At the LHSA show in San Antonio, I had a discussion with a gentleman who claimed that he tested the latest model 90 Pre-aspheric against the current 90APO aspheric, and found the former to be sharper. Poppycock, I thought to myself, there is NO way that is even possible. (Many of you already know I am a huge supporter of the 90 SAA given its astounding sharpness. I am also a fan of the 90TE because its compact size, very light weight, yet still very good performance.) At any rate, I queried him a bit further on his testing methods, doubting his veracity. He claimed he used an Air Force test target. Yeah, right. I let it go. Fast-forward to last week when Andy Piper posted on his newly acquired 90 Pre-aspheric, and also obtained very positive, and IMO credible, results. Well, I couldn’t stand it and had to know for myself, so I gobbled up a very clean 90Pre-A for myself to test – the most recent version prior to the 90APO, with the E55 filter size. Here is what I found out, and I apologize in advance, as I still do not own a slide scanner so have no way to post the images...

First some general observations. The 90AA and 90Pre-A look virtually identical in size and shape. In fact, I had to look for the “APO” engraving on the lens shade to identify the 90AA. However, while both are the same physical shape and size, the 90Pre-A is notably lighter in weight. Obviously the diminutive 90TE is the hands-down winner in the size and weight department. Focus tension on all the lenses is similar, being stiffer than shorter M lenses I own. Helical throw is about 30 degrees less on the 90AA than the 90Pre-A, so the 90AA will focus faster but not as finely. I have not yet decided which I think is better.

* Center performance: I got a surprise here. When compared at the same apertures I could see very little difference on the test target between the three lenses. I could only pick out the TE from the others at f2.8. In the still life I shot, which was lower in contrast than the target, I could just barely detect the difference at f2 for the Summicrons and barely distinguish the three lenses at f2.8; the 90AA being the best and the 90TE being the worst. I want to clarify that I had to look very closely under the loupe to distinguish the difference in center sharpness, so IMO it is probably not significant for most normal shooting situations. * Horizontal edge performance (12mm from center): Interestingly, the lenses performed almost as they did in their centers. On the target, the 90AA was just slightly ahead of the 90Pre-A at f2 and f2.8, and at f2.8 the 90Pre-A was just slightly ahead of the 90TE. At f4 and above I could not distinguish any significant difference between the lenses, and they all performed excellently. On the still life, I could not tell any differences when comparing the lenses at the same apertures anywhere along the 12mm-out section.

* Corner performance: Here is where the new lens design shined, and it clearly outclassed the 90Pre-A and the 90TE up to f4 and was slightly better at f5.6. Surprisingly, the 90TE performed better here than the 90Pre-A at f2.8 and f4, but they were indistinguishable at f5.6. At f8 and above all lenses were indistinguishable.

* Bokeh: While I realize bokeh is very subjective, I’ll share my opinions on it for those that are interested. The 90TE was a little clumpier than the other two, but still very decent, and the 90Pre-A seemed a tad smoother in the transition than the 90AA. So, all in all I’d give the 90Pre-A the nod, but all were very good.

Conclusion: All lenses performed very well at the center at all f-stops. At f8 and above the lenses are essentially indistinguishable across the entire image. At f4 and f5.6, the 90AA is a slight bit better than the other two at the very corner, and very slightly better at the horizontal edge. At f2 and f2.8 the 90AA is clearly better, but only in the corner. I cannot help but ask myself if the extra cost of the 90AA is justified by its significantly better performance only in the extreme corners at the wider apertures. I think not, since I rarely place subjects at the very corner of my frame when shooting the 90 at any aperture, let alone f2. Not to mention the fact that the 90AA cost me about as much as the two older designs combined. I bought the 90AA specifically for its wide-open performance, but I think it’s headed for the auction block. As good as it is, IMO it is just not enough better than the others to justify the extra cost for my uses. With the money I get from the 90AA, I’ll probably buy a fairly good film scanner and start showing you guys some of my stuff... I can hear the groans (and feel the flames) already ;-)

-- Jack Flesher (jbflesher@msn.com), November 29, 2001

Answers

Jack,

Thanks for this effort. You are helping quite a few people save a lot of money. What is the magnification of your loupe? For all practical purposes, ultimate lens quality is desirable when one goes for very big enlargements. If at f2 old and new generation M 90 'crons behave essentially equally, I do not have any reason to spare cash for the coming '90 R apo-asph....;-) We do need user based "non-Putts" comparative tests, if only for sanity sake. Thanks again!!!!

-- Jacques (jacquesbalthazar@hotmail.com), November 29, 2001.

"Conclusion: All lenses performed very well at the center at all f- stops."

Thanks for posting that. Judging by various informed opinions and lens tests, *all* recent Leica M lenses perform very well. There are differences that can be detected by rigorous testing but often the user can only rwally benefit from them under ideal conditions (tripod + fine grain film).

I feel sure that, if you'd been able to include the current model 90mm/2.8 Elmarit-M in your test, you'd have found its performance to be even close to the 'crons in most respects. It's an economical alternative for those who don't need f/2 and is particularly good at close distances because of its flat field. (I'm biased, you understand, being a 90mm Elmarit buff!)

