surface.

Consider this picture:

http://www.rickdenney.com/images/cedar-stump-crater-lake-lor.jpg

Can you tell by looking what focal length I used? I can't, and I took the picture. Without the EXIF, I wouldn't remember. What would it take for me to enlarge the stump with respect to the other side of the crater? I'd have to move closer to it. Zooming it wouldn't help--all that would do is cut off more of the outside of the image. If I wanted the stump to be smaller, I'd have to back up. No, I CHOSE the camera position because I wanted the stump to be THIS size with respect to the background, and then I found the focal length that would give me the composition I wanted from that spot. (It was the 150 setting on the 70-200/4L, as it happens.)

The reason the camera companies draw rectangles on a wide-angle picture to show longer lenses is because that is the exact image those longer lenses will show from that position. When you have conducted the experiment yourself, you will believe it. Get away from the keyboard and go take some freaking pictures.

When you finally get this, you are going to read your old posts and cringe. Don't let it bother you--it happens to the best of us.

Rick "wishing young photographers had some real training" Denney

rdenney

8th of April 2005 (Fri), 08:56

From: http://www.cs.mtu.edu/~shene/DigiCam/User-Guide/4500/ON-CAMERA-LENS/optical-zoom.html

The following are the images of the same scene taken with focal lengths 24mm (WC-E63 wide angle converter), 38mm, 100mm, 200mm and 310mm (TC-E2 2x tele converter) and 465mm (TC-E3ED 3x tele converter). As you can see, the wide angle side exaggerates foreground objects, while ``pushes'' background objects away. The moderate telephoto shot (i.e., 75mm) looks more natural. The longer the focal length we use, the more compression we will get. ... With this comparison in mind, if you wish to exaggerate the foreground, consider using a wide angle lens. ... If you wish to have a compressed effect that pulls distant objects close together, use a telephoto lens.

You realize, of course, that the photographer had to move the camera to keep the VW Bug and the birdhouse the same size in each of the pictures. If he had kept the camera in the same place, and take the pictures with those lenses, you'd be able to pull the long-lens picture out of the short-lens picture using only a pair of scissors.

Rick "who would rather see Stephen Anthony's pictures, so that he knows how they were taken" Denney

PacAce

8th of April 2005 (Fri), 08:57

rdenney

8th of April 2005 (Fri), 09:09

But I do know enough about optics to know that for a given fixed fl lens, its image circle is always the same on a given camera body--regardless of the sensor size behind it.

This is assuredly NOT true. Let's say you own two lenses: A Canon 24mm/2.8, and a Canon T/S 24mm. The tilt/shift lens has a much wider image circle than the other lens, because it has to accommodate the movements it allows. If you shift the lens, you still have to fill the frame. If you tilt the lens, you can't move the image circle such that it doesn't cover the frame. Thus, the 24mm T/S has a big monster front element, like a much wider-angle lens, to accommodate that wide illuminated circle. But if you zero out the movements, guess what? It produces exactly the same image as the 24/2.8. So what if it illuminates more of the inside of the camera body? I might even be able to cover 6x4.5 with that lens without using the movements. Does that make a Canon with that lens mounted on it a medium-format camera? Of course not! And the reason is that the frame in the camera is still just 35mm or APS.

Magnification is controlled solely by focal length. You agree with this, but please understand that magnification and image circle are not directly related. A Symmar 210 magnifies the image EXACTLY the same as a Super Angulon 210, but the Super Angulon has a much larger image circle. Yet both are specifically intended for use on 4x5 cameras. Assuming no movements, both will produce exactly the same image on 4x5 film.

Tessar designs have a 45-degree illumination circle. Plasmats are often 45 or 50 degrees. Super Angulons (in the f/5.6 form) have, if I'm remembering correctly, a 120-degree illuminated circle. And this is true no matter what focal length. But the magnification of the scene within the illuminated circle is based on focal length. Wider illuminated circles at the same focal length just include more scene.

Rick "who thinks you should forget what you think you know and start over" Denney

rdenney

8th of April 2005 (Fri), 09:17

You know, sometimes I think we must just be confusing each other--because there are times I think we are saying the same thing; then others when you seem to not know what you are talking about--which doesn't fit with your obvious depth and breadth of experience and knowledge.

With all due respect, you are measuring whether I know what I'm talking about by whether you can make my words fit what you think you know. A more intellectually honest approach would be to test what you think you know for yourself, rather than doggedly stick to it. I'm speaking from what I personally know to be true, having conducted the experiments myself in various ways over many years. That my point of view is aligned with simple physics doesn't injure my confidence!

Rick "who is not confused" Denney

Steven M. Anthony

8th of April 2005 (Fri), 09:49

I really do wish you would actually test some of your ideas with real stuff. You are theorizing based on what you think you know, instead of describing what you personally know to be true by having seen it with your own eyes.

And all I'm trying to do is understand your pov.

Steven M. Anthony

8th of April 2005 (Fri), 09:53

There really is no debate, merely a lack of willingness to understand some very basic concepts hidden under clumsy definitions.

I'm not sure how you can claim a lack of willingness here. It's possible that I just don't get what these others are saying--but it's not due a to lack of willingness to understand.

Steven M. Anthony

8th of April 2005 (Fri), 10:21

You realize, of course, that the photographer had to move the camera to keep the VW Bug and the birdhouse the same size in each of the pictures. If he had kept the camera in the same place, and take the pictures with those lenses, you'd be able to pull the long-lens picture out of the short-lens picture using only a pair of scissors.

No--I hadn't realised that. But now that you mention it, it IS obvious. I've learned something new--thanks.

But I'm still confused... I see that perspective can be altered with camera position (like the compression effect I brought up). But take slin100's architectural photos: To me, the overall perspective (forget about compression) of the full frame 55mm shot is quite different from the 80mm shot. But the images were taken from the same spot. So it seems that focal length AND position influence perspective.

This thread has kind of morphed into an X-factor discussion, which was not my intention. My interest really was to get at the notion of camera format. I think where it morphs is when you talk about a camera like a 10D--the body and lens system are 35mm-format, but the sensor isn't 35mm-sized.

I think we are agreed that (and I'll be extremely specific here) the image circle of a Canon 50mm/1.8 lens, when mounted on a 10D is always the same size--regardless of the size of the sensor behind it. The same would hold true for any fl lens (with the exception of the trick lenses mentioned above--but even for those, the circle will stay the same size for a given configuration of said lens, regardless of the size of the sensor behind it).

