In search of greater depth of field

[Jesse]

Hello,

I bought a Nikon E400 with APO objectives last year and I've been enjoying it. I spent a lot of time and money setting up a Nikon f-mount adapter for video and photography, then I spent some more time and money playing with the 4x APO objective on-camera, on the end of my 70-200mm zoom lens.

What I'm finding after all of this experimentation is that my depth of field is exceedingly narrow. I need to focus stack 100+ photos in order to get a decent image when the lens is on-camera.

Are there alternative lenses/objectives with similar magnification that will allow significantly greater depth of field?

Thanks. Still learning here.

[rjlittlefield]

Are there alternative lenses with similar magnification that will allow significantly great depth of field?

Not really. The physics of light -- diffraction -- imposes a strict tradeoff between DOF and resolution in a single exposure. If you go for greater DOF, which you can do by stopping down, then you get proportionally less resolution in exchange. The shallow DOF provided by a microscope objective is the unavoidable price you pay for getting the high resolution of that objective.

If you don't need the full resolution of your current APO objective, then you might use a smaller-NA objective of the same magnification to sacrifice a little resolution for say 2X improvement in DOF. But if you want a bigger change, you're stuck.

It may occur to you to wonder, then, if a different camera would help, perhaps because of smaller or larger sensor. The answer again is "no". The tradeoff between DOF and resolution is the same for all sizes of sensors, and for all focal length lenses.

There are a few advanced techniques for getting more DOF at same resolution, but they come at high cost and have tradeoffs of their own. Using shorter wavelength illumination can help, especially in the limit of say scanning electronic microscopes. But to use even UV light requires special lenses and sensors, must be carefully handled to avoid damaging your own eyes, and of course does not produce natural colors. Other purely optical technical techniques such as "wavefront coding" use carefully controlled aberrations and a bunch of post-processing, but they also add noise. In the end the only way to get the same image quality that you do from focus stacking, is to use focus stacking.

100+ photos for an APO objective at 4X is not unusual.

--Rik

[Jesse]

Here are some example photos I just took of a ladybug. I'm using an LED light panel for illumination.

1/160 F11 ISO 1250:


1/160 F8 ISO 500:


These are cropped to remove the majority of the image circle since this objective doesn't play well with full frame and my lens choice.

I'm getting some pretty impressive detail of the compound eye, but not even the entire eye is visible with this depth of field.

Alternatively, I see photos like this (not mine):



The above was taken with a MP-E65mm f/2.8 1-5x lens at f16 1/200 ISO 200.

It's a little difficult to tell given the resolution, but I feel like there is significantly greater depth of field in the above photo. More than just switching to f16 would give me.

I'm just curious if there is some way to get similar performance without having to buy a canon camera and lens combo.

[zzffnn]

That ladybug image was taken with good diffused flash lighting and at a different angle. Probably generously and carefully sharpened as well. Not really comparable to your LED lighting.

[Jesse]

That ladybug image was taken with good diffused flash lighting and at a different angle. Probably generously and carefully sharpened as well. Not really comparable to your LED lighting.

I was wondering about that. You think better lighting would make a huge difference?

Where can I find a tutorial or learn more?

[Chris S.]

Jesse,

It sounds as if you are changing the f/stop on your zoom lens. Is this really the case? If so, we need to clear up some confusion.

When you mount a microscope objective on a lens such as your zoom, you are using the zoom as a "converging lens" or "tube lens." Stopping down the converging lens does not change depth of field--it just causes the corners of your image to go black.

Not sure what the specifications are for your 4x apo, but if it is a 4x/0.20 (0.20 is the numerical aperture, or "NA"), it is equivalent to about f/10. This f/10 is invariant of the f/stop of your converging lens.

To stop down a microscope objective, you might try placing an iris as close as possible to the rear of your objective. This works well for some objectives, but I have no idea whether it will play well with yours, or perhaps cause vignetting.

That ladybug you compare to is reportedly shot at f/16, which of course has more depth of field (but less resolution) than f/10. This said, I suspect that the effective f/stop is even smaller than f/16. (I haven't shot with this lens, so someone who has, please confirm or correct me; but I imagine that it reports nominal f/stop, and adds some extension at higher magnifications.)

You think better lighting would make a huge difference?

More diffuse lighting would make a big difference in the aesthetic quality of the image, but would do nothing to increase depth of field.

--Chris S.

[zzffnn]

Good light (amount, discussion and angle) may help visualize more compound eyes. Flash should eliminate vibration too, if you have any.

Edit: sorry, I did not read carefully enough to notice that you were using microscope objective. You need to stop it down for more depth in that case, as Chris S said.

