Vignetting

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Bob S
Posts: 16
Joined: Sat Aug 07, 2010 2:15 pm

Post by Bob S »

Steve, there are two types of vignetting happening, and you are probably getting confused between them. It is easy to get confused when you try to use optical parts in ways that they were not designed for.

The first issue is simply a limitation of the microscope objective. It was designed to have a 25mm diameter image circle when used with a 200mm tube lens, because that is what was needed for the microscope. You are asking it to fill a 44mm diamter image circle for a full-frame camera, and it is not going to do it. Instead of being disappointed you could be glad that it actually covers a somewhat bigger circle than it was designed for.

There are three ways to deal with this issue. One is to simply ignore the dark corners and crop the image later. Another approach is to use a crop-frame camera, which will concentrate its pixels into the smaller image circle available. The third is to use a longer focal length tube lens, giving higher magnification and a larger image circle, but perhaps with lower resolution because of the limited NA.

Once you understand this then you can deal with the second issue. This is the vignetting when you stop down the lens that you use as a tube lens. The point here is that it is not the size of the front glass that matters, but the diameter and the longitudinal location of the entrance pupil. The entrance pupil is image of the iris diaphragm as seen from the objective's location.

You can get a rough idea of where the entrance pupil is located by placing the lens horizontally on a table, with the iris open, with a white background behind it. Look in from the front of the lens and you will see a white circle; this is the entrance pupil. Now place a marker (such as your finger) above the top front center of the lens and move your head from side to side. You will see that the white circle seems to move from side to side relative to the marker. Move the marker back a ways and try again. Keep going until you find a place far enough back that the entrance pupil does not move relative to the marker when you move your head. That is the location in depth of the entrance pupil. For most 200mm lenses for SLRs that location will be quite far back, near the film plane.

The laws of optic tell you that any light that is going to reach the sensor has to go through the lens entrance pupil. For a normal distant subject this is no problem, but when you stick a mocroscope in front of the lens you are creating a strange optical situation. The light rays from an off-axis point on the subject will come out of the microscope objective at an angle to the optical axis. (Imagine the microscope objective as a pinhole and you can see why that happens.) If the rays from the off-axis image point come out at an angle that causes them to not make it through the tube lens entrance pupil, then those rays cannot reach the sensor, and that off-axis point will be vignetted. The entrance pupil is quite far back, and when you stop down the lens you make the entrance pupil smaller, so it takes less of an angle to make the rays miss the hole, so you get more vignetting.

So one of the qualifications for a lens to be a useful tube lens is to have the entrance pupil large enough and far enough forward to avoid vignetting. These features are irrelevant to the "normal" use of a camera lens, but are vital for this application.

What do you do to avoid this type of vignetting? The first point is, don't stop down the tube lens; it doesn't help anything and it creates vignetting.
If you still have vignetting after taking care of the objective lens image circle issues and opening the tube lens all the way, then you need a different lens. A lens with a lower f/number (wider opening) will have a lrger diameter entrance pupil than a lens with a larger f/number, but you still need to deal with the depth location. You can tolerate a smaller entrance pupil diameter if it is farther forward, closer to the microscope objective. As a rule of thumb, an SLR lens with a long barrel is likely to have its entrance pupil quite far back. A long focal length lens with a short barrel (such as an enlarger lens or large-format lens) will probably have its entrance pupil near the center of the barrel. A simple achromat (+5 is 200mm) will have its entrance pupil at the lens location.

Bob S

stevekale
Posts: 172
Joined: Wed May 11, 2011 2:40 pm
Location: London, UK

Post by stevekale »

Thanks. The points you and Rik made about the entrance pupil of the "tube lens" needing to be well forward makes a lot of sense. I can readily see that this isn't the case with the 70-200mm. I can also see how one might get away with it on a smaller sensor. It is also readily apparent when looking into the lens that using anything other than 200mm narrows the lens aperture significantly (regardless of aperture setting) and leads to more vignetting.

I will post some vignette tests of the two lens this weekend. I want to redo the tests I did on the 70-200. I very quickly tested it with the focus ring set at the closest setting and also at infinity. If I recall correctly, focusing at infinity shifted the front lens forward but led to more vignetting than the test I did with the focus ring set at the opposite setting. (On the 135mm lens, the focus ring moves the rear element while the front element stays still.) When using regular lenses as tube lenses, where do you set the focus?

