Advice on first microscope objective(s) for new SLR camera

[nielsgeode]

Hi all,

Since my Zeiss Stemi 2000C suffers from diffraction and poor quality images I would like to go for something better. For what I have read on this forum (and the internet) the best options are either an infinite CF microscope objective with a camera lens or a finite objective attached to the camera with no additional optics.

After going through the various topics, some things are not entirely clear to me:

I read that there is no absolute winner in terms of image quality (finite or infinite), but I also read:

The infinite set-up is much harder to hold steady than a finite objective on a bellows, expecially with long camera lenses.

Are there people that can confirm this? Is this such a big difference and how much is difference in reality if the setup is really stable?

A quick search on Ebay did not give me a lot of finite objectives, but the infinite ones seem to be readily available. I have the chance to buy a set of two (unused) objectives for USD 592 (EUR 450)
NIKON CF PLAN 20x/0.46 A
NIKON CF PLAN 10x/0.30 A
What do you think of the quality and the price of these two objectives? How well suited are they and would you recommend buying them or is it maybe better to go for the finite solution or a different infinite objective?

Last but not least: to what extend is image quality dependent on the quality and the length of the camera lens? For example, a high quality Canon 200mm f2.8 is $800, a lot of money for a lens only to be used for attaching the microscope objective. Is there a cheaper solution that will give the same image quality? And what happens if you use the Canon 100mm makro in combination with these Nikon objectives? Do you get the same image quality, but only half the magnification?

I am asking all these questions because I am about to completely change my camera gear from Minolta/Sony to Canon. This means spending a lot of money, and I'd like to do it properly the first time.

Thanks!
Niels

[Pau]

I read that there is no absolute winner in terms of image quality (finite or infinite), but I also read:

The infinite set-up is much harder to hold steady than a finite objective on a bellows, expecially with long camera lenses.

Are there people that can confirm this? Is this such a big difference and how much is difference in reality if the setup is really stable?

I don't see the point, a microscope objective on a camera lens makes a very rigid system, at least if the camera lens isn't a cheap zoom. Some bellows are also very rigid, but cheap ones aren't.

Last but not least: to what extend is image quality dependent on the quality and the length of the camera lens? For example, a high quality Canon 200mm f2.8 is $800, a lot of money for a lens only to be used for attaching the microscope objective. Is there a cheaper solution that will give the same image quality? And what happens if you use the Canon 100mm makro in combination with these Nikon objectives? Do you get the same image quality, but only half the magnification?

You don't need at all that canon tele!. The tube lens is much less important than the objective, some members use with great results old manual teles, enlarger lenses or some other things like the now difficult to find but very cheap when available morfanon 172mm

I am asking all these questions because I am about to completely change my camera gear from Minolta/Sony to Canon. This means spending a lot of money, and I'd like to do it properly the first time.

Just a pair of points: don't buy the EOS 60D because a vibration issue, and ...you can also take a look to the newer Sony NEX models with EFSC that have proven to be excellent for macro/micro

[canonian]

I'm not able to answer all your questions, but in my experience an expensive tubelens is not always necessary. With a Nikon CFI 10X I got excellent results using just a Vivitar 100-200mm zoom lens or a Pallas magenta 135mm F2.8 prime with a stuck diafragm, using it wide open. Both lenses didn't cost me more than €50, and the needed conversion adapters and step down rings no more than €20.

An example with a vintage 135mm prime as tubelens:
http://www.photomacrography.net/forum/v ... hp?t=14851

[nielsgeode]

Just a pair of points: don't buy the EOS 60D because a vibration issue, and ...you can also take a look to the newer Sony NEX models with EFSC that have proven to be excellent for macro/micro

I'm aiming for the more high-end models. Currently I have the Sony a900, but precision focusing through its viewfinder is difficult. I had the chance to borrow a Canon 5D for two weeks and I like the remote live view on my PC *a lot*. My Sony a900 doesn't have this (no live view at all) and the images are really noisy at > 400 ISO. These are the two main reasons why I am considering to switch to Canon and I will probably buy the 6D then. However, Sony recently introduced its new flagship the a99 which does support live view, but I can't find if you can also get the live view on your pc (via USB) and I want to read reviews on how the a99 performs with noise at high ISO (800 - 3200) values.

[Pau]

OK, FF has better noise performace and will be the best for many kinds of photography, but for use with microscope objectives APSc is more flexible because the objectives are designed to form an image circle even smaller than APSC sensors. Someones will cover FF with good quality as Rik says, but most ones don't. If you need to use a 300mm tube lens, it will be problematic, as tele photo lenses are very long and other alternatives are much scarce.

Canon LV tethered is very nice, I use it all the time, but other makers can do it too or at least can fill a HD monitor with live image via HDMI, I'm not sure about Sony.

Again, noise at base ISO is very low with modern APSC cameras and at studio there is little reason for higher ISO

...In photomacrography, like in the live, all is question of tradeoffs

[rjlittlefield]

See http://www.photomacrography.net/forum/v ... 953#119953 for answers regarding the objectives.

