Bit of a problem.

Earlier images, not yet re-categorized. All subject types. Not for new images.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Bit of a problem.

Post by DaveW »

I have been trying to photograph some insects hand held by available light but I can't quite seem to get enough depth of field to cover all of them, particularly getting the wings sharp. Do you think I ought to start using flash to freeze movement and be able to use even smaller apertures or what? Or will I then have diffraction problems?

Bee

Nikon D200, 70-180 Micro Nikkor @ 170mm, 1/160th second @ f22, ISO 400

Image

Hover Fly

Ditto, 1/160th second @ f16

Image

DaveW

rjlittlefield
Site Admin
Posts: 24018
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: Bit of a problem.

Post by rjlittlefield »

DaveW wrote:I have been trying to photograph some insects hand held by available light but I can't quite seem to get enough depth of field to cover all of them, particularly getting the wings sharp. Do you think I ought to start using flash to freeze movement and be able to use even smaller apertures or what? Or will I then have diffraction problems?
Dave,

At 1:1 and f/22 on a D200, you're already taking a hit from diffraction at the actual pixels level. See http://www.janrik.net/insects/ExtendedD ... deoff.html for illustration.

The shots you've posted here are not so high magnification and/or they're only web resolution. Both aspects let you get away with smaller apertures. But the effects are not easy to predict and there are matters of taste involved. My guess is that you're on the edge.

I suggest running a test series using a dead insect. Then you can directly compare the photos to see how the DOF vs sharpness tradeoff works for your equipment and your eyes.

--Rik

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

Thanks Rik,

I presumed that the limited resolution of a computer monitor would pretty well obscure diffraction effects at this magnification. Different story if I wanted to print them out though.

I have just had a look at your article, but find it a little difficult to follow as there seems to be no explanation as the what the two different lines of butterfly wing sections relate to? Both cover the same aperture range but one set seems a bit fuzzier than the other and both have an F8 + EDOF sections to them. It's probably just me being thick, but what is supposed to be the difference between the two sets of sections as the article does not make that clear to me?

Not counting the f8 + EDOF section, looking at the top set the f22 - f32 sections are sharpest on my screen throughout their full width. On the bottom set it's f11 - f16 that are the sharpest throughout their widths. I realise this might not be true of prints because the screen has less resolution than a print, but as most images will never be printed out these days but used on screen, or projected, perhaps a revision of diffraction limits for screen and projection use is now needed because it is pointless wasting DOF in producing a higher resolution than the medium they are being reproduced on can show?

Look forward to your answer.

All the best,

DaveW

rjlittlefield
Site Admin
Posts: 24018
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Dave,

The key to the illustrations at my link is this one sentence in the first paragraph under "Illustration": A single full frame is shown at right; crops showing more detail appear in the series below.

The first set of strips is reduced to about 1/3 size, versus what came out of the camera. Every pixel in that set of strips represents a 3x3 block of pixels from the camera. This is equivalent to full-frame being 1024 x 683 pixels.

The second set of strips is actual pixels as they came out of the camera, corresponding to full-frame being 3072 x 2048 pixels.

On my monitor (an LCD screen), the images look like this: The upper set of strips looks almost equally sharp at f/22 and f/16; the f/32 strip is detectably fuzzy. The lower set of strips is equally sharp at f/8 and f/11; the f/16 strip is detectably fuzzy. "Detectably" means that I can see fuzziness if I look closely. It takes about 1 stop farther for the fuzziness to be immediately obvious.

I agree that one should think about intended use when picking an aperture. If I'm only interested in 1024 x 683 with the illustrated setup, then I should shoot f/22, maybe f/32 if DOF is important. But if I'm interested in printing, or to allow zooming in, or to display on larger monitors, then I should shoot f/11 to get all that this lens can capture.

As for revising the "limits", well, I don't know any simple limits. The relationships between magnifications (lens and total), sensor size, f-number, required resolution, and DOF are complicated, and there's no way to get around that. The allowable f-numbers on my two Canon cameras are 3X different (corresponding to the ratio of the sensor sizes). The allowable f-numbers are different by roughly a full stop between some Canon and Sigma lenses (due to differences in how the lenses focus and stop down). Getting 3:1 by bellows extension is wildly different from getting 3:1 by sticking a reversed 50 mm in front of a 150 mm. If anybody said simply that "f/11 is best", they would have failed to specify a half-dozen things that matter, and some of them matter a lot.

Quite frankly I'll be happy if most people get to the point of understanding that "Too small of an aperture can make things fuzzy. Run your own tests and see what works best with your equipment and your intended use."

--Rik

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

Dave and Rik.. remember that (unlike Canon) the Nikon D200 and that lens will display the actual "effective aperture" that results from focusing close. So an indicated f22 would really not be a big deal in the diffraction/DOF trade-off needed for this type of shot.

