Wednesday, July 22, 2009

And the Ribbon for Best Lens Goes to----

---I can't make up my mind.

When I started this blog entry several days ago I was certain I knew the answer.

I wrote: "---the 35ish year old Canon f1.4 50mm lens--even if it did trip me up a few times before I figured out its quirks. "

I had traced the most confusing of the quirks to a little catch hidden under the lens's locking ring. If I accidentally twist the locking ring and release the catch, it limits how far the iris can close. The iris adjusting ring turned freely and it might say f16--the limit on the lens-- but the opening stops at where the iris happens to be when I accidentally trip the catch

While I'm sure this enables some neat function when the lens is mounted right side up on a proper camera, in macro work the depth of field--the height where everything is in reasonable focus-- is proportional to the iris opening, So why did one set of images look sharp when the next set look so... You get the idea. A strip of tape fixed that problem but it took a while to figure out what was going on.

A second quirk was caused by a second catch. It is spring loaded and keeps the iris open on my old FTb Cannon camera until I press the shutter. In theory, easily fixed by a rubber band. In practice, not so easily fixed since there was no good place to hook the other end and still be able to adjust the stage up and down. Took a trip to the hardware store before I could kluge up an attachment rod that moved up and down with the stage.

That is where I ended several days ago. I would have have written a few more lines and published the blog if there hadn't be one other mystery to work out. The Canon lens imaged 4.5 mm on the 24mm long sensor for a magnification of 5.3X while the Nikon lens imaged a little less than 6mm for a magnification of 4.1X. Another clear advantage to the Canon-- but why? Both were 50 mm lens. Using geometric optics formulas it shouldn't have made any difference--50mm is 50 mm period.

But real world didn't line up with theory. Which meant it had to have something to do with the principle planes.

What are principle planes? you must be asking.

Today you would use a computer to trace all light rays needed to calculate a magnification through the series of glass (and sometimes other materials) lens elements that make up a camera lens. Before computers and with more effort you could do the same calculations with a hand calculator. Before calculators and with much much more effort you would use a slide role--the precision machined five foot long version that came complete with its own magnification optics so you could extrapolate between the lines for the necessary accuracy. Before that you would have used a book filled with tables of sines, cosines, tangents and logarithms, plus paper and goose quills.

Sometime near the end of the goose quill era someone came up with the idea of principle planes. With them you could use the simpler formulas of geometric optics that Issac Newton and Carl Friedrich Gauss devised--formulas showing where you would find an image of a candle flame and how big it would be.

With a single symmetrically convex spherical lens-- one with the same amounts of glass ground off on either side and one of the easiest lens to grind--the two principle planes are together in the middle of the lens. With a plano convex lens with one flat side and one strongly curved side, the planes separate. One is against the flat side; the other is at the tip of the curved side. Add another lens element to make a doublet and you can calculate another set of principle planes. Add a third element... and so on. Add certain combinations of lens element and the principle planes don't even end up near any of the glass elements. Like in a telephoto lens. One of its principle planes to floating in space somewhere in front of its lens barrel.

At this point, "So what" and "Why should I care," are valid thoughts.

In normal photography you wouldn't care--the reason you won't find principle planes in the index of most how-to-photography books.

But with micro photography--bigger than lifesize photograph-- it's different. Take a lens. Measure the distance from its front principle plane to your bug's eye--z mm. For a magnification of 10X do the simple calculations. Distance from from the second principle plane to the camera ccd =10z. Oops. Your 300mm telephoto lens is not going to work. 10z puts your camera through the ceiling and into the girl's upstairs bedroom.

But enough with the handwaving theory and back to real world.

I worked out the position of the principle planes of the Nikon 50 mm using the fact that for 1/1 magnification the distance between the object--a machinist rule-- and the sensor must be 4X the focal length (8 inches) plus any separation of the principle planes. Measured out to be 8 1/16 inches. That lens is clearly symmetrical with the principle planes at the center of the lens

With the Canon lens it was different. One principle plane was close to mounting flange, the other plane was about a inch farther up in the lens. Another clear advantage for the Canon lens--20% more magnification. But during these tests, I came to realize what I took for lower lens resolution was less than perfect focusing. Three dimensional flowers might be more photogenic than lens charts but they aren't ideal for resolution comparisons.

A super cheap resolution chart would be a fresh flower pedal with interesting detail held flat to a card with double sided sticky tape.

But first I had to work out a major problem with the Nikon lens--a major lens flare in the center of the image. How had I managed to miss that earlier?

Turns out that earlier I'd mounted the camera, extension tubes, bellows and lenses in that order. This time I had the bellows above the extension tubes. When I flipped them the glare spot disappeared. Bellows folds make excellent baffles for stray light.

Summery of what I've learned-- I better have my technique down pat before I blog about lens comparisons. And good macro/micro photograph is both science and art.