This is the basis for my assertion that camera/lens format is independent of sensor/film format. Another way to say the same thing is to assert that sensor/film format (i.e., size) is independent of the camera format it is contained in.

Slin100's architectural photos also illustrate why the X-factor can be correctly referred to as a "crop factor." The 10D sensor effectively creates the "55mm cropped" image when a 55mm lens is attached to it. It simply provides a "pair of sissors" version of the information passing through the lens--hence "crop factor."

Belmondo

8th of April 2005 (Fri), 10:25

But I'm still confused...

Yet another point of agreement. We're making progress, albeit slowly.

rdenney

8th of April 2005 (Fri), 10:34

And all I'm trying to do is understand your pov.

If I thought otherwise, I'd have given up long ago.

I did a diagram to illustrate the effect a while back for another similar thread, and here it is:

http://www.rickdenney.com/scratch/focal_length_and_format.gif

In the top diagram, we see a 10D with a 50mm lens. There are two items in the scene, a large foreground object that vertically fills the frame, and a smaller background object.

In the next diagram, the camera position is the same with respect to the two objects. The focal length is longer, and the nodal point is further away from the film. Thus, the magnification is larger. But the foreground object still fills the frame vertically, because the greater magnification offsets the effect of larger frame.

Importantly, the smaller background object is exactly the same percentage of the frame height in both diagrams. This is because the perspective remains the same--controlled by the camera position. There is no "compression effect" in comparing the Elan/80 and the 10D/50.

In the bottom diagram, we've taken the 50mm lens (and a 50 is a 50 is a 50, right?), and put it on the 35mm camera. Notice that to fill the vertical dimension of the frame with the foreground object, we have to move closer. And when we do that, the background object becomes smaller with respect to the foreground object. But the magnification of that 50mm lens is the same--pull the camera back to where it was, and the image would look just like it did at the top, except that the larger format would be able to include more scene (i.e., the foreground object would no longer fill the frame).

These diagrams are simplified, to be sure, and the vertical dimension is exaggerated with respect to the horizontal. But they are still in scale. Straight lines are straight lines.

Does that help?

Rick "eagerly awaiting the moment of understanding" Denney

Steven M. Anthony

8th of April 2005 (Fri), 10:35

Yet another point of agreement. We're making progress, albeit slowly.

And this comment helps how...?

Steven M. Anthony

8th of April 2005 (Fri), 10:48

Does that help?

Yes. But there is a diagram you don't show (and I still don't know how to get them to post without a link...), and I'll describe it in a bit. First, a comment on what you posted.

The top 2 diagrams illustrate the "equivalancy" you refer to when discussing the "X-Factor." But what the top diagram fails to show (although you mention it later in your text), is that there is more of the scene being projected beyond the bounds of the 15mm sensor width. When you show (or think of) that, you can see why calling it a "crop factor" is also an accurate term.

rdenney

8th of April 2005 (Fri), 10:58

This thread has kind of morphed into an X-factor discussion, which was not my intention. My interest really was to get at the notion of camera format. I think where it morphs is when you talk about a camera like a 10D--the body and lens system are 35mm-format, but the sensor isn't 35mm-sized.

It had to, because your assertion that format is a function of lens and camera is based on a misunderstanding of perspective and magnification. But I think you are getting there.

The 10D is NOT a 35mm camera. It is an APS camera that happens to accept lenses originally built for 35mm camera. If all the same glass was provided in EF-S mounts (and I mean absolutely identical in front of the mount), it would do the same things it does in EF mount. So, what makes a 10D different from an 1Ds? Sensor size, and sensor size alone. There is no difference if I put my adapted medium-format lenses on it. It's still an APS camera that, with the correct adaptor, will accept lenses originally intended for an Exakta 66.

Look, it's been said before. Definitions are tools. The definition establishes which word we agree will be henceforth used to describe what thing. The thing is what it is, and the word we agree to use in reference to it is something we decide, usually by common usage. Dictionaries just write down what people have already decided to do with the language.

So, is it useful to use the word "format" to describe the structure of the camera body and the design intent of the lens? I don't see how it is useful, because there are just too many exceptions. Also, there is nothing innately characteristic about the box or the placement of the lens mount that affects the resulting image. Boxes and lens mounts vary radically. Thus, using the word "format" to describe the box and the lens mount does not illuminate anything about the nature of the box, and it is refuted by too many exceptions.

But if we use the word "format" to describe the size of the frame, we have nailed the one characteristic element of that particular camera that distinguishes it from all others. The 10D is different from a 1Ds how? It has a smaller sensor. Bury both in the dirt and let aliens dig them up 1000 years from now, and they will think the two mostly identical, except that one had a smaller sensor than the other.

You will respond to that statement that they are therefore the same, and that the sensor size is just a cropped version of the larger sensor. Fine. We can always crop. But when we do, we are doing what photographers have always called wasting format. If we cut a 6x6cm square out of the middle of a 4x5 negative, they'll say we wasted format, and would should have used another lens. But when we put a rollfilm back on the same camera, we have changed its format.

And if we use the term "format" to describe the box and the lens mount, what are medium-format and large-format photographers supposed to use? Some of them don't even have boxes, and their cameras have parts that interchange across film sizes all over the place.

So, the use of "format" to describe solely the size of the image frame provides us with a word that 1.) most people understand because it fits with common usage, 2.) identifies the one fundamentally distinguishing characteristic of the camera, and 3.) avoids the confusing of too many exceptions.

Rick "going back to the original topic, since you asked" Denney

rdenney

8th of April 2005 (Fri), 11:07

Yes. But there is a diagram you don't show (and I still don't know how to get them to post without a link...), and I'll describe it in a bit. First, a comment on what you posted.

The top 2 diagrams illustrate the "equivalancy" you refer to when discussing the "X-Factor." But what the top diagram fails to show (although you mention it later in your text), is that there is more of the scene being projected beyond the bounds of the 15mm sensor width. When you show (or think of) that, you can see why calling it a "crop factor" is also an accurate term.

Call it a crop factor if you want; the term is not inherently wrong. But it can be misleading, and it has obviously misled you, because you have said in this thread that an 80mm lens will show more "compression" on a 35mm frame than a 50mm lens on a 10D, and that is what my diagrams clearly refute. But back to the term, crop factor. When I put a 6x9 rollfilm back on a 4x5 camera, do I try to convert my understanding of the lenses using a "crop factor". NO! I just know that a 100mm lens is normal for 6x9 and 160mm lens is normal for 4x5, and therefore they are equivalent to each other when used on their respective formats. No large or medium-format photographer would ever think of this in terms of a "crop factor". That's why I said your perspective is solely that of a 35mm photographer who happens to have bought a digital SLR with an APS-sized sensor, and you are trying to relate everything back to a single frame of reference--35mm. I'm trying to bring your terminology into the wider world of photography--it will help you communicate, and it may help you understand the difference between perspective, magnification, and field of view.

