El Nikkor 105mm Image Quality in Macro

[bvf]

I was doing some macro work at a magnification of about 4:1 in a metal El Nikor 105mm f/5.6 and a Baader U, and I was disappointed with the results. The images were definitely soft compared with the same subject photographed in visible.

 

Although the El Nik 105mm is rated as good for UV, I had noticed there was a significant focus shift when using my tri-colour UV filters (315BP25, 345BP25, 380BP20). Interestingly, when changing from 380BP20 to 345BP35 you have to rack the lens in (i.e. move it towards the camera) whereas when changing to the 315BP25 you have to rack out (i.e. move the lens away from the camera). Perhaps the problem was that the Baader U lets pretty well the whole of the UVA range through and the El Nik couldn't focus across the range because of its focus shift. So I did comparison shots with the 380BP20 and 345BP25, refocussing for each filter.

 

This is the full image:

 

 

And here are the comparisons, looking at the comma:

 

 

You can see that the Baader U is soft compared with the visible image. And that the 380BP20 image is sharper than the Baader U image. But what really surprised me is how soft the 345BP25 image is: I thought that perhaps I hadn't focussed carefully enough so went and repeated it, and got the same result.

 

So looks like I better use the 380BP20 for macro work rather than the Baader U.

 

So it appears that the El Nik 105 performs OK for macro at the long wavelength end of UV, but image quality drops off noticeably at shorter wavelengths. (I couldn't try the 315BP25 filter because I don't have a light source that will let me focus with this filter.) And so the softness when using the Baader U is probably down to a combination of focus shift and the poorer IQ of the El Nik at shorter wavelengths.

[Andy Perrin]

Hm. At 345nm and especially 315nm you could probably stop down more? What happens when you do that?

[nfoto]

Nikon always stated these enlarger lenses had an 'extended' UV response, but if memory serves, they were thinking to 380nm or thereabouts. So getting much deeper would likely begin to show optical aberrations from design.

 

The Baader U capture does some (weak) indications of lateral CA, whilst the narrower band filters appear cleaner in this regard. Again not unexpected.

 

Getting properly focused images would be much easier, and more reliable, by using tethered shooting and watching LiveView on a computer monitor. Using an automated rail then would be the icing on the cake, so to speak. I assume you do use flash to avoid very long exposures and the concomitant danger of camera movement?

[bvf]

Hm. At 345nm and especially 315nm you could probably stop down more? What happens when you do that?

 

Andy - I did an aperture comparison using the Baader and visible - results below. In the visible it looks like f/8 is just better than f/11. Using the Baader, f/11 may be slightly sharper. But f/8 and ISO 100-200 is usually about the right exposure with my 4 flashguns close-up - if I go to f/11 it means raising the ISO, which I'm reluctant to do.

 

 

 

I did the same test with my beloved Cassar S 50mm. In visible at a macro magnification of about 10:1, this shows optimum aperture is f/8 - f/11:

 

 

 

When comparing the UV filters with the Cassar S, the Baader and 380BP20 images are similar, but some slight softness with the 345BP25:

 

 

 

 

Nikon always stated these enlarger lenses had an 'extended' UV response, but if memory serves, they were thinking to 380nm or thereabouts. So getting much deeper would likely begin to show optical aberrations from design.

 

The Baader U capture does some (weak) indications of lateral CA, whilst the narrower band filters appear cleaner in this regard. Again not unexpected.

 

Getting properly focused images would be much easier, and more reliable, by using tethered shooting and watching LiveView on a computer monitor. Using an automated rail then would be the icing on the cake, so to speak. I assume you do use flash to avoid very long exposures and the concomitant danger of camera movement?

 

Birna - I was focussing using LiveView on the camera, with the image enlarged. With stacking the focussing is not so important as long as you start with the focus in front, say, of the subject and end up with thefocus behind. And yes, I use flash to.

 

I did comparative focus stacks of a wallflower using the El Nikkor with the Baader U and the 380BP25. The difference is not as great as I expected, but the 380BP25 image is slightly better - seems to have less flare. I white-balanced the 380BP25 image as for the Baader U, and had to increase the contrast significantly. In these images, Baader is first then 380BP25:

 

 

 

 

[nfoto]

You obviously have done the homework in a thorough, exemplary fashion. Thus we are basically left with two questions, whether the lens itself is the problem, or the 345BP25 filter.

[dabateman]

You obviously have done the homework in a thorough, exemplary fashion. Thus we are basically left with two questions, whether the lens itself is the problem, or the 345BP25 filter.

 

I question the filter. When you mentioned focus point, it reminded me that you are forgetting filter thickness. Some of these omega filters are close to 8mm thick.

The Baader venus filter is about 1mm thick.

Though your Cassar test, does high light some issues with the EL 105mm.

[bvf]

I question the filter. When you mentioned focus point, it reminded me that you are forgetting filter thickness. Some of these omega filters are close to 8mm thick.

The Baader venus filter is about 1mm thick.

Though your Cassar test, does high light some issues with the EL 105mm.

 

I don't have the specs for the 345BP25 or 315BP25 any more - the items have disappeared from ebay. The 380BP20 description is still there, and that's 3 mm thick. The BP345BP25 looks about the same. The 315BP25 is definitely a lot thicker, perhaps 5mm - but I haven't been able to use that in my tests. But in any case, I re-focussed for each filter so any change in focal point due to thickness would not be a factor.

 

It may just be that 345nm is outside of the El Nik's design range and so the lens is just not very good at that wavelength in a macro deployment. Perhaps I'll compare results from the Cassar with the El Nik to see whether I should use the Cassar for macro work - although I'd prefer to have the greater working distance afforded by the El Nik.

[ulf]

Have you tried the EL-Nikkor 80/5.6?

It has the same or better UV-transmission and might have less focus shift.

[bvf]

Ulf - no, haven't tried the 80mm version. Perhaps I wll treat myelf - there are a couple of metal ones on ebay for £100-£150.

 

I tried a comparison between the El Nik 105 and my trusty Cassar S 50mm. Results are below. I've added a visible light image taken with the El Nik, which shows there's nothing wrong with the lens in the visible region. And for fun, I've added a stereo anaglyph of the Cassar S image. I reckon the Cassar image is definitely sharper (even allowing for the over-exposure of the El Nik image. For the time being, I'll have to stick to the Cassar S for UV macro.

 

The flower is the sub-centimeter Gill-over-the-Ground (a.k.a. ground-ivy, creeping charlie, alehoof, tunhoof, catsfoot, field balm, and run-away-robin - and for the botanists as Glechoma hederacea).

 

PS: I don't know why the images have come up to small - I resized them to 600 pixels as usual.)

 

Visible, El Nikkor 105mm, f/8:

 

 

UV, El Nikkor, f/8, Baader U:

 

 

UV, Cassar S, f/8, Baader U:

 

 

Stereo Anaglyph, UV, Cassar S, f/8, Baader U: