Get close or Crop ?

[Stagecoach]

As a 'learning' macro shooter I constantly come across the DOF problems with my subjects. I'm fully conversant with the affect of aperture and camera to subject distance with respect to DOF but there are times when I just want more DOF.

One of the options available is to increase camera to subject distance (increasing DOF) and then to crop to achieve equivalent subject/frame ratio. The obvious concern here is that you gain DOF at the cost of pixels.

I intend to undertake some tests doing this to try and determine if the image quality loss with my rig is acceptable but would be interested to hear others views on this. Is it something generally done by macro shooters ?

I fully understand that such options as stacking and using a camera with a higher pixel count will help.

[PhotomanJohn]

Grahame - There is really no free lunch when it comes to DOF. The DOF "increase" gained by the greater distance will be lost when you crop in and enlarge the subject in post processing. As you mentioned, you may also limit the image quality because of the pixel count and requiring more definition out of the lens. This effect will vary depending on the equipment used.

My suggestion is to move in, use a higher aperture, get the important part of the subject in one plane parallel to the camera"s sensor and go for it. I am a big fan of focus stacking as you mentioned and would suggest that is one of the very best options for increasing DOF.

Have fun.

John

[shreds]

Undoubtedly there become a point where focus stacking is the only way forward.

http://www.cognisys-inc.com/home_cogn.php

[Stagecoach]

Hi John & Ian,

Thanks for the replies. I am aware of the options such as focus stacking, subject on the best plane etc. but am looking at the specific option of increasing subject distance and then cropping to achieve the same frame.

It's obvious that IQ is going to be lost but would a macro photographer who does not have the option at the time to stack, increase aperture or change camera rig consider moving out to get that bit of increased DOF if it was the only option ?

[Andrew76]

If light isn't an issue, change the aperture. I'm not a macro expert, in fact, I think you've actually given me advice in the past, Grahame, but why sacrifice IQ if you don't need to?

Of course, if light IS an issue, then you have your answer.

[DanK]

would a macro photographer who does not have the option at the time to stack, increase aperture or change camera rig consider moving out to get that bit of increased DOF if it was the only option ?

There is no one answer to that. It would depend on which macro photographer you asked, the pixel density of the sensor, the use to which the image would be put, etc. For example, because of differences in pixel density, a subject taken at MWD will have roughly 40% as many pixels on a 5DMKII as on a 7D. (The image would be the same size on both sensors, that is, 1:1.) So as someone who does a lot of macro, I would be less willing to do what you are asking about with the FF camera. If I were only going to show the image on the web, I would care much less about losing pixels by moving back than if I were going to print the image.

In practice, I have never moved out deliberately for greater DOF, but I have sometimes been forced to do it by circumstances--e.g., a critter that won't let me get close enough. However, it is an interesting question, because DOF does increase quite markedly as you move back from MWD.

[jcuknz]

"For a given image size DoF remains the same" YeahRight
Seems there are exceptions to the rule see posts 17 and 18.
Thanks Grahame

[PhotomanJohn]

Grahame - In thinking about your question again, I guess there is an improvement by moving back (while using the same lens). The DOF increases by the square of the change in distance while the image size decreases linearly. So even though there is a loss of apparent DOF when you have to enlarge the image more, the DOF has increased more than enough to compensate for that so you have a net gain in DOF.

I obviously had not thought about this before and thanks for the mental exercise. Let's see how many folks agree or disagree with this.

John

[jcuknz]

Grahame - In thinking about your question again, I guess there is an improvement by moving back (while using the same lens). The DOF increases by the square of the change in distance while the image size decreases linearly. So even though there is a loss of apparent DOF when you have to enlarge the image more, the DOF has increased more than enough to compensate for that so you have a net gain in DOF.

I obviously had not thought about this before and thanks for the mental exercise. Let's see how many folks agree or disagree with this.John

If you are correct with that then it explains why way back I felt I had achieved greater DoF with one shot I took wide and cropped despite the "Given Image Size" mantra which I had forgotten about at the time

"Dinner for Two"
My cat had just been sick and not to waste an opportunity!

[DanK]

The DOF increases by the square of the change in distance while the image size decreases linearly.

is that right? I have not worked through the math for calculating DOF, but I don't think this is right for image size.

Let's pretend that sensors are square, just to make this easier to type. Let a be the angle of view of a given lens, and let d be the distance to the subject. Let L be the length (or width) of the frame captured. Then I think that:

L=2d(tan a/2).

However, the area of the frame captured would then be A=L^2=[2d(tan a/2)]^2. So the area of the frame increases as the square of the distance, and the fraction of the frame occupied by the subject decreases as the square of the distance.

Am I missing something here?

In any case, DOF and number of pixels are not equivalent--increasing one won't compensate for decreasing the other. They have different effects on the image.

[PhotomanJohn]

Dan - You are correct that the area increases by the square but when I said image size I was referring to length not area. DOF is a linear measurement (front to back) and I believe varies with the square of the change in distance while the depth of the subject image varies directly with the change in distance to subject.