Rick "who thinks of each format in its own terms" Denney

slin100

8th of April 2005 (Fri), 11:08

But I'm still confused... I see that perspective can be altered with camera position (like the compression effect I brought up). But take slin100's architectural photos: To me, the overall perspective (forget about compression) of the full frame 55mm shot is quite different from the 80mm shot. But the images were taken from the same spot. So it seems that focal length AND position influence perspective.

The compression effect is not considered a formal part of photographic definition of perspective. Photographic perspective concerns itself with the size relationships of objects in the image relative to each other, not relative to the viewer. Take, for example, an image taken from a certain distance where a foreground object appears twice as tall as a background object. Without moving, if the focal length is changed, the expanse of the scene captured will vary, but the foreground object will always appear twice as tall as the background object. That is the basis behind the assertion that camera position solely controls photographic perspective.

There is a broader definition of perspective that the common layman subscribes to, which basically includes distortion effects like compression. IOW, how things look. This is the definition of perspective that you are applying. It would behoove us all if you would adopt the narrower, photographic definition when using the word, "perspective."

I think we are agreed that (and I'll be extremely specific here) the image circle of a Canon 50mm/1.8 lens, when mounted on a 10D is always the same size--regardless of the size of the sensor behind it. The same would hold true for any fl lens (with the exception of the trick lenses mentioned above--but even for those, the circle will stay the same size for a given configuration of said lens, regardless of the size of the sensor behind it).

This is the basis for my assertion that camera/lens format is independent of sensor/film format. Another way to say the same thing is to assert that sensor/film format (i.e., size) is independent of the camera format it is contained in.

You cannot draw this conclusion from your basis. Let me put it this way. The photons that are captured are the same, modulo a crop. This is the point you've been making all along, and all of us agree with you on that point. But after this, you and the rest of us diverge.

What you seem to be missing is how those photons are interpreted. The only way a meaningful interpretation between two different formats can be performed is to assess them using a standard print size. It doesn't matter what the size of the print is (4x6, 8x10, web thumbnail), but it must be kept fixed.

Because an APS-C sensor is smaller than a 35mm frame, we must enlarge it more to reach the standard print size. Because more enlargement is required, the data on the APS-C is, in effect, being examined more critically than the 35mm frame. This implies that the various standards of judgement (DOF, sharpness, motion-induced blur) must be met with tighter controls at acquisition time. This means that shutterspeeds must be faster to counter camera shake (the 1/fl rule :)) and that apertures must be closed down to counter the narrower DOF resulting from the smaller CoC. All this flows from the fact that the size of the sensor is smaller. Therefore, the format is defined by the sensor.

slin100

8th of April 2005 (Fri), 11:11

Rick, I really admire your patience as well as the clarity and eloquence of your words.

Belmondo

8th of April 2005 (Fri), 11:24

And this comment helps how...?

I thought the point of these protracted exchanges was to determine points of agreement. I recognized one and felt obliged to point it out.

It seems to me that nobody is going to be satisfied (or silenced) until someone actually takes two camera of different formats, and tests the theory. Hopefully I'll have an opportunity to do it later today with my 1.3X and 1.6X cameras. I just want to do the math first.

Peace.

Steven M. Anthony

8th of April 2005 (Fri), 11:32

It had to, because your assertion that format is a function of lens and camera is based on a misunderstanding of perspective and magnification. But I think you are getting there.

No, it is a function of recognizing that sensor/film size is independent from the properties of the camera/lens system.

The 10D is NOT a 35mm camera. It is an APS camera that happens to accept lenses originally built for 35mm camera.

I think an equally valid assertion is that the 10D is a 35mm format system that happens to have an APS-C-sized sensor in it. It's a "one man's celieng is another man's floor" issue. If you define the format by the size of the recording medium, the 10D is APS-C format. If you define it by the body/lens system the recording mediun sits in, it's a 35mm format.

So, it looks like the answer to my original question is "either!"

If all the same glass was provided in EF-S mounts (and I mean absolutely identical in front of the mount), it would do the same things it does in EF mount. So, what makes a 10D different from an 1Ds? Sensor size, and sensor size alone. There is no difference if I put my adapted medium-format lenses on it. It's still an APS camera that, with the correct adaptor, will accept lenses originally intended for an Exakta 66.

Look, it's been said before. Definitions are tools. The definition establishes which word we agree will be henceforth used to describe what thing. The thing is what it is, and the word we agree to use in reference to it is something we decide, usually by common usage. Dictionaries just write down what people have already decided to do with the language.

So, is it useful to use the word "format" to describe the structure of the camera body and the design intent of the lens?

For me, yes. Maybe because all I've ever used is 35mm format cameras. If I did other formats like you have, maybe it wouldn't be useful. I find it easy to visualize the image circle filling a 35mm film-sized rectangle on the image plane of my 10D and the fact that its sensor is recording only part of it. Sure there are exceptions involved, but there are exceptions involved in the format=film size pov, too--when your experience has been limited to 35mm format.

...Bury both in the dirt and let aliens dig them up 1000 years from now, and they will think the two mostly identical, except that one had a smaller sensor than the other.

And they might well say "Hey, two 35mm cameras, except one has a smaller sensor than the other..."

You will respond to that statement that they are therefore the same, and that the sensor size is just a cropped version of the larger sensor. Fine. We can always crop. But when we do, we are doing what photographers have always called wasting format.

This phrase actually supports the validity of my pov--the 10D's APS-C-sized sensor is "wasting" part of the 35mm format the system was designed for...

If we cut a 6x6cm square out of the middle of a 4x5 negative, they'll say we wasted format, and would should have used another lens. But when we put a rollfilm back on the same camera, we have changed its format.

An equally valid pov...

And if we use the term "format" to describe the box and the lens mount, what are medium-format and large-format photographers supposed to use? Some of them don't even have boxes, and their cameras have parts that interchange across film sizes all over the place.

Not every design needs a format designation. And at least one format is essentially just a box (the pin-hole camera).