There are lots of test results from multiple people showing that the tube lens is not very important. The reason is that the objective stops down the optical system to typically around f/20 effective. As long as the tube lens does not have excessive amounts of CA you'll be fine. See Infinity objective on low-end zoom telephoto works fine for an early test report with much discussion added over the last 2-1/2 years. To get the very last bit of performance from an objective of course you'd use its designed-to-match tube lens that would go in a microscope. But those are expensive and require unusual mounts also.

About the A99, I suggest a Google search on "sony a99 tethered live view". Everything I see says that it's not supported.

For use with microscope objectives, I agree with Pau's reasons and summary that APS-C will work better.

--Rik

[nielsgeode]

To get the very last bit of performance from an objective of course you'd use its designed-to-match tube lens that would go in a microscope. But those are expensive and require unusual mounts also.

Why not use a "normal" non-stereo microscope with matching objective and tube lens and then put the camera simply on the microscope? That should give the same results as using a special dedicated adapter, shouldn't it?

For use with microscope objectives, I agree with Pau's reasons and summary that APS-C will work better.

I guess that if you have the right objective that is capable of projecting on a full frame sensor, you will get the same (or better) quality than with APS-C? The other option, of course, is to buy a second dSLR with an APS-C sensor just for the micro shooting. However, I can imagine that you can then just as well spend the money on a better objective. What's your idea about this? Because I will also do a lot of "normal" photography and macro photography with the Canon 100 mm macro (or MPE-65) and with these lenses full frame does make a difference, I will either buy a full frame body or a full frame and an APS-C body (Canon 1Ds is nice, but too expensive ) .

Niels

[rjlittlefield]

Why not use a "normal" non-stereo microscope with matching objective and tube lens and then put the camera simply on the microscope? That should give the same results as using a special dedicated adapter, shouldn't it?

It would give exactly the same result if you could fit the sensor of the camera down inside the eyepiece tube at the place where the eyepiece normally sees it. But since you can't do that, it takes additional optics to relay the image from that position to the sensor, or different mechanics to allow mounting the camera in the right place.

--Rik

[Charles Krebs]

I guess that if you have the right objective that is capable of projecting on a full frame sensor, you will get the same (or better) quality than with APS-C? The other option, of course, is to buy a second dSLR with an APS-C sensor just for the micro shooting. However, I can imagine that you can then just as well spend the money on a better objective

Microscope tubes are either 23mm or 30mm in diameter. Microscope objectives are made to create an image that will "fit" in one of these tube sizes. An APS sensor has a diagonal of ~27-28mm, a 24x36mm sensor has a diagonal of 43mm. There are no microscope objectives that were intended to directly "cover" a 24x36mm sensor, regardless of price. There is a fair number of objectives that do a fine job covering the APS sensor. There are a few that can be used on a full frame sensor ("over-achievers" I call them), but you had better get someone with first-hand experience with the particular objective, and look carefully at the corner quality and light fall-off characteristics.

Then you need to consider working distances, another key characteristic for the type of subjects you have in mind. Also remember that, for example, a 10X objective on APS sensor takes in a field size of 1.5 x 2.3mm. To record that same field size in a full-frame camera would require a 15X objective.

[nielsgeode]

There are a few that can be used on a full frame sensor ("over-achievers" I call them), but you had better get someone with first-hand experience with the particular objective, and look carefully at the corner quality and light fall-off characteristics.

Is it correct that you are saying that even the finest objective performs better on an APS-C camera than on a full frame sensor?

Probably I am wrong, but I think that if you take an image from a full frame sensor and crop it in photoshop to the size of an APS-C sensor, it should look the same as if the picture was takenon een APS-C camera, shouldn't it? And because you then 'ignore' the egdes and use the same projection from the objective on the full frame sensor as on the APS-C camera, the "objective quality" in both images should be the same. Please do correct me if I am wrong...

Then you need to consider working distances, another key characteristic for the type of subjects you have in mind. Also remember that, for example, a 10X objective on APS sensor takes in a field size of 1.5 x 2.3mm. To record that same field size in a full-frame camera would require a 15X objective.

This is of course true and I think for me it might be an advantage as well as a disadvantage because I still have to see how often I will really need > 10X magnification. However, I guess that simply switching the 200mm tube lens for a 100 or 150mm will do the job with equal quality?

[Charles Krebs]

Is it correct that you are saying that even the finest objective performs better on an APS-C camera than on a full frame sensor?

No. What I am saying that there are no objectives, even the "finest", that were designed to create an image large enough to "cover" a sensor with a 43mm diagonal. Some can be used on a full-frame sensor but do not expect that to be the universally the case. This is far beyond what the manufacturer intended. The only way to find out which ones might work OK on full-frame is to try it and check the results. Some of the ones that will be OK on a large sensor are not necessarily in the "premium" lines and are modest in price. And some of the extremely expensive specialized objectives have a field number no larger than 22mm (diameter of quality image), and definitely would not work well if directly projected onto on a full-frame sensor. (When I say they "would not work well", what I mean is the center would be great, but the edges and corners would be poor).