Where do I go from here? Despite how it might look, I dud take some non-macro pictures recently. Like about ten gigabytes worth this weekend. Butterflies and kids looking at butterflies at Olbrick Gardens on Friday. Circus World and a Wild West reenactment on Sunday. RAW images and burst modes can fill up memory cards quick.

So it might be a while before I'm blogging again about macros. But it will happen.

Sunday, July 19, 2009


The reversing ring showed up in the mail Saturday. This ring screws into the lens's filter threads and mounts it backwards--something that according to the books should improve the image quality. Since I am pushing the resolution of the 50mm garage sale lens I shall have to do some experiments at lower magnification to test that out.

At 5X magnification, five millimeters of the iris adjustment mechanism is focused down on to ~3800 pixels. Since it takes two pixels minimum to resolve a feature this works out to a theoretical resolving power of 380 line elements per millimeter. And since the numbers I've seen posted in the newsgroups and forums for camera lens resolutions range from 40 to 100 lines elements per millimeter, the slight blur in the image I worked to get rid of wasn't caused by my less than perfect focusing. At 5X magnification the 10 megapixel sensor has far more resolution than the lens.

One test for this problem is how much sharpening is needed. On this image I needed a 2.8 pixels radius and a strength of 400 using Adobe Camera RAW. Such grossly overshapening would have totally destroyed a normal image.

Another challenge was illumination. For maximum DOF I shot a f22, the smallest iris opening on the lens. On top of this you add in an exposure factor of 36--the (magnification pus one) squared. I had to clamp my flash 6 inches away from the mechanism and run it a full power to have the needed light. With a 150 watt halogen lamp almost as close--my first attempt--I was so light starved I couldn't see anything in the viewfinder at f22.

The vitals for this image are:

50 mm lens with an extension of 125mm--the maximum the bellows will allow
5 mm by 3.3 mm image area
4 mm from out-of-focus bottom plate to the plate stretching the spring.
ISO 200 at 1/160 sec.
f22 with an exposure factor of 36.

Thursday, July 16, 2009

The Goodies Are Here

Front row--a five part set of extension tubes.

They screw together to hold the lens away from the camera body for greater magnification. When all five tubes are used the total length is 65 mm. Mount my garage sale 50 mm lens and the image/object magnification is greater than 1/1--the traditional boundary between macro and micro photograph.

For about $15 including shipping there is a surprising amount of precision machining in that set. A big bargain if you own the proper lens. More about that below.

Above them is an extension bellows. It stretches out to 144 mm. If I add the tubes I can reach 209 mm for some big time magnification. (With all the expected exposure and Depth of Field challenges of course).

As for quality and considering the price ($52 with shipping) I'd rate it up-to-the-job. While I've used precision stages before with less slop these stages weren't cheap.

Mounted on the bellows is my garage sale 50mm lens. Next to it is the garage sale 35-70 mm zoom.

At the far left is a Canon lens. When more parts arrive, it will be mounted to the macro setup with the camera end pointing out. Depending on the lens design and how the aberrations were corrected, flipped lens can create a cleaner image at high magnification.

Now for the kicker hidden in the fine print. Without the garage sale lens this setup won't work. All type G lens sold with the D60 lack an aperture control ring. You need the electrical connections to open and close the lens aperture--something you will do a lot in macro photography. Nikon does make extension rings and bellows with the electrical wiring but you will pay big bucks. Like $75 a single ring (B&H) vs $15 a five ring set (OEC Camera by way of ebay)

The new macro stuff is sitting on a copy stand. I bought that at a photographic auction cheap--mainly because nobody else was bidding on it. I discovered why when I took it home and tried to do something useful with it . In the days of manually focused film cameras, these stand were necessity for copy work. But with autofocus and almost instant feedback on how sharp the image looks, they lost their resale value.

To finish the set up, there is a x, z stage (up, down and sideways) that I already owned--a leftover from a consultant project.

Now all I needed was something to photograph. To start I dug out the innards of a camera lens that had been sitting in the back of a desk drawer for--enough years for me to forget why it ended up in pieces.

And here is my first macro using this setup

Let the macros begin

Tuesday, July 14, 2009

Should I Cheat?

Should I cheat and delete the last blog before the entire Internet World (humor, humor) learns I have been WRONG!!

Wiki agrees with me--on their macro page they say flat out that, like regular photography, in macro photography short focal length lens have a greater depth of field than long focal length lens. So I am not alone in my WRONGITUDUALISM!!

So what reduced me to this embarrassing state?

Dr Guru 'Ray, Sidney F.' in his Applied Photographic Optics--500+ pages of everything you want to know about lens plus a hell of a lot more--told me that when you get into the macro world where your magnification is close to 1 to 1 you can whack off parts of the DOF equations because they don't amount to much.

After the whacking is complete all you got left is the f#, the circle of confusion. and the magnification. So that means--I hang my head in shame--my S3IS doesn't have a better DOF than the $550 Tamron 90 mm macro lens. Focal lengths don't count in the macro world

But then, since focal length don't count, the S3IS's DOF ain't worse either. It will still hang around my neck. I already own it, so it's a hell of lot cheaper than that new $550 Tamron lens. Besides, with only a small fraction of the fancy lens price, I bought a set of extension tubes and an extension bellows. Shipped yesterday according to this morning's e-mail.