Steven M. Anthony

8th of April 2005 (Fri), 11:42

Call it a crop factor if you want; the term is not inherently wrong. But it can be misleading,

As can "equivalency factor"...

and it has obviously misled you, because you have said in this thread that an 80mm lens will show more "compression" on a 35mm frame than a 50mm lens on a 10D, and that is what my diagrams clearly refute.

That misunderstanding was unrelated to viewing the X-factor as a crop factor. My compression misunderstanding pre-dates (by decades) the advent of affordable digital cameras...

But back to the term, crop factor. When I put a 6x9 rollfilm back on a 4x5 camera, do I try to convert my understanding of the lenses using a "crop factor". NO! I just know that a 100mm lens is normal for 6x9 and 160mm lens is normal for 4x5, and therefore they are equivalent to each other when used on their respective formats.

Good for you, Rick! But since the term "crop factor" isn't inherently wrong and conveys more useful information to me than does "equivalency factor," do you mind if I keep using it?

...your perspective is solely that of a 35mm photographer who happens to have bought a digital SLR with an APS-sized sensor, and you are trying to relate everything back to a single frame of reference--35mm.

But that IS my perspective. Is it wrong to only have had experience with 35mm format? If not, then it isn't wrong of me to think of the 10D in terms of 35mm format, right?

slin100

8th of April 2005 (Fri), 11:44

Steven M. Anthony

8th of April 2005 (Fri), 12:06

No. The two assertions are not equivalent.

If you are able to shift your pov (like seeing a ceiling as someone else's floor) you would see the equivalence.

The problem with this POV is that it has convinced you that you don't need to adjust your technique. You insist that 1/true focal length instead of 1/effective focal length works and that DOF is unaffected. These conclusions are not correct. The proof in these truths lies somewhere out there. We're all trying to help you find it.

Well, I never use the 1/fl rule myself. I know, from experience, how slow a shutter speed I can use with minimal camera shake. I don't recall commenting on DOF.

I think the truth of it is that camera format can be approached from different perspectives (which, I guess, shouldn't surprise us given the subjective nature of reality).

I guess I favor the "format is what it was designed for" perspective. That's the "a football is still a football even if you shoot hoops with it" pov. But I guess an equally valid pov is that once you shoot hoops with it, it becomes a basketball.

slin100

8th of April 2005 (Fri), 12:21

If you are able to shift your pov (like seeing a ceiling as someone else's floor) you would see the equivalence.

IMO, they are not equivalent because of the very different conclusions that are drawn from each.

Well, I never use the 1/fl rule myself. I know, from experience, how slow a shutter speed I can use with minimal camera shake.

Ahh, but is that experience derived from your days of handling a 35mm camera or from a 10D?
I don't recall commenting on DOF.
I apologize, then, for putting words in your mouth. The statement, then, becomes an extrapolation of your POV.

I guess I favor the "format is what it was designed for" perspective. That's the "a football is still a football even if you shoot hoops with it" pov. But I guess an equally valid pov is that once you shoot hoops with it, it becomes a basketball.
I think you're doing yourself a disservice by adopting this POV. Using your analogy in a different way, the relevant question is not what to call the ball, but which game's rules do you use? I think you would say that if you have a football, then you play by football's rules, even though you are shooting hoops. "He shoots! Touchdown!" :)

rdenney

8th of April 2005 (Fri), 12:28

I think an equally valid assertion is that the 10D is a 35mm format system that happens to have an APS-C-sized sensor in it. It's a "one man's celieng is another man's floor" issue. If you define the format by the size of the recording medium, the 10D is APS-C format. If you define it by the body/lens system the recording mediun sits in, it's a 35mm format.

The problem is that you will only be speaking to yourself. Consider the responses you have seen in this thread. Do you think your suggested definition has made anything more clear for anyone other than yourself?

I think you'll find yourself embroiled in these discussions every time you use the term in the manner you are using it, because it is not the common usage and it therefore confuses people. Eventually, people will ignore you. That would be a shame.

Let me give you an example from another of my hobbies:

I'm a tuba player. Tubas (I'll bet you didn't know this, heh, heh) are sized both by length and width. In length, they are sized in three categories: contrabass, bass, and tenor. Contrabass tubas are pitched in Bb or C, bass tubas in Eb or F (half an octave higher), and tenor tubas in Bb or C, in the same octave as trombones. In width, they are categories in quarters. 4/4 tubas are "full-sized", whatever that means. Fatter tubas are 5/4, and the fattest tubas are 6/4. Small tubas are 3/4.

In the tuba world, nobody really agrees on definitions of these designations, but pretty much everyone can tell by looking whether an instrument is a 4/4, 5/4, or 6/4. And not everyone agrees that a "tenor tuba" is a broad category. Some think of it as a specific type of instrument, and leave baritone horns and euphoniums as Something Else. Wars have been fought over the difference (in America at least--in England the words are used differently) between a "baritone" and a "euphonium". The truth is--there is no difference in physical terms. But we can still tell by looking which is which, with few exceptions.

The point? Definitions, to be effective, require broad agreement. They are informative, not normative. So, when you try to define your word your way, against the expections of those with whom you speak, you block understanding rather than enhance it. When I insist on using the term "baritone" to describe what most people think is a "euphonium", I entertain only myself. When I insist on lumping all tubas pitched the same as trombones into the category "tenor tuba", I cut off many to whom I'm talking. When I insist that a particular instrument is a 6/4 when most people think of it as a 5/4, our conversational subject changes to words rather than tubas. Thus, despite my strong desire for a clear and defensible taxonomy, I use the words the same way others do.

But in this case, the taxonomy against which you are arguing is indeed defensible, from the point of view of most to whom you will speak.

By the way, I never said you couldn't use the term "crop factor". I'll only complain if I see confused understanding of perspective, magnification, and field of view to go along with it.

Rick "who thinks there is too little distinction and too much overlap between classes of 'body/lens systems' to support definitions without confusion, unless you include the frame size" Denney

Steven M. Anthony

8th of April 2005 (Fri), 12:53

IMO, they are not equivalent because of the very different conclusions that are drawn from each.

Resulting in different conclusion might be a necessary condition for inequivalence, but it is certainly not a sufficient condition.

Ahh, but is that experience derived from your days of handling a 35mm camera or from a 10D?

Well, they are both 35mm cameras, so it's hard to answer that... :) But the majority of my experience has been with film cameras. But interestingly, as someone else has pointed out on a thread months ago, I find I can hold the 10D steadier than my film camera. One reason offered up on that old thread was that the added weight provided more counterforce for muscles to push against.

I apologize, then, for putting words in your mouth. The statement, then, becomes an extrapolation of your POV.