My thinking is that if we double the distance to the subject the depth of field increases by four times while the depth of the subject is reduced by half. If we then enlarge the image by two to get it back to where it was in the close shot we still have twice the depth of field. I gave it a quick try and when the two images were enlarged to the same size on my screen, the DOF was clearly greater with the shot at the longer distance. I learn something everyday.

John

[Stagecoach]

Hi Andrew,

Light or rather the ability to change aperture would not be an option as I had mentioned in post 4

[Stagecoach]

Hi All that have responded and thanks for the interest,

Firstly, let me explain what prompted this question. Last night I took the shots of the mantis under what were luckily pretty ideal circumstances to allow achieving the best IQ for the image with my equipment. In doing this I did have certain constraints which were that I had to use flash (night time) and I was unable to push my aperture above f29 (flash power) as at the ISO I chose was to restrict noise to what I consider workable.

On monitoring the images on the LCD of course they looked decent but after I had processed them I started wondering if there was some how I could have improved the DOF 'slightly' with the constraints I was working within. This prompted the specific question which was really "should I have taken some at an increased distance to improve DOF and would I have got an acceptable IQ after cropping them ?". Acceptable is a bit of a loose term and of course will vary between all.

I suspect that from the discussion replies one of the major considerations is going to be the use/size of the final image.

Perhaps one of the answers/comments should have been "why did you not take some shots at a further distance when you had ample opportunity and time to try this out". That's a lesson I have just learnt !!

[Andrew76]

Hi Andrew,

Light or rather the ability to change aperture would not be an option as I had mentioned in post 4

Missed that one!

[TonyW]

I had these problems yesterday trying to photograph a praying mantis preying. It had an insect held firmly in its claws. What do people think about the idea of increasing the aperture to get a better depth of field to the extent that the image is under exposed and then correcting that in PP? This would lead to a reduction in the effective bit depth but should not affect the sharpness.

I also have a problem in these circumstances where a tripod is difficult to use and I cannot stop myself from swaying backwards and forwards.

[Stagecoach]

Tony,

My first thoughts would be;

a) The affect of increasing the aperture depends upon which apertures you are using. Although the theoretical sweet spot for my macro is around f11 I'm perfectly happy to go up to f22 as I know that will give me acceptable IQ for images that are generally just for the web. I do not have enough experience to be able to confidently know how much IQ I would lose by pushing my lens to say f29 although I used this last night.

b) Underexposing is going to increase noise and when correcting (if not acceptable to you) in PP is then going to reduce sharpness to some degree.

Once again we have a 'balance' situation here whilst striving to achieve perfection.

As for swaying .......................... something solid to lean on I find is my only option

[Stagecoach]

Having sat and considered the replies and my original question I decided to undertake a test tonight.

Object; To take a macro image at settings around what I would often use and to then double the DOF by increasing the camera to subject distance only and then determine the loss of IQ when cropping to achieve the same image size.

The image on the left was taken at 0.45 mtr giving a DOF of 11.8 mm. The image on the right was taken at a distance of 0.61 mtr giving a DOF of 23.4 mm (twice that of the left image).

Post processing for each was ACR equal exposure reduction and no sharpening. In Elements capture sharpening on each full image at 40% at 4 pixels, brightness increase of 30% to each. Right hand image cropped to get same frame. Each resized to 550 pixel width with 100% output sharpening to left image and 150% to right image.

Here's the result.



There is a noticeable loss of IQ (that could possibly be corrected) in the centre of the flower but only noticeable comparing the two side by side. Would I do this for the flower shot, no, because in my view this had adequate DOF. Would I have done it for my praying mantis shot, yes because the increase in DOF would have outweighed the slight loss of IQ in that specific example I believe.

A worthwhile test from which I have learnt more about the limits of my equipment.

[jcuknz]

I thought somebody would do this test but they have not so here is mine ...
Lens wide open f/8 at 140 and left there for the 18mm AoV shot
Shot against the light, window behind table.
Seems to say something about the lens's IQ at 140 though I think for an ordinary shot that could be corrected in PP.
Perhaps I should repeat the test ZWMF instead of zooming.
The increase in DoF perhaps comes from operating the 14 AoV at f/8 instead of f/4.4
Probably not a good exercise but raises ingteresting questions about the matter and what affects what. With the flies I zoomed with my back not the lens.


Some time later zooming with five legs, my two and the tripods three, plus adjusting the height of camera to roughly match the angle of view for both shots and changing the lighting, with false attempt between above and below .... URRRGH

Whatever having done this test PhotomanJohn has convinced me that my earlier thinking was quite wrong.

Oh Dear! I forgot to add the distances ... left is 20cm from my two dioptre CU lens and the right is enlarged from 40cm
Working at 75mm or 150mm Angle of View with my MFT. Wide open at f/5.8 1/250 400 ISO

A clue to the answer is in the amount of enlargement I had to make to get same size images which was 150% not 200%

[xpatUSA]

About 2-1/2 years ago (before I knew Everything), I posted about a watch shot here:

http://forum.tz-uk.com/showthread.ph...Depth-of-Field

It may be of interest. The shots are more informative than works of art, though.

With some blurb and another image, it takes you from this:



to this:



Camera in question was a Nikon D50 armed with a Nikon 60mm micro-Nikkor.