Comparing macro lenses using MTF - Part VI - 3x

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

Moderators: Pau, rjlittlefield, ChrisR, Chris S.

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Comparing macro lenses using MTF - Part VI - 3x

Post by Miljenko »

As 2x magnification completely changed lens list and order compared to 1x chart, 3x testing introduced whole bunch of new lenses while eliminating many „goldies“ from previous tests. In general, all enlarger lenses dropped out of top eight 3x lens chart. I did list them just because many photomicrographers use them at this magnification but you'll realize how unsuited they are for 3x and over. The best enlarger lens in this list at number 9 is Schneider-Kreuznach Componon-S 50mm f2.8 relayed with vintage Rodenstock Rodagon 150mm f5.6. When stopped down to f5.6 it produces sharp pictures measuring 2240 LW/PH at center and 2411 LW/PH at the edges. Good but far from outstanding.
As shown many times in this forum, various industrial, microfiche and scanner lenses can provide exceptional sharpness and contrast. Unfortunately, frequently the origin, use and manufacturer of such lenses are hard to detect. And worse yet, those „golden samples“ can appear just once on Ebay so knowledge about it's performance is of very limited value for all except for the lucky buyer.

Image

This sequel of my lens test series comes a bit delayed because of five microfiche and one scanner lens that arrived at the last minute so it took couple of days to perform all the additional tests, analyze and include data into already finished charts. As you'll see soon enough, it was worth waiting.The main obstacle when testing those industrial, scanner and microfiche lenses is complete lack of any threads. So my 3D printed adapters I introduced here were very helpful aids: https://www.photomacrography.net/forum/ ... highlight=
I had to print couple of them for different lens diameters and filter frames depending on relay lens threads. And yes, relay lenses again produced better performance than plain extension tubes. A problem here was finding proper shorter tube lenses for those tiny things since their focal length was between 15 and 35 mm. One of the harder requirements was protruding front element that provides optically close mounting. If mounting further from relay lens vignetting often occurs making some otherwise great tube lenses useless. For instance, many standard camera macro lenses have very recessed front element. Fortunately, couple of enlarger lenses in 50 to 150mm range on hand were just the right stuff for that use.

Image
Relay lenses used: TOM50 = Tominon 50mm f4.5, ROD60 = Rodenstock 60mm f4.5, EXT = extension tube only, ROD80 = Rodagon 80mm f4, ROD150 = Rodagon 150mm f5.6, JUP37A = Jupiter 135mm f3.5, POR240 = Porst 240mm f4.5, AGFA = Agfa 107mm f4

Image

Fortunately for me and unfortunately for anybody else is that two best results were achieved by rare lenses of unknown use and design. I suspect Bell & Howell „24x“ lens was used in microfilm reader based on it's external shape. OTOH, UNIC 16.6mm f3.0 has no distinct shape, protrusions or threads so I don't have a clue where and when it was used. Placed third is 3M Company „8.05x“ lens which also could be kind of microfilm reader optics. The very same one is currently offered on Ebay. Those 3 lenses performed best when paired with respectively: Tominon 50mm f4.5 (vintage Polaroid macro lens), Rodenstock 60mm f4.5 (miniature lens with 20mm thread) and old Rodagon 150mm f5.6, of course all of them extended for infinity focus.

Image

Not very far from those 3 is notorious Lomo 3.7x Russian objective producing solid 2670 LW/PH central resolution which drops to still respectable 2010 LW/PH at the edges. Very good results concerning 3x magnification which is out of it's design boundaries. Added to still affordable Ebay price and no need for a relay lens, this is one of the best lenses available in 3x to 4x magnification.

Image

Biggest disappointment came from Tominon 17 and 35mm macro lenses. I was obviously expecting to much from those vintage lenses Tomioka designed for Polaroid technical camera in '70s. It might be, however, that I've purchased below average samples. Both were tested using all available aperture values and those stated gave the best figures.

Image

Somewhere in between are four microfilm lenses without any manufacturers sign (except for „Made in Japan“ sticker) and just numerical designation in 3690-5xx format. I have tested it with and without relay lenses in magnifications from 2x to 5x. The best I could get at 3.2x was 2446/2514 LW/PH from 25mm f3.5 sample when relayed via Rodagon 80mm f4. When paired with Rodenstock 60mm f4.5 magnification drops to 2.4x while central resolution raises to excellent 3480 LW/PH and usable 2250 LW/PH edge value. I believe those four lenses have greater potential so I'll retest them later with other relay lenses / magnifications / settings.
Since I don't have many 4x and 5x lenses and initial test with other lenses at those magnifications didn't look promising, I'll skip that part for now. So the next week comes part VII of my test series with 6 tube lens shootout: 200-210mm paired with 5x infinite Nikon objective.
Enjoy.
Miljenko
All things are number - Pythagoras

ChrisR
Site Admin
Posts: 8580
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Fascinating stuff, thanks for posting.