I don't see how it is an extrapolation. I know that the dof of a 50mm on a 10D is the same as the DOF of a 50mm on an AE-1--and that both will differ from the DOF of an 80mm. Am I wrong on this, too?

I think you're doing yourself a disservice by adopting this POV. Using your analogy in a different way, the relevant question is not what to call the ball, but which game's rules do you use? I think you would say that if you have a football, then you play by football's rules, even though you are shooting hoops. "He shoots! Touchdown!" :)

Absolutely not. I'd say we were playing basketball with a football. I would just recognize that my using the ball in a way its designers didn't intend it to be used doesn't alter the intended use of the ball, or its being a football. Just like to me, the 10D is a 35mm camera with a "format wasting" sensor.

Steven M. Anthony

8th of April 2005 (Fri), 13:02

The problem is that you will only be speaking to yourself. Consider the responses you have seen in this thread. Do you think your suggested definition has made anything more clear for anyone other than yourself?

Oh come now, Rick. Just because I don't take your pov doesn't mean I stand alone in the universe on this issue...

I think you'll find yourself embroiled in these discussions every time you use the term in the manner you are using it, because it is not the common usage and it therefore confuses people. Eventually, people will ignore you. That would be a shame.

People may, indeed come to ignore me, but I doubt it will be because I consider a 10D a 35mm camera or refer to the X-factor as a crop factor. :) And at this point, should the issue arise again, the answer is that both views are valid. End of embroilment!

Let me give you an example from another of my hobbies:

...In the tuba world, nobody really agrees on definitions of these designations...

Is that why the USA has a trade embargo with Tuba?

The point? Definitions, to be effective, require broad agreement. They are informative, not normative.

I can't tell you how many photogs I've explained the crop factor to who say--"Oh, now I get it." Maybe we just run in different circles... (all relevant puns intended)

slin100

8th of April 2005 (Fri), 13:13

Resulting in different conclusion might be a necessary condition for inequivalence, but it is certainly not a sufficient condition.

I believe it is sufficient. Equivalence is a bidirectional property. It means that the necessary and sufficient conditions can be exchanged. A is equivalent to B if A -> B and B -> A both hold. If A -> C, B -> D and C != D, you cannot say that A is equivalent to B.

I don't see how it is an extrapolation. I know that the dof of a 50mm on a 10D is the same as the DOF of a 50mm on an AE-1--and that both will differ from the DOF of an 80mm. Am I wrong on this, too?

Yes. Go to any DOF calculator that has provisions for specifying the CoC. My personal favorite is http://dofmaster.com/dofjs.html. Punch in a 50mm lens, aperture and distance. Then toggle the format between 35mm and 10D. The format selection is a just a shorthand way of changing the CoC. You will see the DOF numbers change.

Absolutely not. I'd say we were playing basketball with a football. I would just recognize that my using the ball in a way its designers didn't intend it to be used doesn't alter the intended use of the ball, or its being a football. Just like to me, the 10D is a 35mm camera with a "format wasting" sensor.
You are playing with a 10D (the football) using the rules of basketball (35mm photography). If you are going to use a football, you should play by football's rules (APS-C photography).

Steven M. Anthony

8th of April 2005 (Fri), 14:26

I believe it is sufficient. Equivalence is a bidirectional property. It means that the necessary and sufficient conditions can be exchanged. A is equivalent to B if A -> B and B -> A both hold. If A -> C, B -> D and C != D, you cannot say that A is equivalent to B.

What conclusion can you draw from one POV that you can't from the other?

You will see the DOF numbers change.

Sure, I guess that makes sense given that CofC and DOF values are subjective measures based, in part, on an 8 X 10 print...

You are playing with a 10D (the football) using the rules of basketball (35mm photography). If you are going to use a football, you should play by football's rules (APS-C photography).

I'm sorry, but I consider myself free to use a football as I see fit--be it to play basketball or to use as a door-stop, or a floatation device...--as the situation demands. Heck, if McGyver only used things for their original, intended purpose, he'd be killed off in the first episode! :)

slin100

8th of April 2005 (Fri), 14:38

What conclusion can you draw from one POV that you can't from the other?

That DOF and the 1/fl rule of thumb are affected by the size of the sensor, to give you two.

Sure, I guess that makes sense given that CofC and DOF values are subjective measures based, in part, on an 8 X 10 print...

But if you admit that, then you must admit there are a whole host of other things that are affected by the size of the sensor. That's why it makes sense to say that the format of a camera system is dictated by the size of the sensor.

I'm sorry, but I consider myself free to use a football as I see fit--be it to play basketball or to use as a door-stop, or a floatation device...--as the situation demands. Heck, if McGyver only used things for their original, intended purpose, he'd be killed off in the first episode! :)
I never said you weren't free. It just that people will look at you funny and not understand you. That was Rick's point about adhering to conventions.

CyberDyneSystems

8th of April 2005 (Fri), 15:23

Source = "A Glossary of Photographic Terms"
Format:

The actual size of the photograph, either slide or negative, produced by a camera; in 35mm photography, the picture measures 24mm x 36mm and has a diagonal of 43mm, While the new APS (Advance Photo System), several new formats were included, including panorama . While it can also be explained as shape and size of image provided by camera or presented in final print or transparency. Governed in the camera by the opening at the rear of the body over which the film passes or is placed. The standard 35 mm format is 36 x 24 mm; half-frame, 18 x 24 mm; 126 size, 28 x 28 mm; 110, 17 x 13 mm; standard roll film (120 size), 2x 2 in.

Steven M. Anthony

8th of April 2005 (Fri), 15:53

That DOF and the 1/fl rule of thumb are affected by the size of the sensor, to give you two.

But whether you view a 10D as a 35mm format camera/lens system with an APS-C sensor in it or an APS-C format camera that accepts 35mm-format lenses, the conclusions one would reach on those issues is the same--as both are tied to the sensor size and both povs acknowledge the sensor is APS-C-sized.

But if you admit that, then you must admit there are a whole host of other things that are affected by the size of the sensor. That's why it makes sense to say that the format of a camera system is dictated by the size of the sensor.

But to me, as I've pointed out several times before, it makes more sense to say it the other way around. Given that we've established the equivalence of the 2 povs, why the effort in getting me to look at it your way?

I never said you weren't free. It just that people will look at you funny and not understand you. That was Rick's point about adhering to conventions.

Well, viewing it my way has helped others undrstand the issue--so not everyone will look at me funny. And besides, in some circles, taking an unconventional view is actually valued!

