- HOW TO MAKE A LENS COMPARISON/TEST -

[magom]

Hi guys,

its been a while since my last topic but I'm always reading some interesting comments and news in this great Forum.

Today finally I have time to make the big lens comparison of my collected lenses, that I want to share with you when it is done. I only have this night to do it.

My ambition:

- Comparison between

Apo Rodagon N 50mm
Scheider-Kreuznach M-Componon 80/4
Leitz Summar 4,5 80mm
Apo Rodagon N 105mm/f4
Sigma 150mm f2,8 Macro
Tamron 90mm f2,8 Macro
Canon 50mm f1,4 with Extension Tubes

- in all reproduction scales (different bellow positions and extension tubes)
- with and without the Raynox 6x,12x and 24x CM-3500 Lenses
- in Retro and in normal position

My needs from you guys:

- Websites with Tables of reproduction scales, f-stop calculations etc. for knowing which specification I am using with each setup in the comparison.
- A tipp how to deal with different distances of my standarised reference table or object.
- more Infos to do the test the elaborate it can be

I would appreciate a lot, if you can post some links or something similar, where I can learn the best an accurate methods to do the test properly.

Maybe we can built up a topic where future guys can find a portfolio of information about doing a lens comparision for sharpness et cetera.

Thanks in advance!

Best regards, Magom

[rjlittlefield]

I only have this night to do it.

I appreciate your enthusiasm, but the task you have outlined seems much bigger than the time and resources you have to do it.

I suggest spending the night with a much smaller set of tests.

To get the best value for your time, I suggest choosing a single magnification that you personally care about, and restrict your efforts to determining center and corner sharpness over a range of f-stop settings.

See http://www.nnplus.de/macro/Macro100E.html for one example of tests that can be done pretty quickly with no special test targets or techniques.
For an example at higher magnification using a focus stacked subject, see http://macrosmuymacros.com/index.php?op ... 25&lang=en.
For a set of very formal tests focusing on MTF, see http://coinimaging.com/Lens_tests.html.

--Rik

[magom]

(...) but the task you have outlined seems much bigger than the time and resources you have to do it.

Hey! Thanks a lot!

You are totally right! But that is okay! Even if it costs me more days, I think it is worth the work.

The links you posted are exactly what I was looking for. Do you also have some Sites where I can find more about the caldulations of distances, f-stops when using attached enlargement lenses like raynox? (I remember haven´t read some topic of morfa one day where he calculated the "new" intensity of light of such a combined lens, but couldnt find it)

[TheLostVertex]

The links you posted are exactly what I was looking for. Do you also have some Sites where I can find more about the caldulations of distances, f-stops when using attached enlargement lenses like raynox? (I remember haven´t read some topic of morfa one day where he calculated the "new" intensity of light of such a combined lens, but couldnt find it)

I think it is best to get an accurate ruler of some sort, and just measure the magnification. Magnification=Field Size/Sensor size. Then you can will have the actual magnification, and you can give the field size(more useful often times).

For aperture, you should be able to take the magnification and the nominal aperture marked on the lens and calculate that. Effective Aperture=Aperture(Magnification+1)

[rjlittlefield]

Do you also have some Sites where I can find more about the caldulations of distances, f-stops when using attached enlargement lenses like raynox?

No, I don't have any good links for that.

A good estimate for effective f-number can be obtained by comparing the combo under test against a known lens looking at the same target under the same illumination. Adjust the exposure time to get equal histograms, then estimate the f-number of the unknown combo as equal to the f-number of the known lens adjusted by the square root of the ratio of exposure times.

However, even that simple procedure can easily give a misleading answer.

I think it is better to just describe exactly how you had the lenses set up. If you test with the Raynox on the Sigma 150 at close focus and set on f/11, then say that. In that particular case, the effective f-number will not be changed by the Raynox, but it won't be f/11 either.

--Rik

[TheLostVertex]

I think it is better to just describe exactly how you had the lenses set up. If you test with the Raynox on the Sigma 150 at close focus and set on f/11, then say that. In that particular case, the effective f-number will not be changed by the Raynox, but it won't be f/11 either.

--Rik

Would I be mistaken in thinking that in that set up(sigma@F11 on the Raynox)that you can just calculate the effective aperture based on the sigma's aperture and the combos magnification? Since the Raynox is not acting as a pupil.

[rjlittlefield]

Would I be mistaken in thinking that in that set up(sigma@F11 on the Raynox)that you can just calculate the effective aperture based on the sigma's aperture and the combos magnification? Since the Raynox is not acting as a pupil.

As I read your words, you're mistaken.

Because the Raynox is not acting as a pupil, the effective aperture is determined entirely by the Sigma.

That will be different from f/11 depending on where the Sigma is focused, so you could think of it as depending on the magnification of the Sigma alone.

But it does not depend on the magnification of the combo. You could stick on any power of closeup lens to get different magnifications, and the effective f-number would not change.

Of course we're talking about effective f-number as seen by the camera. If you look at things on the subject side, you'll find that the entrance cone gets wider to match the increasing magnification of the combo. One way to make sense of that is to think of the same diameter aperture getting closer to the subject. Another approach is to just accept that the angular widths of the entrance and exit cones are always in a ratio equal to the magnification.

--Rik

[TheLostVertex]

But it does not depend on the magnification of the combo. You could stick on any power of closeup lens to get different magnifications, and the effective f-number would not change.

Of course we're talking about effective f-number as seen by the camera. If you look at things on the subject side, you'll find that the entrance cone gets wider to match the increasing magnification of the combo.
--Rik

I think my mistake here maybe that I did not read the lens list, nor infer the correct set up. I was thinking:

Camera-Raynox lens-Sigma lens(Stopped down)
In that order.

