Sensor size and macro photography

[Ken MT]

I have both a Canon 30D and a 5D mk 11. With a 105mm 2.8 mm Sigma macro lens, am I better using the 30D which has a longer focal length, but shallower DOF, or the 5D which has a shorter focal length but will allow me to crop harder. I attach a photo of a spider that was about 20mm in length taken with the 5D, using a 70mm extension extension tube with the macro lens at 1:1. (1 sec at f16)

[Glenn NK]

Ken:

A to the different focal lengths of the bodies - only the lenses have focal lengths. What happens is that the lens projects the same image to both the 30D and the 5D sensors, but the larger sensor of the 5DII captures a wider and higher area of the incoming light.

As to which body to use, I am faced with the same choice. I posted something like this elsewhere:

My 5DII has 21 MP, and I use the 100mm macro lens on it. If I keep the same distance from the subject (I don't want to put the lens on top of a flower), the resulting MP left from the 5DII is 5/8 x 5/8 x 21 = 8.2 MP - pretty close to the 8.1 of the 30D. Of course this isn't the whole story - the DOF of the FF 5DII is less than that of the 30D. So for macro work, the 30D is an advantage.

However in order to utilize all the pixels of the 5DII, I must either use a longer lens (expensive and achieve less DOF), or move in closer which in addition to the nearness being problematic, I achieve less DOF (noted above).

In reality what I do is use a single extension ring (12 mm), and move in a bit closer with the 5DII. The thrill of using the 5DII hasn't worn off yet, but it will.

However, there is an advantage to the 5DII with live view - by zooming in 10x, the slightest movement of a flower can be detected, so I wait until the breeze drops. With the 30D, it's guesswork.

Glenn

[revi]

Ken:
(...)
As to which body to use, I am faced with the same choice. I posted something like this elsewhere:

My 5DII has 21 MP, and I use the 100mm macro lens on it. If I keep the same distance from the subject (I don't want to put the lens on top of a flower), the resulting MP left from the 5DII is 5/8 x 5/8 x 21 = 8.2 MP - pretty close to the 8.1 of the 30D. Of course this isn't the whole story - the DOF of the FF 5DII is less than that of the 30D. So for macro work, the 30D is an advantage.
(...)
Glenn

Why, in this situation, would the DoF be less for the 5DII than for the 30D ? You project the same image from the same lens at the same size on the sensor. You are in effect treating your 5DII as if it is a cropped sensor camera (that the crop is done in hardware in one case, and in software in the other makes no difference).

Remco

[rob marshall]

You guys are confusing me. I always thought that the actual DOF is the same on full-frame and crop-factor cameras. If you took an image on each camera with the same lens at the same distance, the DOF would be exactly the same. The field of view would be different, but not the DOF.

DOF is the portion of the lens output that is in focus and sharp, not the relative percentage of the FOV that it occupies.

[Dave Humphries]

You guys are confusing me. I always thought that the actual DOF is the same on full-frame and crop-factor cameras. If you took an image on each camera with the same lens at the same distance, the DOF would be exactly the same. The field of view would be different, but not the DOF.

DOF is the portion of the lens output that is in focus and sharp, not the relative percentage of the FOV that it occupies.

Hi Rob,

Isn't it related to circles of confusion, which do change with sensor size? (not sure)

I haven't got it fully worked out in my head either, but it is my firm belief that the effective aperture as it pertains to DoF is affected by the camera's crop factor, just as we commonly perceive focal length to be.

So with the lens set to f/8; on the 1.6 crop 30D, you get the equivalent DoF as when using f/13 (or f/12.8 to be more precise) on the full frame (5D Mk2) camera.

Cheers,

[revi]

I think the confusion comes from using different approaches.

DoF will change with sensor size in the 'normal' case where we want the same image occupying all the available sensor space. So the actual size of the image changes according to the crop factor. As we want the exact same image for crop and full size sensor, we need the same Field of View, implying a longer focal length for the full size sensor. And the longer focal length gives a smaller DoF. (note that here we need to actually change the focal length)

In the case above (posts # 1-3), the case is different: we use a fixed focal lenght and a fixed distance between camera and object. Then the image from the full size sensor is cropped to the same size as the crop sensor (actual physical dimentions!). So the 'native' FoV for the full frame might be larger, the effective FoV we end up using is the same for both sensors.