Moppie

8th of April 2005 (Fri), 15:57

I’m just going to restate something I said earlier, because it seems that while the physics have been understood, at least to some degree, the nature of the terms used to differentiate different types of photographic equipment have not.

Definitions are subjective and contextual.

35mm, 120, medium format, large format etc etc, they are all definitions.
But they are not what is being debated, despite what the first page of this thread would have you believe.

A cameras format is defined by two things, its manufacturer, and common language at the time through popular culture. There are some general rules that are followed, but there is nothing set in hard law that gives concrete definitions of different formats.

As a general rule a camera comprising a body and lenses capable of projecting an image on to a 4x6 piece of film will be called a 4x6 format camera.

It’s the definition that best fits what the camera does, and was designed to do.

If the camera has a removable back, and you take off the 4x6 back and replace it with a 35mm one, it will still be a 4x6 format camera. Nothing has changed that will stop the camera making 4x6 format photos, provided the 4x6 back can be re-installed.

However for the time the 35mm back is installed some people might want to re-define the cameras format.

There is nothing written in law, either, official or unofficial that says this must be done and how. Its best technical definition would be a 4x5 format camera that produces 35mm format negatives. But some people might want to call it a Custom 4x6 format, others might want to call it a Hybrid 35mm. I might want to paint it in bright colours and call it the Moppeta 35mm. All are quite valid definitions for that camera.

The negatives it produces of course would be in the 35mm format, which is dependant on the film used, and not the camera (although these could be enlarged, and cropped to look like 4x6 format photos, with out the same level of quality).

Definitions are vague and often random things, they change with time, between cultures and languages. They are related to the social practices of the culture applying them rather than to any scientific law, and so debating them is generally a pointless exercise.

CDS’s headach over the magic 1.6x factor is largely a result of nobody sitting down and defining what it was that was in debate.

This thread has, despite its length and technical nature actually managed to define what it is being debated.

- How does lens size, projected image circle and film/sensor size effect things like perspective, magnification etc.

- How do we define cameras what have a sensor/film size that is of a quite different size to the projected image circle from the lens.

The first one has been answered; the physics involved is clearly explained in this thread.
It is not up for debate, unless your Steven Hawking or Isaac Asimov.

The second can never be answered in a format such as this. It is not something that a single thread in a small forum can ever be influenced by.

Instead popular culture combined with the manufactures responsible will eventually come up with its own definitions.

What they will be I can not tell you, I simply don’t know enough about them, but in the mean time their lack does not seem to stop people taking some great photos and sharing an absolute wealth of knowledge about how it was done.

If the lack of a firm definition for a cameras format bothers you then you might want to just stop for a second ask if it affects how your camera takes a photo.

If you decide it does then sell your camera, and find another hobby. Because how your camera is defined has no effect on how it works, and instead of trying to argue definitions that are beyond your control you should instead be out taking photos and sharing them, and the knowledge you gained doing so with the rest of us on this forum.

You can not argue definitions, you can state them, you can accept them or reject them, but you can not argue them.

How you define something is irrelevant to how it functions.

slin100

8th of April 2005 (Fri), 16:09

But whether you view a 10D as a 35mm format camera/lens system with an APS-C sensor in it or an APS-C format camera that accepts 35mm-format lenses, the conclusions one would reach on those issues is the same--as both are tied to the sensor size and both povs acknowledge the sensor is APS-C-sized.

That's strange. Before, you were saying the conclusions were different. Before, you were suggesting that a 10D is a 35mm camera subject to the guidelines of 35mm photography. If you think this is equivalent to saying that a 10D is an APS-C camera subject to the guidelines of APS-C photography, then the issue isn't settled.

For example, I still don't know what you format you would call a 20D with a 18-55 EF-S. APS-C or 35mm?

But to me, as I've pointed out several times before, it makes more sense to say it the other way around. Given that we've established the equivalence of the 2 povs, why the effort in getting me to look at it your way?

Because I maintain that they are not equivalent.

Steven M. Anthony

8th of April 2005 (Fri), 16:26

I’m just going to restate something I said earlier, because it seems that while the physics have been understood, at least to some degree, the nature of the terms used to differentiate different types of photographic equipment have not...

Very well put. I had been thinking of the hybrid issue myself. And I think you are onto something with the Moppeta... :)

Steven M. Anthony

8th of April 2005 (Fri), 16:39

That's strange. Before, you were saying the conclusions were different. Before, you were suggesting that a 10D is a 35mm camera subject to the guidelines of 35mm photography. If you think this is equivalent to saying that a 10D is an APS-C camera subject to the guidelines of APS-C photography, then the issue isn't settled.

Reflecting on Moppie's recent comment, I think of the 10D as a hybrid 35mm/APS-C. Some aspects conform to APS-C format (like DoF calculations) and some conform to 35mm format (like the 1/fl rule (when considering the camera shake RECORDED ON THE SENSOR)).

For example, I still don't know what you format you would call a 20D with a 18-55 EF-S. APS-C or 35mm?

Right now, I consider it a hybrid, like the 10D.

Because I maintain that they are not equivalent.

Okay.

rdenney

8th of April 2005 (Fri), 17:19

Reflecting on Moppie's recent comment, I think of the 10D as a hybrid 35mm/APS-C. Some aspects conform to APS-C format (like DoF calculations) and some conform to 35mm format (like the 1/fl rule (when considering the camera shake RECORDED ON THE SENSOR)).

Hold on, Steven. Remember that the mechanism that makes it an APS camera regarding depth of field is that the image will be enlarged more to get the same size print. That's the same mechanism that makes camera shake also related to the frame size and not just the focal length. You have to take both or neither.

Frankly, it's okay with me if we say the 10D is a hybrid APS camera. And I quite agree with Moppie's point that a 4x5 camera with rollfilm back is still goig to be called a 4x5 camera, if it still supports its use as a 4x5 camera.

If we decide that because the 10D is an APS camera (or that even just that it has an APS sensor), we will not enlarge the pictures beyond 8x10, and if we decide that we will not enlarge 35mm beyond 13x19, then 1/fl works as well for both (which is to say it is mostly defined by when it DOESN'T work, heh, heh). We can also say that depth of field is more for shorter lenses, so if our aim is to make the same image using an APS camera and a 35mm camera, we have to use a longer lens for 35mm, and we will get less depth of field therefore. But some of that difference will be eaten up by the smaller print and therefore lower sharpness standards for the APS.

