Comparing macro lenses using MTF - Part V - 2x

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Miljenko
Posts: 171
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Comparing macro lenses using MTF - Part V - 2x

Post by Miljenko »

This series of tests took many months and working hours to prepare and complete. This is where I realized that vast majority of lenses perform better when mounted on longer („relay“) lens than via extension tubes or bellows. I started with famous Rodenstock Apo Rodagon D 75mm f4.5 2x expecting results that will be a top reference for all the other lenses at 2x. Unfortunately, data that came out were dissapointing; just plain average! When using extension tubes I couldn't get more than 2258 LW/PH at center and 2363 LW/PH at the edge. Then I recalled tests I've done two years ago comparing pics from one enlarger lens mounted on extension tubes and relay lens. It was obvious that pics made with double lens setup were sharper and punchier. So I repeated that test now with different lens and different testing tools – Imatest instead of eyes. Difference was obvious: the same lens at the same magnification produced 2708 LW/PH at center which is 20% increase. However, edge resolution and CA remained the same. But this event was a sign I should look closer into extension tube vs relay lens phenomenon. To keep the long story short, ALL the lenses tested performed better via relay lenses. After comparing dozen of lenses in both ways (enlarger lenses and other kind as well), I started to test through relay lenses only.
My first lens tested at 2x that stands out concerning resolution was vintage „3M Company 8.05x“. I still don't know for which purpose was it built for, so any ID help is appreciated. Rough measuring indicates it's focal length as 50mm at fixed f/4. I have tested it just extended and it produced pretty average data: 2329/1113/0.35. But when mounted on Nikon E 100mm f2.8 lens (set to f/4), resolution jumped to outstanding 3188/2985 LW/PH. And CA dropped to lowest I ever measured: 0.006!

Image

And then happened magic find in my boss' workshop: Agfa 107mm f4 scanner lens. I managed to make thread on one end and glue the adapter on the other so after dissapointing test as single macro lens, it turned out it's fantastic relay lens. 3M 50mm lens now produced incredible 3409/3550 LW/PH, resolution exceptional even for 1x and lower. And the magnification is even sligtly larger at 2.1x. CA is worse than with Nikon 100mm at 0.19 but still very good. It was obvious sign that I have to retest many of previously tested lenses at 2x, now with a new relay lens. Please check the results below and notice remark stating sort of extension: EXT for extension rings, AGFA for Agfa 107mm, NIK for Nikon 100mm, JUP for Jupiter 37A and ROD for Rodenstock Rodagon 150mm f5.6, another great relay lens.
Thanks to that Agfa 107mm scanner lens I was lucky to find dream combo for 2.1x magnification that includes Schneider-Kreuznach Componon-S 50mm f2.8. It is the lens about which you can find very different opinions. As you can see from the table below, this lens improves in performance considerably when moving from tubes to Nikon to Agfa. I believe out there must exist other 100mm relay lenses that will perform equally good.

Image

For 75-80 mm lenses in order to get 2x magnification, you have to find good 150mm relay lens. I tried couple of ordinary lenses with 135mm focal length with mostly mediocre results. I retained only Russian Jupiter 37A who produces decent results, but the magnification with 80mm lenses is a bit short at 1.7x.
150mm camera lenses are not that common but there are many more among enlarger lenses. Rodenstock Rodagons seemed like good candidates so I managed to get two: the old one with outside aperture gear, weighting over half a kilo and the newer one, much smaller and lighter with the same f/5.6 speed. In short, the old one is better and seems to get the most from attached macro lenses. All the test results from 75 and 80mm at 2x were produced using the old 150mm Rodagon.

Image

So for 2x magnification my recommendation would be Schneider Componon-S 50mm f2.8 and some 100-105 mm relay lens with f/4 or f/2.8. It might take couple of tries to get just the right kind of relay lens to get really stellar performance.
Here is the chart with a remark: all the lenses were tested in reverse orientation except for those that are designed like real macro lenses (Tominons 35 and 50).

Image

Again, Rodenstock Rodagon 80mm f4 scored very high so it might be that I've got really good sample. And my sample of Apo Rodagon D 75mm f4.5 2x might be below average since it's performance at this magnification was mediocre. Apo Rodagon N 50mm f2.8 performed very good as expected but much cheaper, plain Rodagon 50mm f2.8 was not far behind so it is my recommendation for budget conscious photomacrographers. Dissapointing lens this time was EL-Nikkor 63mm f2.8 praised by darkroom masters who praised it big time and made my spent quite a lot for it. Well, it's resolution is decent and even over the whole APS-C frame (and probably much more), it does have low CA but the resolution figures are far from exceptional. My previous experience with EL-Nikkor 50mm f2.8 was even worse. It had about the same central resolution but dropped considerably at APS-C edges.
OTOH, pleasant surprises were again Minolta C.E. 50mm f2.8 and ancient Focotar 5cm f4.5 who performed very decently and can be bought at very low prices on Ebay.
Tominon 50mm f4.5 Macro performed ok via Agfa relay lens but not exceptional as on 1x magnification while Tominon 35mm f4.5 scored 3.3 only at f4.5 (1140/1839/0.14) so I didn't include it in final chart. It was not tested on any relay lens so I consider retesting it that way as well; it might score better.
I have included well known Russian Lomo 3.7x objective here well out of it's design boundaries. Incredibly enough, it performs great at center (third overall!) but mediocre at edges and proves once again it's kind of optical miracle.
Hoping this 2x part of testing series will give some food for thought and eventually improve your ways of using and selecting macro lenses.
Have fun.
Miljenko
P.S. Part 6 with 3x magnification will be online in a week or bit longer because these days no less than 5 microfiche lenses arrived for testing. First tests look promissing in 2x to 4x range.
All things are number - Pythagoras