Think of it like lenses designed and made to be used on an APS camera. If you put them on a full-frame camera many would vignette badly. Some would illuminate the sensor out to the corners, but the image quality would be very poor at the edges and corners. But you might find one or two that give good image quality to the corners. But that is not something you can expect, and must be tested.

I think that if you take an image from a full frame sensor and crop it in photoshop to the size of an APS-C sensor, it should look the same as if the picture was takenon een APS-C camera, shouldn't it? And because you then 'ignore' the egdes and use the same projection from the objective on the full frame sensor as on the APS-C camera, the "objective quality" in both images should be the same. Please do correct me if I am wrong...

If the sensors have about the same pixel density (number of pixels per mm) then this is accurate (assuming the other sensor characteristics are about the same). There are many APS sensors that have greater resolution capability than full frame cameras (think in terms of pixels per mm). There are a few objectives that can't be used on a full frame (because the image will not be large enough) but produce very high resolution in the APS size area. In this case, if the APS sensor has greater resolution than the full frame camera it would actually be a better choice.

However, I guess that simply switching the 200mm tube lens for a 100 or 150mm will do the job with equal quality?

With a full frame camera you will need to select your objective carefully if you do not want to crop. This is the case using the regular tube lens focal length (200mm for Nikon and Mitutoyo, 180mm for Olympus). When you obtain lower magnifications by using a shorter focal length tube lens you will find that the "acceptable quality" image circle size gets smaller as you reduce the tube lens focal length. I don't know of any objective that will work nicely to the edges and corners of a full frame camera with a 100mm tube lens. Even with an APS sized sensor it is not that common to have an objective that will provide good quality over the entire sensor area when using a 100mm tube lens.

[rjlittlefield]

Is it correct that you are saying that even the finest objective performs better on an APS-C camera than on a full frame sensor?

No. What I am saying that there are no objectives, even the "finest", that were designed to create an image large enough to "cover" a sensor with a 43mm diagonal. Some can be used on a full-frame sensor but do not expect that to be the universally the case.

Let me try explaining the issues in a different way.

Key facts: every microscope objective is designed
1) to be focused at a specific distance from the subject, and
2) to cover a certain field of view on the subject at that distance.

These specifications are "in front of the objective", and they basically cannot be changed. Changing the focus distance will add aberrations, and trying to enlarge the field of view will get you into areas where existing aberrations degrade the image.

There are other specifications "behind the objective". There you have more flexibility.

Infinity objectives are designed to give rated magnification with a certain length tube lens, typically 200mm. Changing the focal length of the tube lens will make the final image larger or smaller in proportion to the focal length. By adjusting the focal length of the tube lens, you can fit the high quality image circle to any size sensor.

Finite objectives are more challenging, at least in principle. You can adjust the magnification and image circle size by just changing extension, but this also changes focus distance in front of the objective. Low magnification objectives such as 10X NA 0.30 and below will let you do this without visible degradation, but at higher magnifications (wider apertures) the tolerance gets smaller. With a 50X NA 0.55 objective the tolerance is down to a few percent change. To adjust the magnification without changing the focus distance, you have to add optics behind the objective. This is simple enough in theory, but in practice the required lenses may be difficult to obtain and require custom mounts.

The advantage of APS-C sensors (27-28 mm diagonal) is that in practice they provide a good fit to the high quality field of view using readily available lenses and mechanical adapters. It is a happy coincidence that this also happens at rated magnification.

A few objectives have a sufficiently wide field that they can also cover full frame sensors (43 mm diagonal) using readily available adapters at rated magnification. All other objectives can also be made to fill a full frame sensor, but only at higher than rated magnification and using lenses and adapters that are not so readily available.

In all cases, when the image circle is matched to the sensor size then you get the same field of view on subject, and in the final image you get the same diffraction blur and DOF.

--Rik

[nielsgeode]

A few objectives have a sufficiently wide field that they can also cover full frame sensors (43 mm diagonal) using readily available adapters at rated magnification.
--Rik

Do I understand correctly that even with these objectives, the image quality will benefit if you change from full frame to APS-C sensor?

[nielsgeode]

Just a pair of points: don't buy the EOS 60D because a vibration issue,

What excactly is this vibration issue?

[Charles Krebs]

What excactly is this vibration issue?

The feature we are after is the Electronic First Shutter Curtain. When using most Canon bodies with this feature "active" there is a very slight internal noise at the start of an exposure (the 50D is completely silent). But there is no discernible vibration that results from whatever is making that slight internal noise. On the 60D there is a slight vibration that occurs when the noise occurs. It is not that large, but if you are getting a camera because you want the vibration-free benefits of this feature it would be best to choose a different model.

Pau initially brought this up, and somewhere in the forum is a thread that discusses it.