But if we say that we expect the 10D to support larger enlargement ratios than a 35mm camera, as we surely must considering wide usage, then I think we have to consider all performance aspects in light of that extra enlargement, including camera movement, depth of field, and so on. Of course, that would be a convention, not a law.

If you don't accept that degree of enlargement should be considered, then all cameras are 35mm by your definition, because camera shake in absolute distances on the image plane is related to focal length alone, and it is therefore the same for all formats. We should run and tell those Speed Graphic users that they should not be using shutter speeds slower than 1/125 with their 127mm wide-angle lenses!

By the way, you guys should go look at Steven's pictures. Even though his skull is made of armor plate, and even though he's newly reformed on his understanding of perspective, magnification, and field of view, it's quite clear to me that he has a photographer's eye. So, if I said anything that suggested he was in the "whippersnapper" category, I retract it forthwith.

Rick "who realizes that what matters is what ends up on in the image" Denney

ilya

8th of April 2005 (Fri), 17:42

By the way, you guys should go look at Steven's pictures. Even though his skull is made of armor plate, and even though he's newly reformed on his understanding of perspective, magnification, and field of view, it's quite clear to me that he has a photographer's eye. So, if I said anything that suggested he was in the "whippersnapper" category, I retract it forthwith.

Rick "who realizes that what matters is what ends up on in the image" Denney

you're right. good job sma

Ilya "who can respect good work even if its done by a pita"

RockOne

8th of April 2005 (Fri), 18:30

If after reading this thread you aren't confused then you don't really understand whats going on ? :-) :-) :-) :-).
You guys need to get out and take a few more photos ;-) :-) :-)

Steven M. Anthony

8th of April 2005 (Fri), 19:03

Hold on, Steven. Remember that the mechanism that makes it an APS camera regarding depth of field is that the image will be enlarged more to get the same size print. That's the same mechanism that makes camera shake also related to the frame size and not just the focal length. You have to take both or neither.

Yeah--but for DoF, calculations are based on a constant (8X10) print size. Such a constant does not exist for camera shake! Nor, I assert, is there likely to be an agreed-upon level of "acceptable" camera shake. So with that in mind, I'm having it both ways! :)

... Of course, that would be a convention, not a law.

And there is nothing more futile than trying to legislate convention...

If you don't accept that degree of enlargement should be considered, then all cameras are 35mm by your definition, because camera shake in absolute distances on the image plane is related to focal length alone, and it is therefore the same for all formats. We should run and tell those Speed Graphic users that they should not be using shutter speeds slower than 1/125 with their 127mm wide-angle lenses!

Sure it should be considered. But as someone said before (was it you?) that if one consideres print size, then 1/fl never really works... And what about pixel dimensions. My Sony 707 (5MP) has a tiny sensor, but produces images that are 2560 pixels X 1920 pixels. At 300ppi, that's an 8.5 X 6.4 image. My 10D (6.3MP) has a larger sensor (and more pixels), and produces images that are 3072 X 2048. At 300 ppi, that's a 12.8 X 8.5 image. So, even though the sensor in the 707 is tiny in comparison to the 10D's, the image size is comparable (given the difference in MPs). So how does sensor pixel dimensions work into all of this?

slin100

8th of April 2005 (Fri), 21:00

Yeah--but for DoF, calculations are based on a constant (8X10) print size. Such a constant does not exist for camera shake! Nor, I assert, is there likely to be an agreed-upon level of "acceptable" camera shake. So with that in mind, I'm having it both ways! :)

Actually, the rule of thumb for camera shake is also based on a constant print size. As you adjust the print size up or down, the discernment of camera shake also goes up or down, so you adjust the rule of thumb accordingly. The same factors that influence DOF also affect camera shake.

rdenney

11th of April 2005 (Mon), 08:52

I think that somewhere in the near future, one of the formats will be abandoned and the terms no longer applied to digital media. We'll probably see cameras not much different in appearance from the present DSLRs, but they'll have hellacious resolution capabilities. We'll also see DSLRs with the same technical capabilities as the still-popular view cameras, but they'll be smaller than we know them now. (Yeah, I know, the smaller capture medium will limit such a view camera's camera's capabilities to some extent, but i can't see a 4"x5" CMOS being produced any time in the near future).

The term "view camera" has never implied "large format", particularly not just because it has movements. There have been and continue to be purpose-made medium-format view cameras on the market. The only difference between a medium-format view camera and a 4x5 view camera is the maximum format it will accommodate.

And when I put my 6x9 rollfilm back on my 4x5 camera, in that case it's just a rollfilm back, and I've never heard anyone refer to it as a reducing back.

Of course, there are many cameras with no or very limited movements that accept large-format film (4x5 and up) beyond the Gowlandflex (which was demonstrated to me personally by Gowland himself in 1978 at a, um, well-attended college lecture). Speed Graphics are one such, despite the rising lens standard. They used bellows in all formats at one time, with hard box cameras like the Rolleiflex and the Leica being the exceptions. And even the Gowlandflex has bellows, ala the Mamiya C-330 TLR.

I fear you may be correct that we will see higher and higher pixel densities in lieu of larger and larger sensors, but that is a pity. The mass market has no need of sensors with greater resolution or size than a Digital Rebel--they can make prints from those images already larger than they are used to, and at better quality. Few people in the consumer market make prints larger than 4x6", and when they do, it's an 8x12.

Those willing to spend over $1000 on a camera body, however, have different needs. And those needs are not met by more and more pixels stuffed into an APS sensor. I can already see the limitations of my lenses in my 10D images--and some of my lenses are among the best made. And there is a limit on the number of photons that we can expect to fall into a photosite on the sensor. I daresay one could not make a bigger print from a 8mp 20D than from a 6mp 10D at the same quality, because of those other limitations. Small sensors have reached their image-making potential, and no improvement in technology can increase the number of photons coming from the scene, though they may attempt to simulate them.

No, what we need are not smaller pixels but more of them, meaning larger sensors.

There currently are available digital scanning backs for studio product photography that cover 4x5. They produce images of hundreds of megabytes on a tethered computer, but that is what is needed for that photo of the new zoomy sports car printed in a 133-line process color full-bleed two-page brochure spread at 11" by 17". That is much more demanding than what most people face, and quite beyond the capability of a hand-held digital camera at present, even a 1DsII.

I scan medium-format film at 1128pixels per inch, giving me an image about 2600 pixels square. Despite having about the same number of pixels, the image quality exceeds the images from a 10D significantly, simply because each pixel is more accurate--it integrates more light. You can tell the difference even in images reduced for web display--the larger formats have may have the same number of pixels, but they have a higher information density.

