stack with microscope objective and bellows

[tpe]

This was my first attempt with a microscope objective, untill all the bits of the real ones i have orderd arrive.



the setup looks like this
http://www.scientificillustration.net/i ... era_01.jpg

The shots are pretty washed out and the edges are very unsharp and it was taken with a microscope objective (the postage cost more than the lens) that just had the markings Swift & sons. London. |in. 0.25 on it, and an adaptor made out of an old CD-ROM attached (first cardboard tube then) by a set of bellows to a K&M 5D and stacked from 66 images.

I don't know anything about the lens and just tryed it, the main problems it gave were the DOF had a sort of echo, so there was a sharp plane, then an unsharp plane then a smaller sharp plane. The shots were very milky, and the other problems were from the stacking, where there is that nasty halo around all the hairs etc. If anyone has any ideas about what i should expect from this kind of setup it would be great to hear, and what to so next tim, and how to get it less milky, if i can put an iris in there somewhere where should i put it, and is there any way that i could get just the center of the image to fall on the ccd? Thats on top of any other comments anyone might have etc.

many thanks

tim

[rjlittlefield]

Tim,

First, try extending your bellows a bit more. This will give you more magnification, moving those fuzzy corners off your camera's sensor and giving you just the central part of the image you have posted here.

More extension may also improve the image quality. Most of those old objectives were designed to cast their image around 150 mm away from the shoulder of the mounting threads. Using an extension different from that can introduce some additional aberrations. However, this is typically not a big deal with low power, small numerical aperture lenses like you have. The main thing is that you don't want too little extension because then the camera sees those fuzzy corners.

It's hard to say what's causing the milkiness. First thing to do there is pull the lens off the camera and just look through it by eye. If the image looks milky that way, then there's a problem with the lens -- fungus or bad coating or delaminated or maybe just very dirty (but I presume you already checked that!).

If the lens looks good as a magnifying glass but the camera sees milk, then the problem is probably stray reflections from somewhere in the adapter chain.

To find those, start by taking the camera off the bellows and looking in by eye, with the subject illuminated as if you were going to take pictures. Look for any bright reflections.

If you see some, you need to either mask those surfaces or cover them with dull black paint, flocking, or fabric. I have an old piece of black velvet that I cut little pieces from sometimes. I also have a can of Krylon Ultra-Flat black paint. It's pretty good, though not as good as the velvet.

It's also conceivable, though not likely, that some of the reflections are coming from the camera mount. You might consider cutting a black construction paper mask and mounting it just inside the bellows so that the objective can't see any parts of the camera except the sensor.

The DOF echo is very strange. You say there's one plane sharp, then a fuzzy area, then another sharp plane? I've never seen that before and can't imagine how it could happen without some pretty serious damage to the lens -- say it's cracked or severely delaminated so that it's sort of a "bifocal".

It's normal to get some halo around the hairs on this sort of shot, because of the stacking. What I see in the picture posted here looks pretty typical.

Significant improvements can be made in this picture by post-processing to correct the color balance, subtract out some of the milk, and do a bit of sharpening. With your permission, I can post out an image to show what I mean.

Hope this helps...

--Rik

PS. I notice that your image is being hosted in a _temp directory at your own web site. I presume that means you're thinking to delete it at some point in the not too distant future. But it's better if that doesn't happen. The forum has more value when postings including their images remain intact as long as possible (years). That way they become an archived resource that we can reference in future work. Please, feel free to use the photomacrography2.net server for image hosting, if you don't want to host images indefinitely yourself.

[beetleman]

I will certainly give this first try a big "WOW". Can`t wait to see more. What else are you getting for the setup?

[Ken Ramos]

A great image there Tim. We are beginning to see more and more work like this it seems. Really great stuff.

[Charles Krebs]

Tim....
everything Rik said

If the objective is marked "0.25" I'm guessing that is the NA, and it is probably a 10X. This shot appears to be considerably less than 10X (on sensor... just a guess on my part) so you probably had less then the 150-160mm extension that would be "best". That will likely (as Rik mentioned) contribute to the soft corners. Remember, microscope objectives were really designed to produce a "quality" image circle in the 20-25mm diameter range (significantly less than the diameter of most DSLR sensors). Some may not even be be that great at the corners with a 20mm diameter image. More recent "higher end" objectives will often provides a larger useful image. For example... Olympus made "wide field" eyepieces that had a field number of 26.5mm so apparently the objectives to be used with these provided good image quality (hopefully!) with an image circle of at least 27mm.