I also want to measure the amount of vignetting. I tend to prefer the 4x5 aspect ratio and so crop away corners in any event. It will be interesting to see what degree of vignette encroaches on the area I care about.

Rik, what do you use to measure/monitor the image degradation at the edges you mentioned above?

Btw, has anyone tried using the Canon 200mm I L series prime as a tube lens? The old version (rather than the current II version) can be had relatively cheaply used.

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

stevekale wrote:When using regular lenses as tube lenses, where do you set the focus?
Infinity is the nominally correct place. You can focus somewhat closer and still retain good quality, but there are limits to that approach depending on magnification and NA of the objective. See AF motor focusing with a microscope objective for some discussion and illustration of that point.
Rik, what do you use to measure/monitor the image degradation at the edges you mentioned above?
I put it on the end of a 55 mm f/1.8 where vignetting wouldn't be so much of a problem and for sure the sensor field would be big enough to see the edge of the high-quality region. See http://www.photomacrography.net/forum/v ... 9686#59686 for details.

--Rik

stevekale
Posts: 172
Joined: Wed May 11, 2011 2:40 pm
Location: London, UK

Post by stevekale »

Some basic tests with the Nikon Plan Achromat CFI 10x objective. I did not have access to good sunlight and so even illumination was perhaps an issue but here goes:

Canon 135mm f2.0L prime focussed at infinity and aperture f2.0:

Image

135mm focus at close as possible:

Image

Also, for what it's worth (imperfect lighting perhaps makes it worth nothing) here is a grab of the ACR screen with samples (centre was 254):

Image

Now the 70-200mm f2.8L set at 200mm and aperture at f2.8:

200mm infinity:

Image

200mm near:

Image

and grab from ACR:

Image

And, last but not least, an ACR grab of the 70-200mm set to near focus at 200mm (at f2.8 ) with the image cropped to the maximum possible 4x5 aspect ratio (my preferred):

Image

The latter would appear to be ok as far as vignetting is concerned.
Last edited by stevekale on Sat Oct 22, 2011 11:54 am, edited 1 time in total.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Steve I don't see pictures #3, 6 or 7.
The link for the first one is
https://www.sugarsync.com/pf/D6327763_0149166_61531.jpg

Why https ?


By the way a Nikkor 135mm f/2.8 doesn't vignette on 24x36 at infinity, with some stops to spare, iirc. (Same objective)

stevekale
Posts: 172
Joined: Wed May 11, 2011 2:40 pm
Location: London, UK

Post by stevekale »

ChrisR wrote:Steve I don't see pictures #3, 6 or 7.
How about now? I didn't notice that Grab saves a TIFF file rather than a JPG. With Safari as a browser, it didn't matter. Even though I had mislabelled the address as .jpg they still displayed ok.

ChrisR wrote:Why https ?
It's just the way Sugarsync delivers the public link address.

ChrisR wrote:By the way a Nikkor 135mm f/2.8 doesn't vignette on 24x36 at infinity, with some stops to spare, iirc. (Same objective)
Does it matter how the "tube lens" is focusing? So long as it focuses and does so with no (or limited) vignetting?

(I am merely trying to see if I can leverage existing equipment. Having said that, can the Nikkor be mounted on a Canon?)

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Yes can see them now.

Yes you can focus a tube lens a bit closer than infinity, but you're then using rays which are diverging, from the objective. The extent you get away with it varies, of course. Rik et al have done some tests.

Yes you can use any Nikon lens on a Canon. I have chipped and unchipped adapters - no advantage with the former because "focus confirm" is of no real use in macro. Somebody said some extra exposure modes work with the chip, I think. Maybe it was for TTL flash? Manual and Aperture Priority both work though.
You can even use Nikon G lenses on a Canon - see recent post in Equipment.
Some older Nikkors work better on a Canon than they do on some Nikons - at least you can always use the meter and fire the shutter.
With Nikon lens on Canon body you can still focus to infinity.

stevekale
Posts: 172
Joined: Wed May 11, 2011 2:40 pm
Location: London, UK

Post by stevekale »

Those lenses have quite a reputation and seem to be rather cheap used. Not bad for something that has essentially remained unchanged for about 30 years. Any reason to differentiate between the AI and the AI-s? (or even the pre-AI version)

Harold Gough
Posts: 5786
Joined: Sun Mar 09, 2008 2:17 am
Location: Reading, Berkshire, England

Post by Harold Gough »

ChrisR wrote:Why https ?
If you access Ebay on a public page that will be http. Within your account it will be https for your protection:

http://en.wikipedia.org/wiki/HTTP_Secure

So the sugarsync website must have increased protection over that of general websites.