Dave, as Rik mentioned with digital it's very easy to set up a quick test of a small detailed subject and determine what apertures you can live with.. strongly recommended.
Last edited by Charles Krebs on Sat Apr 14, 2007 8:00 pm, edited 1 time in total.

rjlittlefield
Site Admin
Posts: 24018
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Charles Krebs wrote:Dave and Rik.. remember that (unlike Canon) the Nikon D200 and that lens will display the actual "effective aperture" that results from focusing close. So an indicated f22 would really not big a deal in the diffraction/DOF trade-off needed for this type of shot.
I did not know that!

Do I understand correctly, then, that the Nikon pair working at 1:1 would display "f/22" when using the same actual aperture that my Canon would call "f/11"?

If so, then add one more factor to the list of things that matter a lot.

--Rik

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

Do I understand correctly, then, that the Nikon pair working at 1:1 would display "f/22" when using the same actual aperture that my Canon would call "f/11"?
Yes... assuming the lens used actually "loses" two stops going to 1:1 (the 70-180 doesn't quite go to 1:1). This is if you "lock" the aperture dial on the lens and set the aperture via the camera body. (This is not possible with old lenses that do not have the row of electronic contacts).

Mike B in OKlahoma
Posts: 1048
Joined: Fri Aug 04, 2006 10:32 pm
Location: Oklahoma City

Post by Mike B in OKlahoma »

Interesting....I guess the electronics in the camera use the focusing distance and the focal length of the lens to come up with a magnification to plug in and get the effective f/stop? WOW!

My answer to your dilemma is "Remember [former US President] Theodore Roosevelt!"

Nothing to do with politics, but Theodore is quoted as saying:

"Do what you can, with what you have, where you are."

With the equipment I have, the subject I have, and the image I want to capture, I do what I can. Most of the time this means using flash, and simply because I get into the habit of using it, I suspect this means I use flash sometimes when I really could take a natural light shot--Often much preferable in good light.

Sometimes it means shooting in mid-day when you normally wouldn't use photography, but using the bright mid-day light, supplemented by flash to fill in the harsh shadows, and stopping way down.

In less desirable light, it may mean cranking up my ISO higher than I'd choose to.

For a really spectacular subject, if time allows, I will often try to get it more than one way, so that if one technique fails for a reason I didn't foresee, maybe the other technique will work. In practice I seldom have that much time for unusual shots, but occasionally it happens. More often, I do have time to try it several times using the most obvious method, so that if I bobble one due to my movement, subject movement, mis-focusing, etc., I have a chance to get it again.

It also can help in terms of maximizing DOF to carefully consider where my DOF is relative to the interesting part of the subject--Be really sure that you aren't wasting DOF on empty space or an uninteresting part of the subject. What I've read, and experienced is that at macro distances, the sharp area in front of and behind the subject is about equal, unlike more conventional photography distances where the rule of thumb about 1/3 in front, 2/3 behind works reasonably well. Rik and Tom can probably mentally calculate the DOF for their f/stop and magnification, but I can't do that! :lol:

Finally, remember that it's better to get the shot correctly at "inferior" quality due to high f/stop or high ISO, than to get a blurry mess because your shutter speed was too slow. Ideally I'd like every one of my shots to be suitable for making into a ten foot poster, but I can't always do that!

Oh, and most important of all--Don't necessarily pay too much attention to people who pontificate on internet photography forums! :wink: (especially if they quote dead Presidents)
Mike Broderick
Oklahoma City, OK, USA

Constructive critiques of my pictures, and reposts in this forum for purposes of critique are welcome

"I must obey the inscrutable exhortations of my soul....My mandate includes weird bugs."
--Calvin

twebster
Posts: 442
Joined: Wed Jul 26, 2006 8:02 am
Location: Phoenix "Valley of the Sun", Arizona, USA

Post by twebster »

Hi ya' Dave, :D

Regardless of the actual DOF you have to work with, it is absolutely imperative that you position the camera parallel to the subject plane and features you most want in focus. I say this after viewing the image of the syrphid fly. The body and wing on the right side of the fly are sharp but the left wing is OOF. However, look at the flower petal behind the left wing. It is nearly sharp. It appears to me that your plane of focus was tipped slightly to the left. When shooting macro misalignments of millimeters (or less) can place important subject parts outside the DOF.

Best regards, :D
Tom Webster

Phoenix "The Valley of the Sun", Arizona, USA

The worst day photographing dragonflies is better than the best day working! :)

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

Thanks all very useful information.

Thanks Charles, that explains something I noticed yesterday, that though I had my 70-180 Micro Nikkor mounted on the camera having the smallest marked f-stop on it's aperture ring at f32, I could go down further to f45 on the camera top LCD and in the viewfinder using the cameras sub-command dial when the lens was at it's maximum magnification.

I asked a fellow cactophile friend of mine who for his job is a Professor of Geometric Computing at a British University what was the maximum resolution a digital projector could produce or that my computer monitor could show and got this answer:-

"Few digital projectors will do a resolution greater than about 1280
pixels wide and somewhat less high. Even allowing for future
improvements, there's not much point keeping more than 2M pixels, I
think - for projection /monitor."


As I say, I doubt I will ever wish to print out my images and I think printed images will decline in future in favour of electronic presentation anyway, just as with film v digital, then DOF v diffraction may move more towards DOF and diffraction that can be seen on a screen/monitor rather than that presently required for printing.