[rjlittlefield]

This post adds information to what Chris S has posted above...

It's a little difficult to tell given the resolution, but I feel like there is significantly greater depth of field in the above photo.

Yes, it is difficult to tell, and many people have been misled because of that.

The catch is that the other fellow's image of the ladybug contains little or no information beyond what can be seen in the <1MP image that is posted. If you had access to his full-resolution camera images, and tried looking close at them, you would soon discover that there's no point in looking closer because all you see is the same level of blur, made bigger.

At 5X and f/16, the MP-E 65 has an "effective aperture", on the sensor side, of f/96. On the subject side, that corresponds to using a 5X NA 0.026 objective. That's obviously quite a small NA, and it's that small NA that produces both the large DOF and the poor resolution.

I'm just curious if there is some way to get similar performance without having to buy a canon camera and lens combo.

Sure. You can stop down a microscope objective by cutting a piece of black paper to cover the rear of the objective, then punch a small hole in the center of the paper to let just that portion of the light through. Using your 70-200 zoom lens at 200 mm, effective f/96 will occur when the hole is 200/96 ~= 2 mm diameter.

If you do this, you'll probably be struck by how "soft" the image becomes. This is because along with strictly limiting the best resolution you can get, diffraction blurring also reduces the contrast of whatever fine detail is still resolved.

To compensate for diffraction blurring, it is routine to apply very strong digital sharpening. See http://www.photomacrography.net/forum/v ... 774#208774 and http://www.photomacrography.net/forum/v ... 198#203198 , and the larger threads that surround those particular posts, for some relevant discussion. Much of those discussions deals with a sharpening process called "deconvolution", but the main point for your purposes is that any image that has been severely blurred by diffraction both can and should be treated with a lot of digital sharpening to bring out what detail there is in it.

Alternatively, I see photos like this (not mine)

It's important to distinguish between the image that you're seeing, and the image that you might be imagining. As discussed above, the MPE-65 image would not get much more detailed if you could look closer. If you're imagining that it would, then first you need to adjust your expectations.

Then, if you do like that sort of image, to reproduce it you just need to get some optics that can be stopped down to the same extent. The old classic solution involved bellows and a reversed enlarging lens, and that still works pretty well for studio work. Or you can use the black paper disk, though that gets quite fiddly.

The great selling points of the MP-E 65 are only that it's built like a tank, adjusts from 1X to 5X at the turn of a ring, and has an automatic diaphragm so you can frame and focus at f/2.8 but shoot at f/16 an instant later at the press of a button. It has no magical means for somehow breaking away from the DOF-versus-resolution curve that is imposed by diffraction.

--Rik

[Jesse]

Jesse,

It sounds as if you are changing the f/stop on your zoom lens. Is this really the case? If so, we need to clear up some confusion.

When you mount a microscope objective on a lens such as your zoom, you are using the zoom as a "converging lens" or "tube lens." Stopping down the converging lens does not change depth of field--it just causes the corners of your image to go black.

I did not know that. I was wondering why changing the aperture seemed to have little if any affect though. Thank you.

[Macro_Cosmos]

Lighting makes the biggest difference. You can have the best equipment in the world and bad lighting which stems from inadequate skill will just ruin the image. It's totally normal for a stack to be 100+ exposures depending on how deep it is. I'd be delighted if I get to run a stack of only 100-ish, which then usually I find out that the stack isn't deep enough, there's many areas out of focus. It's always better to overdo it.

Both your photo and the example you show lack diffusion and strategic lighting placement. Simply bombarding the subject with light usually yields unsatisfactory results. You can see that from those tiny hairs in the sample photo you showed, there's not much detail there. Also, lots of hotspots are evident around the leaf and ladybug.

The easiest place to start is utilising household objects, usually found in the kitchen! A plastic circular container cut open and wrapped with tracing paper (or those transparent sheets for baking and steaming various Chinese treats) gives pretty good results. My equipment is quite elaborate and cost me a lot but my light shapers are very cheap -- jelly containers and plastic bottles. I swear by them, these "cheap" little modifiers are crucial to getting the result I like.

When you enter a supermarket looking at those funky containers, thinking about how they can improve lighting, you're on the way to getting what you want.

In my opinion, greater depth of field isn't needed, you just need a little more playing around, experiment with household objects, and you will be satisfied with the improvements. There's always improvements that can be made! Don't stop experimenting.

[Jesse]

Lighting makes the biggest difference.

Do you have some photos of your setup in use? I'm having trouble visualizing it.

Also, do you use a stackshot or other automated focus rail? I'm currently using a sunwayfoto mfr-150s and while it's better than handheld for accuracy, I'm finding it very tedious for 100+ photos.