ray_parkhurst
Posts: 3402
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

My first glance says that the aperture setting in these tests is fairly small, such that diffraction is going to limit performance. This brings up the question...how is effective aperture calculated for stacked lenses? I remember Lou did some work on this a while back and found significant advantages to stacking in these mag ranges due to larger effective aperture. I will try to look up that thread.

Edited to add: here is one thread discussing the EA benefit of stacking:

http://www.photomacrography.net/forum/v ... ked+lenses

In the linked thread, Lou confirms that EA for stacked lenses is f*m, rather than the smaller f*(m+1) for extension.

So for 2x, you get EA=2f with tube lens, and 3f with extension, a clear 50% advantage! You can see this advantage clearly in the performance table.

Edit #2: The improved performance of the enlarger lenses probably also has to do with their performance being better at larger enlarging magnifications, but I suspect most of the improvement at 2x is due to EA.

Edit #3: Here is another thread from Lou describing work with stacked lenses at 1x:

http://www.photomacrography.net/forum/v ... ked+lenses

Lou Jost
Posts: 5933
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Thanks Ray for mentioning those posts. I am a big fan of stacked lenses for the reasons stated. It is interesting to see these nice numerical tests!

But it remains a mystery what makes a good tube lens. Wish there were a simple test but it looks like the quality of a tube lens depends on unpredictable aspects of the front lens.

elimoss
Posts: 41
Joined: Wed Sep 12, 2018 11:31 am

Post by elimoss »

So one of the links referenced above states that Rik has mentioned in a FAQ why it is preferable to place the aperture of coupled lenses in between them. Can someone point me to that explanation?

rjlittlefield
Site Admin
Posts: 23543
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

elimoss wrote:... Rik has mentioned in a FAQ why it is preferable to place the aperture of coupled lenses in between them. Can someone point me to that explanation?
That's probably in FAQ: Stopping down a lens combo .

--Rik

jcb
Posts: 49
Joined: Sat Jun 11, 2011 2:38 am
Location: France

Post by jcb »

I am trying to get an idea of the parameters that make a possible relay lens. From the above I get :
  • relay focal length = prime focal length * magnification
    relay focal length / relay aperture > 19mm (Agfa 107mm/5.6 best aperture)
The second criterion is probably debatable and there are surely others that may be useful to screen relay lens candidates.
Any ideas ?

Miljenko
Posts: 171
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

jcb wrote: relay focal length / relay aperture > 19mm (Agfa 107mm/5.6 best aperture)
Keep in mind JCB that although Agfa 107 performed best at f5.6 when working alone, it gave it's peak at f4 when serving as the relay lens.
Best.
Miljenko
All things are number - Pythagoras

JKT
Posts: 417
Joined: Fri Oct 28, 2011 9:29 am
Location: Finland
Contact:

Post by JKT »

Your LOMO results are exactly as I have noticed - 2x is too far for it - the edges fall apart. Which brings up a question: Would you be willing to test the LOMO with magnifications 2.5, 3.0, 3.5, 4.0, 4.5 and 5.0? That should tell the useful magnification range for it and I'm likely not the only one interested.

That kind of test series would likely be very useful for those lenses not requiring tube lense ... or at least the ones that shine on some magnification. With such results it should be easy to choose a lens collection based on the required magnification range and acceptable compromise level.

jcb
Posts: 49
Joined: Sat Jun 11, 2011 2:38 am
Location: France

Post by jcb »

So this should be :
  • relay focal length = prime focal length * magnification
    relay focal length / relay aperture > 25mm
I guess that the Agfa 107mm, like the the rodagons, is a more or less symmetric lens. It seems that the telephotos are at a disadvantage in this comparison.
Based on the results with the rodagon 60mm (f4 so that focal length/aperture = 15mm), I also guess that the relay lens physical aperture may be linked to the prime lens physical aperture but this remains conjecture for now.

Miljenko
Posts: 171
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

JKT wrote:Would you be willing to test the LOMO with magnifications 2.5, 3.0, 3.5, 4.0, 4.5 and 5.0?
No problem, JKT. Once it's on bellows, retesting on various lengths is a piece of cake. I wanted to do exactly that anyway. I'm busy right now finishing 3x set of tests, but I might switch to Lomo3.7 some time next week.
Best.
Mike
All things are number - Pythagoras

JKT
Posts: 417
Joined: Fri Oct 28, 2011 9:29 am
Location: Finland
Contact:

Post by JKT »

Miljenko wrote:Once it's on bellows, retesting on various lengths is a piece of cake.
That's what I suspected. Thanks in advance!

Post Reply Previous topicNext topic