With the modest NA's of the objectives that are useful on bellows, I haven't found that a bit more or less bellows makes a really big difference in the overall image quality produced by the objective. But the intended extension for these older "finite" objectives is usually 150-160mm, and using less will very often result in the "quality" portion of the image circle being too small, and yield poor edge and corner definition. On a microscope the image would typically be "enlarged" (or reduced for very small sensors) before it entered the camera so as to make that 18mm diameter "sweet spot" cover the film/sensor.

is there any way that i could get just the center of the image to fall on the ccd?

... either longer bellows or perhaps a 1.4X teleconverter will accomplish this, but then the magnification might be too high for the subject.


In the link you gave to picture the set-up, it appears that the diffusion "dome" might extend well beyond the subject. If this is slightly within the "view" of the objective (or just outside of the view of the objective) flare levels can increase very dramatically and give that "milky" look.

Quite nice for an initial attempt. Nice angle and composition. OK, so the edges are a little soft, the central core looks good. Color and contrast could be improved with a little image editing. Once you sort a few things out I think you will be very pleased.

Charlie

[tpe]

Thanks all, this is a world of information that is just impossible to get anywhere else. Sorry about the delay in getting back, since the last post i have lost a hard disk, then the camera packed up then the motherboard packed up so it has been difficult getting pictures on line, but I did however receive 90% of a new second hand microscope . So hope to get some more stuff up as soon as possible.

Charles, first of course thanks very much the help is invaluable, and it amazes me that both you and Rik managed to make so many correct conclusions about the setup from so little information. The lens is a strange one. Since the picture i have received a 10x and a 4x and this one gives something inbetween when they are all used on the same scope if that means anything. Actually the teleconverter idea has prooved very usefull, not so much on this setup but on my new microscope. I probably shouldnt be doing it but the picture quality looks better than on the standard 35 mm adaptor and i get more controll over the magnification.
Yes the diffusion dome did go byond the subject, but now has been changed, and it makes a good deal of difference.

Ken, yes i think it is getting into the mainstream now, actually i think it is great that software is really beginning to help with optics (instead of just make cutsy filters) and this and amateur astronomy have really got a big lift, making acceptable results at a reasonable price.

Many thanks beetleman, I have just got a reichert microscope, some zeiss objectives, for my bellows and a Reichert 2.5x objective for the reichert. I am adapting the microscope and a childs fiberoptic lamp to try and get a reasonable lighting setup around the stage, so fingers crossed .

RiK you and Charles have been a real help, i am hanging on every word from both of you . Bellows, and length of extention, great, now using a new extention length and it looks better, unfortunatly camera dies half way through the stack. I am going through the chain and painting everything black now (especially as the extention tubes are err to say the least, novel uses for their materials ).

The bifocal thing is still a bit of a mystery, i guess it is delaminating or something similar, or perhaps reflection between the lenses of the objective, it looks to pre date any kind of coated optics.

The halo is really bothering me still, It would be so nice to take some pictures of the antenne of the mosquitos out side my window, they also have some wonderfull jeweled eyes, that i had never seen before trying this, that are quite beautifull, but i guess the limitations of this type of technique are keeping away from geometry where there are many gaps and objects crossing over each other.

tim

btw will be sure to make the image links permenant, it also explains the three image rule, if the images are hosted by photomacrograpy i could guess the bills are going to climb.

[rjlittlefield]

Tim,

I'm saddened to hear about your troubles, but also glad to see that things seem to be getting back to normal in your world.

Looking again at your initial post, I see that you wrote that the lens is marked with Swift & sons. London. |in. 0.25 . I believe that last part is actually 1 in. 0.25 , meaning 1 inch focal length, and NA 0.25 as Charles suggested. On a standard 160mm scope, that would give magnification very close to 5X.

I'd be interested to hear about those extension tubes made of novel materials. The most extreme I've ever done involved PVC plastic pipe. That stuff is shiny! Very challenging to get it dulled, blackened, and baffled. But it did succeed in coupling a C-mount video camera to a traditional microscope (sans eyepiece). For several days I watched protozoa on the TV as if it were an aquarium! (Hhmm, those memories would be, what, about 25 years old?)

About the halo, some amount is unavoidable but each software package has one or two tunable parameters that affect the amount you get. We can talk about those in more detail when you get some more stacks to play with.

--Rik

[arlon]

pretty cool for a first try!