I hope my explanation is clear enough, I'll try and add a figure later tonight

Remco

[rob marshall]

I think the confusion comes from using different approaches.

DoF will change with sensor size in the 'normal' case where we want the same image occupying all the available sensor space. So the actual size of the image changes according to the crop factor. As we want the exact same image for crop and full size sensor, we need the same Field of View, implying a longer focal length for the full size sensor. And the longer focal length gives a smaller DoF. (note that here we need to actually change the focal length)

In the case above (posts # 1-3), the case is different: we use a fixed focal lenght and a fixed distance between camera and object. Then the image from the full size sensor is cropped to the same size as the crop sensor (actual physical dimentions!). So the 'native' FoV for the full frame might be larger, the effective FoV we end up using is the same for both sensors.

I hope my explanation is clear enough, I'll try and add a figure later tonight

Remco

I just did a test. I used my Sigma 105mm macro lens set at f/2.8. I mounted it first on my Canon 50D (crop factor) then on my Canon 5D (full-frame). My subject was a plastic measure so I could get the measure lines. The cameras were on a tripod, and I didn't move the distance between lens and subject, and I tried to focus on roughly the same spot of the measure. I also used the 5D a second time and moved it closer to get the same framing as the 50D. Here are the full shots, straight out of the camera.

50D full shot


5D full shot


5D moved closer to fill the frame same as 50D


I then edited each shot, rotated them for ease of reading, and did a screen grab of the main focus area so you can see clearly how much is in focus. Any thoughts of the differences - if there are any?

50D partial view


5D partial view


5D partial view, but framed as a 50D shot (moved camera closer)

[revi]

Let's have a shot

3rd one (camera moved closer) clearly has a lower DoF (for me): 9-10 mm, against 12-15 mm for the others.
The other two are virtually the same, perhaps slightly more DoF in the second, but do I detect traces of sharpening there?

Remco

[Glenn NK]

I'd like to add a correction to my first post.

With both cameras using the same lens, and both cameras mounted the same distance from the subject, the 30D has a slight edge in DOF.

Check it out for a 100 mm lens @ f/4 and a subject distance of 20 cm (200 mm approx 8 inches):

http://www.dofmaster.com/dofjs.html

30D DOF = 0.03 cm = 0.30 mm
5DII DOF = 0.05 cm = 0.50 mm

The 5DII has a wee bit more DOF, but neither is significant. 0.30 or 0.50 of a mm is virtually nothing at all (I do a lot of up close flower photography, and 1/2 mm is pretty useless in terms of DOF).

However (as I stated) in order to NOT lose the 5DII's pixels that occurs when cropping, the 5DII must be moved in closer. This is where the DOF is lost - the closer the camera is to the subject, the less will be the DOF.


Let's put the cameras so they have the same field of view (using the same lens of course):

Subject distance = 20 feet for the 30D
Subject distance = 5/8 x 20 ft = 12.5 ft for the 5DII:
Both using the 100 mm lens at f/11:

30D DOF = 5.24 ft
5DII DOF = 3.2 ft

It can readily be seen that distance is the major player in this situation.

Of course we could put the 5DII at 20 feet, and crop to get the same image, but we'd then be using only 5/8 x 5/8 x 21 MP = 8.2 MP.

Glenn

[rob marshall]

Let's put the cameras so they have the same field of view (using the same lens of course):

Subject distance = 20 feet for the 30D
Subject distance = 5/8 x 20 ft = 12.5 ft for the 5DII:
Both using the 100 mm lens at f/11:

30D DOF = 5.24 ft
5DII DOF = 3.2 ft

It can readily be seen that distance is the major player in this situation.

Of course we could put the 5DII at 20 feet, and crop to get the same image, but we'd then be using only 5/8 x 5/8 x 21 MP = 8.2 MP.

Glenn

Yes, distance is the factor. If you shot from a longer distance, everything in the frame might be at infinity whatever the camera/lens, and everything would be in focus. But the closer you get to the subject the less is in focus.

Can we say then that if you have a crop-factor camera and a full-frame, but only one macro lens (say 100mm), you are better off using the crop-factor; because to get the same subject framing for the FF you would need to move in closer, and by doing so would lose some DOF over the crop-factor body?

I also have a Panasonic G1 (four thirds) which is a x2 crop. I've noticed that the DOF on that is often better on that than the Canon 50D (although the lenses are obviously not the same) Surely it's the end results that count, not how they are achieved?