And there are not a few pros who have paid twice the cost of a 1Ds for a Kodak digital capture back for their Hasselblads, with a sensor sized at 36x36, though it has no more pixels than a 1DsII. Why spend so much? Because format is king.

One thing DSLR's and digicams have done, though, is open the question about what is format. There is such a variety of different sensor sizes and shapes, that some means of knowing what focal lengths mean became necessary. It was a marketing move, in my opinion, to report focal lengths in equivalent focal lengths for 35mm. And that has led to some confusing distortions, this thread exposing several. It would have been much better if they reported the actual focal length, and then reported the relationship to the normal lens for that sensor size. Thus, the 5-15mm zoom on a digicam that is currently labeled as "5-15mm (35-105mm)" or, worse, "35-105 equiv.", would be reported "5-15mm (0.8 - 2.5X)", with X being "normal".

The irony is that the film format most challenged by digital cameras is 35mm, and 35mm's fall from the heights will be steep and fast. Quite soon, it will be relegated to the same niche backwater as medium format, and even medium format is on the run. It would be funny if all lens focal lengths were reported in their equivalence to a format that no longer exists, heh, heh.

Rick "who would buy a 30D if it had a plain 1Ds sensor in it, but who, until then, will stick with his 10D and film for larger formats" Denney

Steven M. Anthony

11th of April 2005 (Mon), 10:35

I can already see the limitations of my lenses in my 10D images--and some of my lenses are among the best made.

What sort of lenses and what sort of limitations?

rdenney

11th of April 2005 (Mon), 13:10

What sort of lenses and what sort of limitations?

Hmm. Well, let's see. The 50mm/2.5 Compact Macro is quite excellent for copy work, but at small apertures (f/22) diffraction is a problem. Enlargement enlarges diffraction effects just like everything else. I reproduced a photo album using that lens, but had to shoot at f/22 because I couldn't turn down my studio lights enough on the copy stand, and my setup was makeshift and I couldn't move the lights back any farther and maintain the proper angle.

And I can sure tell when I'm not using my 70-200/4L at optimum aperture (which is about f/11).

The Sigma EX 12-24 is one of those lenses like the talking dog: The miracle is not that he speaks with good diction, but that he speaks at all. So I don't expect perfection from such a radically wide lens, but I can still see its faults clearly. My example is glow-in-the-dark better than Michael Reichmann's (though I've caught him being unwilling to correct an obvious focusing error in one of his tests of ex-Soviet lenses, so my faith in his results generally has been shaken).

Here's my lens rundown for the 10D:

Sigma EX 12-24
Sigma 14/3.5
Zenitar 16/2.8 (this one is sharp!)
Canon 20-35 USM
Sigma EX 28-70/2.8 (a left over from film days--don't use it much any more)
Canon 50/2.5 Compact Macro
Pentax 50/1.4 Super Takumar (adapted)--as sharp as the Canon 50/1.4
Canon 70-200/4L
Jupiter 85/2 (okay, I'll give you this one, heh, heh, but at f/2 it does some nice things in the bokeh department)
Canon 135/2.8 Soft Focus (which, of course, can be turned off)

No, this isn't a list of the best primes ever made, but there are some lenses in this collection that perform as well as any made in their focal lengths.

But let's think of it another way: Theoretically, the maximum lens resolution that the 10D sensor should be able to record is in the range of 60-75 lines/mm. We are accustomed to thinking of that as not a particularly demanding performance level. But, in fact, it is quite demanding. Few lenses of any quality will do better except at their optimum apertures and with optimum technique (or, when tested on an optical bench, which is rather hard to take out into the field!). But if we had a 56x56mm sensor for my medium-format camera with the same number of pixels (say, 2600x2600), our limiting resolution would only be about 22-28 lines/mm. Most lenses, even real cheapies, will do that well at most apertures. All my silly Ukrainian junque can do that well by f/8, even the lenses I bought off ebay for $25. Yet I can enlarge the resulting image to a larger print with the same results. And a 56x56 sensor with the same pixel density as the 10D would allow a system that is utterly lens-controlled, just like the very best film. Of course, I'd only be able to store about 7 or 8 RAW images from that puppy on my 1-Gig CF card, heh, heh. But it would be worth it: I could make 32x32" prints with the same quality (from six inches) as the 8x10 from the 10D.

Rick "size matters" Denney

Steven M. Anthony

11th of April 2005 (Mon), 14:56

But let's think of it another way: Theoretically, the maximum lens resolution that the 10D sensor should be able to record is in the range of 60-75 lines/mm.

Ahem... Before I can translate your whole comment, help me out here. 60 - 75 lines/mm means what? That if you had 75 lines drawn in the space of 1 mm that you could resolve each seperate line? Squeeze more lines in and they run together?

slin100

11th of April 2005 (Mon), 15:51

Ahem... Before I can translate your whole comment, help me out here. 60 - 75 lines/mm means what? That if you had 75 lines drawn in the space of 1 mm that you could resolve each seperate line? Squeeze more lines in and they run together?
Roughly, yes. It's a value that characterizes the resolving ability of an imaging system. Technically, it's line pairs per mm, not lines per mm.

See these two links for information.
http://normankoren.com/Tutorials/MTF.html
http://www.luminous-landscape.com/tutorials/understanding-series/understanding-mtf.shtml

rdenney

11th of April 2005 (Mon), 16:08

Ahem... Before I can translate your whole comment, help me out here. 60 - 75 lines/mm means what? That if you had 75 lines drawn in the space of 1 mm that you could resolve each seperate line? Squeeze more lines in and they run together?

Yes. Remember those old test charts? They put parallel lines such that the thickness of the lines was the same as the space between them, with 75 of then showing up in a millimeter at the image plane. It's one of the basic measures of lens resolution. It's not as descriptive as a mean transfer function (which is a measure of what's there versus what the lens conveys--what does the lens transfer). The MTF incorporates a more rigorous method of measure contrast and its relationship to resolution. But it's relatively easy to compare lines/mm, and it's generally proportional to the MTF.

And yes, it's line pairs per mm, but that mostly reflects the question: Can you detect the space between two lines?

The 10D sensor has 133 pixels/mm, and you figure it will show half that unless the lines are at a perfect 45-degree angle. So, it's on the high side, if anything, and it's right up against the aliasing limit. There's all sorts of theory about how many lines/mm a given pixel array can theoretically detect.

Rick "who thinks the 10D sensor is good enough to detect lens quality in all but a few cases of the best lenses used optimally" Denney