cactuspic Loc: Dallas, TX

By rereading this thread, I realized I misstated that Zeiss Luminars do not have aperture controls. My bad.

Charles 46277 Loc: Fulton County, KY

The shots I took here with the vase were at f11--an instinctive goal number. I can see at the highest magnification that there was a choice between focusing on the surface of the paint, and focusing on the surface of the background, whereas at 2.5:1 uncropped it all looked sharp. On small formats, f11 has some diffraction effect, which is more noticeable the closer you look (enlarge), but here that did not seem to be a critical problem even in the last blow-up. The parts in focus were still quite sharp, but it became more clear that some of it was not in focus. On further enlargement (not shown here), the image definitely begins to break down.

In large format (especially landscapes or natural scenes), it is very common for photographers to use f22 for everything unless there is a special reason to deviate, in addition to using camera adjustments for focus. In small formats (35mm or APS-C), I tend to go for f11 unless there is a reason not to, as I like overall sharpness (like postcards). If nothing in the picture is nearby, I might go to f8--maybe even sharper. Either way, the idea of selective focus with emphatic bokeh areas is still for me a special effect in most cases, not standard procedure (except in traditional portraits).

cactuspic Loc: Dallas, TX

The aperture shown on the lens dial is just the nominal aperture. In most photography that is also the effective aperture, or close enough. In macro photography, things get a little more complicated. You lose two stops of light to get to 1x and a stop of light for every additional multiple. For example, if you you lens says f/11 at 1x magnification, its effective aperture is f/22. At 2x the effective aperture is f/32; at 3x the effective aperture is f/44, and at 4x the effective aperture is f/64, etc even though the nominal aperture still reads f/11. In the old days, we had to calculate the magnification factor when computing exposure. These days our cameras will just increase the exposure time automatically unless set for completely manual.

I am not sure f/22 has the same impact on diffraction in large format photography as it does in using a 35mm sensor, but I will defer to more knowledgeable hogs on that question.

In addition to the sites listed by mawyatt, the extreme macro site by Johan J Ingles-Le Nobel is very informative. http://extreme-macro.co.uk/

mawyatt Loc: Clearwater, Florida

Edit: This was written while the above was posted, so some duplication.

At Macro and near Macro work the effective aperture becomes a barrier to realizing sharp images. Generally a good approximation for Effective Aperture is EA=Lens Aperture(1+Magnification), thus an f8 lens at 1:1 has an EA of 16. Under normal use where Magnification is <<1 then the EA and Lens Aperture are similar, but near Macro or below this is not the case.

With the small pixels in modern digital cameras, diffraction can start to degrade images at f16, f11 or below. So the EA can become a real barrier at higher magnifications. Good microscope objectives have a lower aperture to deal with diffraction issues. It's usually a good idea to shoot near the lower lens aperture when near Macro levels to keep diffraction in check.

Best,

Mike

mawyatt Loc: Clearwater, Florida

Agree, Johan's site is a great all around resource for macro work, I should have included him....my bad!!

Thanks for noting this.

BTW should mention that some Nikon Macros indicate EA rather than Lens Aperture. Don't know about Canon, Sony and others.

Best,

Mike

Charles 46277 Loc: Fulton County, KY

Yes. And also, note that with auto exposure (here I used f11 in AV mode) we are not particularly aware of the small effective aperture. When we exposed manually, we had to calculate the effective aperture for the sake of correct exposure, but now the camera can do that for us, by the shutter speed--and it does not tell us the real (effective) aperture we are getting. We have to figure that out for ourselves. This is one of the advantages to using close-up accessory lenses (filters) that do not affect aperture--the nominal aperture is the effective aperture. They are fine for many applications.

mawyatt Loc: Clearwater, Florida

Yes, the Raynox Close Ups are really good and inexpensive. I have the 150 and 250, but admit I mostly use them a "tube lens" for the infinite corrected microscope objectives I have. These are modern objectives that are designed to image into another lens, tube lens, before the image can be passed onto the camera sensor. The Raynox 150 & 250 work brilliantly as tubes lenses and cost well below a "proper" tube lens, so they can serve two purposes!!

Best,

Mike

Charles 46277 Loc: Fulton County, KY

By all means post images if you like.

mawyatt Loc: Clearwater, Florida

With Charles's permission, here's an image (with the $18 objective mentioned) of a test chip I designed around about 8 years ago in a QFN package. The gold wires are 25 microns in diameter.

Best,

Mike

(Download)

Charles 46277 Loc: Fulton County, KY

Well, you got your money's worth there. Can you explain how you did it? I am not sure how that lens is used. What are tube lenses?

mawyatt Loc: Clearwater, Florida

This lens from AmScope does not require a "tube lens". It's placed a given distance from the camera sensor (160mm) using a bellows. The depth of field (thin in-focus ribbon) is only 55 microns wide with this lens at 4X, so multiple images were taken at ~25 microns increments moving the camera/lens closer to the subject utilizing a computer controlled motor driven focus rail system from Wemacro (I also use a custom motorized rail with the Wemacro controller and software). This produced 100 individual images which were subsequently stacked with Zerene software rendering a final image all in focus. This was then processed in PS.