Even the requirements regarding illustrations for "books" may change as conventional colour printing becomes dearer v. distribution by electronic means. Just look at the cost of a full set of printed Encyclopedia Britannica now against the whole edition on CD/DVD. Plus such printed editions are out of date the year after they are published, electronic ones could be constantly updated almost as your virus protection software is done now!

One further non-related question for my own interest. In the days of the Nikon F2 all camera manuals used to specify the type of metering cell used e.g. Silicon Blue, Gallium Arsenide, Cds etc. Digital cameras don't seem to do this anymore, at least I cannot find it in my Nikon D200 manual. What cells does Nikon use for it's Matrix Metering and also Canon for it's similar systems, anybody know?

DaveW

beetleman
Posts: 3578
Joined: Fri Aug 04, 2006 4:19 am
Location: Southern New Hampshire USA

Post by beetleman »

Well, let me say one thing. When I first looked at the pictures, I thought to myself, These don`t look to bad for DOF...they look rather normal to me. The fly is awesome Dave. If you take any kind of angle shot of an insect, some part of it will be OOF (unless you stack). I still don`t understand some of this DOF information, but I find whenever I read it, I pick up a little more information every time it is discussed...thanks everyone
Take Nothing but Pictures--Leave Nothing but Footprints.
Doug Breda

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

As Tom says the closer you get Doug the more you should get the insect in "plan view" to minimise DOF problems. That is one of the problems with larger insects like dragonflies taken at an angle in trying to get the wings in focus as I understand it.

Others will correct me if I an wrong, but DOF does not really reduce proportionately as you get closer. Further away you can take an object shaped like an American Football (or Rugby ball) end on and get the whole length in focus. As you get closer it has to be spherical shaped like a Soccer ball to cover it with the DOF, and closer still like an American Football or Rugby ball standing on end for all of it's depth to be in focus.

Quite a problem with things having depth as you get closer.

DaveW

rjlittlefield
Site Admin
Posts: 24018
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

A potpourri of thoughts...
DaveW wrote:Others will correct me if I an wrong...
Well of course, Dave -- we're your friends, and what are friends for if not to keep one on the straight and narrow path... :wink:
...but DOF does not really reduce proportionately as you get closer.
It is more correct to say that DOF does not reduce proportionately with magnification. Actually it reduces proportional to magnification squared. If you change the magnification by 2X, DOF shrinks by 4X.

Suppose you have a bunch of subjects that are the same shape but different sizes, and you photograph them at magnifications appropriate to their size, so that their images all end up being the same size. Then the DOF, expressed as a fraction of the subject's total depth, will be proportional to the subject size. Spherical beach ball, you can get the whole thing sharp; spherical pollen grain, you can get only a tiny slab in focus at once.

To a very good approximation, distance (being closer) does not affect DOF. Shoot the same subject at 1:1 with 50mm and 150mm lenses, and you'll get the same DOF at the same f-number, despite being 3X closer with the 50mm lens.
Few digital projectors will do a resolution greater than about 1280 pixels wide and somewhat less high. Even allowing for future improvements, there's not much point keeping more than 2M pixels, I think - for projection /monitor."
Well, that one fact is true, but there are other facts and other viewpoints.

Projectors do seem to be stuck at 1280 pixels wide. Why that is, is a bit of a mystery, considering that...

The HDTV specification 1080i calls for 1920 x 1080 pixels. And it's no problem actually getting that resolution today in upper-end monitors and laptops (monitor spec WUXGA). Very large monitors such as the NEC LCD5710-B are quite expensive ($12,000 for 57" diagonal), but smaller monitors are quite affordable -- consider the Dell 2407WFP, 24", 1920 x 1200, selling today for $609.

It sounds to me like your professor friend is thinking about classroom presentations and public lectures. His answer is very appropriate for that situation -- particularly considering that folks in the back of the room won't even be able to resolve the full 1280 pixels!

I generally think in terms of a viewer who has his/her own monitor and maybe even wants to zoom & pan to look closer. That situation argues for rather more pixels.

You'll have to decide for yourself what your target audience is.
Quite a problem with things having depth as you get closer.
Yea, verily! :D

--Rik

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

Thanks Rik,

"Consider the Dell 2407WFP, 24", 1920 x 1200, selling today for $609"

I'm reading you on that very monitor Rik!

"It sounds to me like your professor friend is thinking about classroom presentations and public lectures. His answer is very appropriate for that situation -- particularly considering that folks in the back of the room won't even be able to resolve the full 1280 pixels!"

He knows that was the very type of situation I would use my images for as he does the same thing himself. Plus he views his images on a 30" monitor because the education department probably buys his gear!

DaveW

rjlittlefield
Site Admin
Posts: 24018
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

OK, that about nails it.

I remember an instructor, decades ago, telling us about presentations:
No matter how big your slide is, the person in the back of the room will see it like a notecard at arm's length. Plan accordingly.
I just did the calculation. Turns out, a notecard at arm's length is equivalent to well under 600x1000 pixels.

"Plan accordingly", indeed!

--Rik

Post Reply Previous topicNext topic