Add in what I already own--suitable garage sale and closeup lens plus an x,y focusing stage left over from an ancient optics project--I can now take macros until the cows come home. Mixed metaphor--until the memory cards melt.

You have been warned! Twice!

Monday, July 6, 2009

Why are three cameras hanging from my neck?

Last Sunday I went on one of our monthly flickr walks--this time to the Allen Centenenial Gardens on the UW campus. It's a public garden that for some reason had failed to register in this photographer's brain cells. Not only had I never photographed there, until recently I didn't know the garden existed.

We flickr-walkers weren't the only ones with cameras. It's that sort of place. So after I had stepped away from the group to take a series of reflected images, I returned to the group to find ones of the orher flickrwalkers explaining to a non flickrwalker what she would need to take macro flower photos.

In rough order of expense, she would have to buy at least one:
1--set of closeup lenses
2--set of extension tubes
3--macro prime lens
4--macro zoom lens.

A mini tutorial, well explained--but with one unspoken and wide spread assumption. After you paid all those big bucks for your DSLR the rules demand that you use it to take all your pictures. Big bucks > better camera> photographic masterpieces. Right??

Not necessarily, or so I believe. When I joined the conversation, I lifted the Canon S3IS that was hanging around my neck, and added a fifth option. She could buy a highish end Canon non-DSLR like mine. Then, after she downloaded the totally free CHDK firmware extension she would have all the goodies like RAW mode that the Canon marketing folks stripped out of their non-DSLRs. Plus a whole pile of extra goodies like scripts that the Canon engineering folks never managed to invent.

The best of all worlds and for only a fraction of the price of a macro lens. Or as it turned out, totally free. Her boyfriend owned an S3IS that was gathering dust somewhere.

A wild unusual idea, I know. Carrying more than one camera around your neck. So wild and unusual I was asked to pose for a series of photos of my three cameras to document the concept.

OK, what is going on here. One of the holdovers from 35 mm film days is that 24 mm lens is a wide angle lens, a 50 mm lens is a normal lens (whatever that means) and a 200 mm lens is a telephoto lens. Correct if you are shooting 35mm format. Wrong if your sensor dimensions are not 35mm by something less mm.

Since the vast majority of digital cameras use a smaller sensors we now have the "35 mm equivalent" spec. By the way this is not a new invention of the digital age. With larger and smaller film formats the focal lengths are also different.

Very different, for instance, if you are shooting with a 8 by 10 inch view camera. If you study the history of photography, you would be very hard pressed to come up with any 19th century telephoto images. Except for a few images of the moon shot through astronomical telescopes long enough lens weren't available.

And to jump off topic for a moment, why is 8 by 10 inches such a common print format. First reason--in the 19th century you could only make contact prints and 8 by 10 inches makes for a very viewable size. Second reason, you could wander in to the local general store and buy your glass plates--8 by 10 inches was also the standard window pane of the time.

Which brings up an interest question--how many masterpiece of photograph were lost when some frugal farmer scrapped off the emulsions of a stack of worthless glass negative to repair his greenhouse after a hailstorm. A common practice in years gone by.

But back to the point of this blog (Yes there is one)

Instead of talking about "35 mm equivalents' we could talk about fields of View, Like how wide an angle the photo will take in using a a 1/2.5 (5.75 X 4.31mm) sensor-- which happens to be the sensor size on my S3IS. Then we would easily see (with the help of a little geometry) that the 6 to 72 mm zoom lens of the S3IS had a field of view that ranges from from ~62 degrees to ~4 degrees. If you insist on doing it the old way, the chart--found in a garage sale book-- shows that 35 mm eqivs range from 36 mm to 432 mm.

So what! Little cameras have little sensors and little lens and don't cost much. DSLR have big sensors that cost lots and lots more but they are always better because_______!


Short focal length lens have a much greater longitudinal Depth of Field (DOF). That's optics talk for saying that more of the flower or bug you pointed your camera point at is now in focus. Many of the so called Internet gurus will tell you different but optics is optics and the shorter the focal length the greater the DOF.

So there, all you DLSR-with-zoom-lens lovers!!! My S3IS with a 6mm super macro mode will always have a greater DOF. Ain't just cheaper. It is always--well sometimes always--better.

Here's proof for the doubters.

The top image was taken with a new 90 mm Tamron zoom at f8 in my local camera store . From top to bottom the artificial flower is about 1 inch tall

And this image was taken with my S3IS, again at f8.

To quote the friendly salesman after I showed him the two images. "Looks like you just saved yourself five hundred and fifty bucks."

And to show how the S3IS works in the field. The flower was about 3/4 inch in diameter. The love bugs--they were too busy and preoccupied to be measured

Macros are fun. So expect more posts about them.

You have been WARNED !