[Ken MT]

OK, now which camera body do you use, or is it a case of taking both?
Have just found a good, but complicated, article in Tutorials on macro lenses

[hjphotog]

Frankly, I don't have enough fingers and toes and patience to figure out all that math and calculation. I shoot mostly living things that won't stand still long enough. My litmus is how does something look in the viewfinder, in live view and on an editing screen. And with digital, I've found a 7-inch hdmi remote monitor is better than the f-stop check button on my camera. I use my 7D for macro when my subjects only sit a second or two -- bumble bees, nervous insects; also it's great for hand-held closeup work, even with manual focus.
The 5DII does a splendid job when the subject is cooperative.
But my secret is that the farther away I set up my tripod, the more depth of field I get. Right now, me and another codger photographer are experimenting with extensions on some really big lenses which creates long distances from the subjects and some very deep depth of field. I got some GREAT macro shots using my 500 F4 with extension rings. The problem was every piece of dust, every gnat, every piece of pollen on the focus plain was in the shot. We're working that out. But not yet ready to go public.
Soft flash, fill flash helps. Bracket focus and bracket exposure. Angle of the sun is essential. In other words, just like taking any other shot.
That's the beauty of high-res digital: taking a dozen to two dozen shots for one subject doesn't use up $100 worth of film.
I've used this for close-up and macro for years and even taught it in college. My students end up with crossed eyes and question marks above their heads when I explain the calculations for settings. But showing them what to do with the camera and how to select lenses , extension tubes, lighting conditions and patience and they turn out very cool work.

[Arlen]

Where this choice really matters most is at the minimum focus distance of the lens. Both 35mm cameras will be the same distance from the subject, so both will have the same DOF. But the small sensor camera will yield an image that looks more magnified, due to it's restricted field of view. To get that same field of view from the full-size camera, you can crop the image in post processing.

Then it becomes a matter of megapixels. If the full-size camera has sufficiently more resolution, it's cropped image will still have more pixels than the uncropped image from the small sensor camera. If it doesn't, then the small sensor camera wins the megapixel war.

Of course this comparison only considers depth of field and resolution. There are other differences between the cameras that may well be decisive.

On the flip side of this, if you don't need to be at the minimum focus distance to get the magnification you want (seldom the case for me): then at a similar field of view in-camera, the small sensor camera will be a greater distance from the subject, and so have greater DOF.


Edit: Most of this has been said in one form or another above. I'm not sure if the way I have explained it helps make it clearer or not.

[Colin Southern]

Where this choice really matters most is at the minimum focus distance of the lens. Both 35mm cameras will be the same distance from the subject, so both will have the same DOF. But the small sensor camera will yield an image that looks more magnified, due to it's restricted field of view. To get that same field of view from the full-size camera, you can crop the image in post processing.

Assuming equal megapixels, the crop-factor camera wins at macro if DoF is what's needed.

For the same FoV, the crop-factor camera can be positioned further from the subject and will thus have the greater DoF.

One can of course crop the image from a FF camera positioned at the same distance from the subject as the crop-factor camera (or in fact from any distance that gives the required DoF), but one also needs to remember that in doing that, one typically throws away a LOT of pixels (typically over 1/2) (thus putting the FF camera at a significant disadvantage to the crop-factor camera).

I think that a lot of folks tend to overlook how densely packed the pixels are on most crop-factor camera.

[Hero]

Assuming equal megapixels, the crop-factor camera wins at macro if DoF is what's needed.

But then again with an equal amount of pixels the FF sensor's ones are a lot bigger. And big pixels translate to larger f-numbers before diffraction sets in. I'm not enough of a maths-head to tell you the precise numbers, but it should compensate at least a bit.

[Colin Southern]

But then again with an equal amount of pixels the FF sensor's ones are a lot bigger. And big pixels translate to larger f-numbers before diffraction sets in. I'm not enough of a maths-head to tell you the precise numbers, but it should compensate at least a bit.

Probably, but I'd rather deal with diffraction issues (pretty much a non-issue with some appropriate sharpening) than lack of DoF.

[thequacksoflife]

curiously some of the best macro's I ever saw were done using a panasonic Fz20 with a macro attachment. the small sensor of the bridge camera being a real help the photographer told me......