When in fact, I think we maybe discussing:
Camera-Sigma Lens(Stopped down)-Raynox Lens

So would it be correct then to say that in case 1 it would be the magnification of the combo, and in case two it would not be?

[rjlittlefield]

When in fact, I think we maybe discussing:
Camera-Sigma Lens(Stopped down)-Raynox Lens

Certainly that's what I was talking about.

The setup Camera-Raynox-Sigma did not even occur to me. The mounts and optical designs are not built to work that way. It works well to stick a teleconverter between camera and macro lens, but the Raynox's are designed to go on front of ordinary lenses. They can also be used as the rear component ("tube lens", "converging lens") in combination with an infinity objective stuck on front, but that's again a different situation.

So would it be correct then to say that in case 1 [Camera-Raynox lens-Sigma lens(Stopped down)] it would be the magnification of the combo

I suspect that's close but not exact. Lens combos always change the pupil factor so you'd have to throw that into the calculation also.

--Rik

[TheLostVertex]

Certainly that's what I was talking about.

The setup Camera-Raynox-Sigma did not even occur to me. The mounts and optical designs are not built to work that way. It works well to stick a teleconverter between camera and macro lens, but the Raynox's are designed to go on front of ordinary lenses. They can also be used as the rear component ("tube lens", "converging lens") in combination with an infinity objective stuck on front, but that's again a different situation.

After reading your reply, I realized my mistake and reread the list of optics. Sure enough what I thought was being talking about actually made no sense with the components

So would it be correct then to say that in case 1 [Camera-Raynox lens-Sigma lens(Stopped down)] it would be the magnification of the combo

I suspect that's close but not exact. Lens combos always change the pupil factor so you'd have to throw that into the calculation also.

I was thinking that it would be a close approximation, but not exact.

Atleast my face is only half-egged today.

[magom]

Hey guys,

you are totally right, it costs a lot of time to do all the calculations.

Here are the Calculations found on morfa's flickr page, done for my APO Rodagon N 105/4:

http://www.flickr.com/photos/johnhallme ... 4592459772

APO Rodagon N 105/4 + Raynox 24x CM-3500
105/40=26,25
+24 diopter which means a focal length of 1000mm/24=41,7mm
1/F=1/f1 + 1/f2 => 1/F= 1/105 + 1/41,7=> F= 30,9462
30,9462mm/26,25=1,1789 f1,18


APO Rodagon N 105/4 + Raynox 12x CM-3500
105/40=26,25
+11.8 diopter which means a focal length of 1000mm/11.8=85mm
1/F=1/f1 + 1/f2 => 1/F= 1/105 + 1/85 => F= 49,756mm
49,756mm/26,25=1,895 f1,895


APO Rodagon N 105/4 + Raynox 6x CM-3500
105/40=26,25
+6 diopter which means a focal length of 1000mm/6=166,7mm
1/F=1/f1 + 1/f2 => 1/F= 1/105 + 1/166,7 => F= 69,7733
69,7733mm/26,25=2,658 f2,658



Could you give me an explanation how I can calculate the exact distance between entering Lens or chip with subject? I don't know how to make the exact distance when using lenses of different focal lenghts. Also infos about the bellows pullout would be great.

I don't now how to compare exactly the different lenses...

Thanks a lot!

[rjlittlefield]

Could you give me an explanation how I can calculate the exact distance between entering Lens or chip with subject? I don't know how to make the exact distance when using lenses of different focal lenghts.

Nobody knows that.

In theory it's a simple calculation that depends only on knowing the exact focal length and the separation between front and rear principal planes.

In that case the distance from sensor to subject is just:
d = FL*(1+mag) + FL*(1+1/mag) + separation

You can make sense of that formula as
FL*(1+mag) = extension behind the lens
FL*(1+1/mag) = extension in front of the lens

In practice the calculation is impossible because separation is never part of the published lens specs. Worse, modern lenses often change focal length as you turn the focus ring, and that's not part of the published specs either.

As a practical matter, the only way to match magnifications for macro work is by experiment and interpolation.

--Rik

[magom]

As a practical matter, the only way to match magnifications for macro work is by experiment and interpolation.

--Rik

That is super intresting Rik, thank you. I will do it like you say. Thanks for the formula by the way.

Do you or does anyone know a good way or workflow in Photoshop to do a crop comparison? When I crop the images made with not accurate and same image ratio.

I [/quote]

[magom]

Here is a first short test for establishing a first idea of the setup. Note that I was using Av-Mode with the Canon 7D. I think it is better using a flash setup, isn't it?



As an interpretation of the first short test I can say, that the lens in reverse position works better when wide open. At f5,6 and above, the normal position has a better sharpness. I have tried to find the focal point via Microstage and Live View. Is that okay?

The problem I think is, that I had to resize (down) the reverse-shots when doing the compilation. Maybe that distorts the images. Any ideas about that?


I hope you guys don't laugh about my first testings. Help an tips are needed

[rjlittlefield]

Do you or does anyone know a good way or workflow in Photoshop to do a crop comparison?

I usually put each image as a separate layer of a single file, tweak their position if necessary to get exact alignment, then crop the whole file. That gives me a bunch of crops all exactly the same. Then I expand the canvas and move the layers sideways so they all show at the same time.

It's simpler to keep things straight if you label the layers while you're constructing the file. An easy way to do that is to construct the original layered file using Scripts > Load Files into Stack. That will name each layer as the file that image came from.

--Rik