If your 3D designs are available online, perhaps they could be added to Saul's list in the FAQ section?
Chris R

RobertOToole
Posts: 1723
Joined: Thu Jan 17, 2013 9:34 pm
Location: United States
Contact:

Post by RobertOToole »

Very interesting work Miljenko.

It was especially interesting to see the poor results with the Tominon 35 and 17 lenses. Previously I had two or three poor samples of the 35mm f/4.5 and for some reason I decided to try another, probably due to the low prices. Thanks to google I found a new old stock unit complete with box and Styrofoam. This sample was fantastic matching or beating the Canon MP-E 65, Olympus 38mm, Canon MP-35, and some other highly regarded performers.

The 17mm? I owned three before I found a really sharp sample (from the same seller as the 35mm).

Its not just Tominon I have found, I owned 4 Schneider 45/4 APO Componons and one unit was so soft I was forced to toss it in the garbage, it was too soft to resell.

On the lensRentals blog they posted an article about performance variance between samples. With fast 50mm primes they found more sample variance between units of the same brand then they did between brands!

BTW they also reported in another post that Canon was best in terms of tight image quality control, sigma was second ahead of Nikon, and M43, especially panasonic was the worst, they were either great or garbage. Zeiss and Leica also produce lemons in each batch.

Best,

Robert

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

Post by elimoss »

Looks like the Tominons @5.6 on extension would be working at an aperture disadvantage of at least 1 stop relative to the stacked combos and the Lomo (at less than normal extension). So it should not be surprising that it resolves less, although this seems like more than we might expect.

Perhaps the tominons would perform better stacked, but not in a reversed direction, since they are already reverse designs...?

ray_parkhurst
Posts: 2810
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Miljenko...I was hoping to see some curves like you showed for the 63mm EL-Nikkor in your introductory thread on MTF testing. I really liked the format as it pulled together all the relevant info I look for into one graph. Are you planning to publish these graphs at some point? This would be an order of magnitude more useful to me than just a summary ranking. Even if you haven't made such a graph for all the lenses tested, the ones you already have would be extremely useful, especially for the top performers in each of the categories. TIA...Ray

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

ChrisR wrote:Fascinating stuff, thanks for posting.

If your 3D designs are available online, perhaps they could be added to Saul's list in the FAQ section?
I believe I have given enough guidelines for succesful designing and printing lens adapters. Every industrial lens has different diameter and they are going to be interfaced to different filter rings so my actual adapter probably won't suit anybody else. If you feel ilustrated step by step manual would be more helpful, I can do that in a near future.
Best.
Miljenko
All things are number - Pythagoras

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

RobertOToole wrote:Very interesting work Miljenko.
On the lensRentals blog they posted an article about performance variance between samples.
Best,
Robert
Thank you, Robert. And I'm very well aware of Roger's thorough lens tests and comparisons. That's why I quoted him in 2011 when he started doing that. Five years after I published my take on the subject. It was in 2006. when I took five identical Sigma lenses, tested them using Imatest and published findings on Dpreview lens forum. Thread soon exceeded 150 replies and was extended to second thread. The other outcome was more serious: Sigma managers emailed Sigma representative in Croatia protesting against their kindness to borrow me five lens samples. In Dpreview thread readers confirmed their experiences with lens sample variations. There is a trace of this case in one of my Pbase galleries: http://www.pbase.com/miljenko/image/58520006/original
Best regards,
Miljenko
All things are number - Pythagoras

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

elimoss wrote: Perhaps the tominons would perform better stacked, but not in a reversed direction, since they are already reverse designs...?
Unfortunately, Tominons 17 and 35 have casings which are actually extension tubes so putting them onto relay lenses produced severe vignetting. :-(
All things are number - Pythagoras