Harold
My images are a medium for sharing some of my experiences: they are not me.

ChrisLilley
Posts: 674
Joined: Sat May 01, 2010 6:12 am
Location: Nice, France (I'm British)

Post by ChrisLilley »

stevekale wrote:Those lenses have quite a reputation and seem to be rather cheap used. Not bad for something that has essentially remained unchanged for about 30 years. Any reason to differentiate between the AI and the AI-s? (or even the pre-AI version)
The differences between AI and AIS mainly apply to Nikon film cameras(which can do shutter priority and program mode with AIS,but not with AI) or Nikon DSLR users who want to add aftermarket chips to the manual Nikkors (again, to enable S and P modes, to get exif info on focal length and aperture, and allegedly to get better matrix metering).

The difference between AI/AIS and pre-AI again applies mainly to Nikon bodies - pre-AI will not meter unless a chip is added, and due to the thicker projection at the back of the aperture ring will likely break the aperture follower (higher end Nikon DSLRs) or the minimum-aperture-sensor (midrange Nikon DSLRs) except for the lowest-end Nikon DSLR which have a push-in rather than turn-round minimum aperture sensor.

For use with an adapter on Canon, none of those things apply. Canon bodies don't have the deliberate limitation that disables metering on Nikon bodies if the body does not know the aperture and focal length. (On the other hand, Nikon bodies don't have the deliberate limitation that disables focus confirmation if the lens does not have electronic contacts). Similarly, the Nikon to Canon adapter does not have either an aperture follower or a min-aperture sensor so there is nothing to break and pre-AI lenses can be mounted safely*.

Having said that, AI lenses are generally older than AIS lenses (the transition happened around the late 1970s to early 1980s) and pre-AI lenses are older than that (the transition happening in the late 1960s to early 1970s). Thus, there are improvements in coatings along the way. the earliest pre-AI lenses are single coated while the later ones are multicoated. The coatings were further improved over time,so typically an AIS lens will have the best coatings compared to the AI or pre-AI versions.

There were also mechanical changes (pre-AI lenses are very solidly constructed, chunky metal things while the later lenses are more lightly constructed (although still all metal, not plastic). The total focus throw from infinity to closest focus was reduced in some AIS lenses compared to the AI versions; good for people who value fast if imprecise focus for reporting etc, less good for people who prefer slower but more precise focus.

Lastly some optical designs were revised over the years, usually an improvement, so for some lenses the later versions are preferred (but those changes do not line up neatly with the pre-AI, AI and AIS classification).

A good guide to the changes in Nikkor lenses over the years, handy to identify exact versions based on their serial number, is
http://www.photosynthesis.co.nz/nikon/lenses.html

Turning now to the specific lens of interest, the Nikkor 135mm f/2.8
http://www.photosynthesis.co.nz/nikon/lenses.html#135

the optical design was changed in January 1976 and was then unchanged through AI and AIS versions until the lens ceased production in 2005. So a 'type K' pre-AI version with serial number starting 7xxxxxx, or an AI,or an AIS version, will all have the same optics.

The focus throw was unchanged at 270° throughout that time, too.
http://www.photosynthesis.co.nz/nikon/specs.html#135

An AIS version is to be slightly preferred, due to the likelihood of better coatings. (I own a late-model type K 135/2.8 but no other versions so I can't tell you if the coatings changed on that one, sorry)> but if you see an AI version at a good price don't hesitate to go for it.

* to be complete, I should add: with the exception of non-retrofocus wide angle lenses which project a great deal into the body and require full-time mirror lockup.

My apologies if that answer was rather more in-depth than you might have wished :) I see a lot of misinformation in that area on the net and would prefer to err on the side of completeness.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I think all the same comments apply to the Nikkor 135mm f/3.5, except that it's cheaper. If I remember correctly the vignetting is the same - none.
As there are fewer elements and we're only using the centre of the glass in a "tube" lens, perhaps it makes no difference.
I was looking for a cheap f/2.8 version but found myself with an f/3.5 I'd bid $27 for!
I could compare (sometime!) if you need..

Post Reply Previous topicNext topic