The only "optic" used in creating this image was the AmScope 4X which goes for $18. The magnification of this lens can be changed over a reasonable range by simply changing the lens to sensor distance. With infinite corrected lenses, the tube lens FL must be changed...so not quite as simple. These lenses (inf. cor.) are considered having the overall best IQ at higher magnifications, but cost considerable more that $18 and require some form of tube lens (additional $).

So from an IQ/cost ratio this AmScope is hard to beat IMO!

Edit: Forgot to answer about "tube lenses". These are lenses that usually end up in the tubes of modern high end microscopes. They have an advantage that the distance between the infinite corrected objective lens and the tube lens isn't generally critical which allows some flexibility is use with adding things in the optical path. The designated "tube lenses" aren't cheap but some folks have found that the cheap Raynox 150 (208.3mm) magnifying lens works well as a tube lens, as do some of the older 200mm fixed lenses from Nikon (200mm F4 "Q") and others. Sometimes certain infinite corrected objectives (Mitutoyo for example) can be "pulled away" from their design magnification by using a different FL tube lens like the Raynox 250 (125mm) or Carl Zeiss 135mm F3.5 or Vivitar (Komine build) 135mm F3.5. The effective magnification becomes design magnification times actual tube FL/design tube FL.

Hope this helps.

Best,

Mike

mawyatt Loc: Clearwater, Florida

Here's an example of what can be done with these techniques. Can has followed good advice of folks (and minor advice from myself) over on http://www.photomacrography.net and has created a really nice setup based around the Wemacro focus rail. The Mitutoyo and Nikon objectives he shows are expensive, but the Lomo (Russian) is reasonable I believe.

https://www.dpreview.com/news/6016272115/incredible-microscopic-close-ups-of-a-peacock-feather

Here's a link to Robert O'Toole site with images with the $18 AmScope lens.

https://www.closeuphotography.com/seventeen-dollar-plan-4x-objective/

So for the cost of a good dedicated macro lens one can create a pretty powerful and flexible studio setup to dive deep into the macro world and produce stunning images like Can, Robert, Lou and others have done.

Best,

Mike

Charles 46277 Loc: Fulton County, KY

OK, those results are spectacular--perhaps even seeming unrealistic in the peacock feathers you cited. I am not sure I would want to go that far on my amateur work. I used to use a Canon bellows, but now I can just attach the digital camera to the back of a 4x5 camera for a bellows. Extension tubes added to the camera depth of a couple of inches go far enough for most macro options, but for a 150mm lens I could use the view camera bellows. Perhaps one could even make a macro panorama sequence left to right, moving the camera back (or top to bottom) and stitch the images together... The Kodak Cine 25mm does not need any extension for 2x magnification, but on the bellows could go much higher (even using an 8x10 camera). Stacking focus at the very high magnifications seems to require extraordinary minute controls, such as you describe, as well as complex software. My view cameras allow short lenses (even touching the ground glass if not careful), out to very long extensions. A flat bed on a flat stand might be most stable for macro. To be honest, I have not had much inspiration to go much beyond 1:1 or 2:1--partly because of the challenges, and partly because of the artistic merit--science and commercial applications are less interesting to me. Your reference to Robert O'Toole shows a simple set up that looks promising. It would give a bit more distance from lens to subject compared to the Kodak 25mm.

mawyatt Loc: Clearwater, Florida

Charles,

Yes Can's peacock feathers are stunning, others over at the site I mentioned have used peacock feathers as well, some as a image quality reference.

I have done some stitching awhile back, the result was 19 stitches of 220 stacks per image. Final rendering after cropping and post processing yielded an image of almost 19,000 by 16,000 pixels in full resolution 16 bit TIFF format. My goal is to get to a 30,000 by 20,000 pixel chip image. All this requires extreme precision and repeatability, the special lens setup created for this need to be Telecentric (magnification independent of lens to subject distance). I can't show this chip image as it's highly proprietary and state of the art technology.

Proper lighting becomes a real issue at higher magnification as well, uniformity being extremely critical for chip images, much more so than insects or other subjects. Some chips have tiny 100 micron spherical solder balls for electrical connection, many thousands on certain chips. These are like tiny spherical mirrors everywhere and show ugly specular artifacts if the lighting is not absolutely uniform. All this creates a difficult lighting setup where I've used quadruple diffusion with dual light tents, one inside the other and up to 12 studio strobe flashes all synched up with RF and optical triggers. When all these fire everything light up, so best to not be in the same room!!

Cheers,

Mike

Charles 46277 Loc: Fulton County, KY

So you can't do this with a Brownie, right?