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

ray_parkhurst wrote:Miljenko...I was hoping to see some curves like you showed for the 63mm EL-Nikkor in your introductory thread on MTF testing.
I perfectly understand, Ray, That was my idea at the beginning, but after dozen of tests performed in such thorough way, I simplified the procedure for two reasons: 1. It takes enormous amount of time for each lens to do it that way (about two weeks, two hours a day) and 2. After 6 months of testing enlarger lenses I noticed a pattern concerning the aperture. The useful range is f4.8 to f6.8 without exceptions. So there was no need to go below and above those figures. This procedure change saved me huge amount of time and provided for testing much larger number of lenses than planned at the beginning. Of course, I can easily generate colorful charts for those lenses having full set of data. Unfortunately, most of those didn't score the highest performance so I see no huge gain in publishing all those charts. Actually, those chart could confuse those less informed. Data I provided are selected as the best figures respective lens has produced during the test.
Or maybe I see it the wrong way? Please elaborate.
Best.
Miljenko
All things are number - Pythagoras

Lou Jost
Posts: 4732
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

By accident I see that I bought a Unic 16.6mm f/3.0 on ebay for $14 on Sept 21, the week before you posted this. I was looking for cheap weird lenses. So they do come up now and then!

ray_parkhurst
Posts: 2810
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Miljenko wrote:
ray_parkhurst wrote:Miljenko...I was hoping to see some curves like you showed for the 63mm EL-Nikkor in your introductory thread on MTF testing.
I perfectly understand, Ray, That was my idea at the beginning, but after dozen of tests performed in such thorough way, I simplified the procedure for two reasons: 1. It takes enormous amount of time for each lens to do it that way (about two weeks, two hours a day) and 2. After 6 months of testing enlarger lenses I noticed a pattern concerning the aperture. The useful range is f4.8 to f6.8 without exceptions. So there was no need to go below and above those figures. This procedure change saved me huge amount of time and provided for testing much larger number of lenses than planned at the beginning. Of course, I can easily generate colorful charts for those lenses having full set of data. Unfortunately, most of those didn't score the highest performance so I see no huge gain in publishing all those charts. Actually, those chart could confuse those less informed. Data I provided are selected as the best figures respective lens has produced during the test.
Or maybe I see it the wrong way? Please elaborate.
Best.
Miljenko
No worries Miljenko. I was just expecting based on your original post that you had already put the work into doing the charts with truncated range of f4.8 to f6.8.

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

Lou Jost wrote:By accident I see that I bought a Unic 16.6mm f/3.0 on ebay for $14 on Sept 21, the week before you posted this. I was looking for cheap weird lenses. So they do come up now and then!
Lou, I didn't write much about the best 3x lens in this shootout because I thought no one knows about it. I'm glad you caught it and can enjoy the "3x" winner. Now it makes sense to tell a little bit more about it. I have tested it with no less than 7 lenses because it was so promising from the beginning. There are basically two problems with this lens: 1. It has to be very close to relay lens front lens element. If moved just a couple of millimeters from it and vignetting occurs. 2. Working distance is short and this makes lighting a challenge because of relatively large diameter. It gets even shorter with 4x-5x magnification which is a shame 'cause it performs equally good at that level.
You should try it with 50mm relay lenses with protruding front element. I had at hand vintage Meyer Optik Trioplan 50mm f2.9 which provided very good performance (2959/2686 LW/PH) when stopped down to f/4. Relayed via Tominon 50mm f4.5 the edge resolution gets a bit down (2567 LW/PH) but the center res goes up to stellar 3100 LW/PH). At 3.6x resolution is still pretty good at 2380/2300 LW/PH but the working distance gets below 10mm.
Anyway, I'm sure this lens will give lots of great shots at 3x.
Best.
Miljenko
All things are number - Pythagoras

Lou Jost
Posts: 4732
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Thanks very much for the extra information, Miljenko. Yesterday I ordered the 50mm Tominon as tube lens. I will watch for the Meyer Trioplan.

I wonder if the working distance can be improved by using an additional relay lens in front of it. This is what my 85mm Repro-Nikkor f/1 is designed for.

Miljenko
Posts: 170
Joined: Sat Jun 01, 2013 11:53 pm
Location: Zagreb, Croatia
Contact:

Post by Miljenko »

Lou Jost wrote:I will watch for the Meyer Trioplan.
You can't go wrong with Meyer Trioplan, Lou. It is a vintage lens used on Altix rangefinder cameras in 60's, later rediscovered as "soap bubble bokeh lens". You can see some everyday pics done with this lens here: http://www.dyxum.com/dforum/samples-mey ... 24014.html
It is so good that recently crowdfunding was raised for reviving this design: https://petapixel.com/2016/04/14/meyer- ... ble-bokeh/
If you find vintage Altix version you'll just need M42 adapter and fine file to slightly modify it's mount and screw it onto the adapter.
Best.
Miljenko
All things are number - Pythagoras

Post Reply Previous topicNext topic