-- Ray Moth (ray_moth@yahoo.com), November 29, 2001.

Jack: Welcome to the club!

My impressions are exactly identical, except that my 'cron seemed better than the TE in the corners at 2.8 - but we all know YMMV when it comes to these tests. Both my 'cron and TE have the longer focusing throw, which I do think helps accuracy.

Before you sell, check for two things with the SAA - less color fringing near the corners and a little more contrast at f/2. Personally I'm happy with the non-AA, but these other small improvements MAY be important - or not.

I think there's a pretty clear pattern emerging. For the most part the last pre-APO/ASPH lenses were designed by or under the supervision of Walter Mandler in Midland (c. 1960-1980). Particularly when you look at the fast lenses (90 'cron, 75 'lux, 50 f/1, 50 'cron, 35 'cron, 21 Elmarit) his (or Leica's) philosophy was to get the absolute best possible resolution at the center wide-open, even at the cost of some corner resolution. Erwin makes mention of "designing for the photojournalist". In the process the aberrations left near the corners, along with other factors, promote smooth 'bokeh'.

For that matter, even the 400 Telyt-R/Viso (designed mid-60's) follows this pattern.

The new generation of APO/ASPH lenses can't get a lot better in the center, because Mandler already did such a good job with the glasses and spherical surfaces he had available, but they do bump up corner performance a lot and wide-open contrast a little. (The 35 'lux ASPH is an exception - a big improvement overall over the 1960 design).

If you need the corners sharp at f/2, you want the newer lenses. If you can live with some softness at the edges, the pre-APO/ASPHs offer a lot of center performance and some pretty nice OOF backgrounds - at a discount.

For folks in general: I think it's critical to test ANY fast tele on your own M body, and if possible try several units, even of the same vintage. My new (old) 'cron focuses much more consistently than the one I let go previously (later barrel, same optics) on my bodies. Given the narrow depth of field with a 75 'lux or 90 'cron, even small production variations in the body or lens (remember those tiny 'correction' numbers on the focusing ring?) can be enough to throw out the focus.

-- Andy Piper (apidens@denver.infi.net), November 29, 2001.

For those interested, links to a 90 Summicron pre-APO/APSH portrait to show overall image quality, and an enlarged detail of same to show sharpness. (1/1000th @ f/2, Pan F, DDX normal)

-- Andy Piper (apidens@denver.infi.net), November 29, 2001.

Well, if that's the pre-asph at f2 it looks like I shouldn't have bothered getting the newer one (could have got a 50 summicron with the spare cash). For what it's worth I think there is a bit of a danger in thinking that if it's not the most recent generation it's not good enough - that's a trap I certainly fell into at first. It's also certainly the case that the "improvements" are not always going to benefit you as a photographer. For example, the flatness of field and into-the-corner sharpness of the 90apo can actually look a little bit odd on a head and shoulders portrait - the slight fall off of the 75 actually looks nicer for this particular task. Also, to return to the running theme of tonality, I still reckon the 90 elmarit over any of the others...Just thought I'd chip in.

-- stephen jones (stephenjjones@btopenworld.com), November 29, 2001.

Stephen, I too still reckon the 90 Elmarit over any of the others -- and I think I have already tried them all out -- I just didn't say this above anywhere because I thought it was just all about 'crons.

-- Michael Kastner (kastner@zedat.fu-berlin.de), November 29, 2001.

I sold my pre-APO 90/2-M several years ago, and bought the 90/2.8 Elmarit-M and have never regretted the decision. I also sold a 75 after a short time, and in both cases it was because of the size and weight of the lenses which for my purposes run contrary to why I use the M in the 1st place. The 90/2 was a great lens, but f/2 showed noticeably less contrast than f/2.8. OTOH I've got both 90/2 and 90/2.8 for the R system and the 90/2, though similarly less contrasty at f/2, gets used most often. The reasons are finder brightness and the fact that I've got the 80-200/4. I am also a fan of the thin T- E, mine is the short-throw version (late Canadian) and I really like carrying that lens even if it's performance doesn't quite equal the current version.

-- Jay (infinitydt@aol.om), November 29, 2001.

Thank you Jack!

Posts like this are worth their weight in Cash for us new-comers to the Leica M-system.

What distances did you photograph at?

-- Peter Olsson (peter.olsson@lulebo.se), November 29, 2001.

Jacques: The loupe I used for the comparisons is a 16x Peak scientific, and I realize that may not be enough magnification to adequately distinguish minor differrences, but I also think it gets me in the ballpark. Film was Fuji Provia 100F, and camera was on a tripod, of course.

Andy: Yes, there was a very slight difference in contrast wide open, but truthfully in my sample, I did not think it very significant. Surprisingly, I saw no color-fringing either, yet I was expecting it! Perhpas in a very high-contrast situation it will become noticeable. The 90TE had notably lower contrast than the others at f2.8, but it was still not objectionable.

Ray: No doubt the 90 Elmarit would have fared very well in this test too. And I think it is safe to assume the 90E would out-perform my 90TE. But since my 90TE performs so admirably I have never felt the need to get a standard Elmarit. If I did not want f2 in a 90 on occasion, I would be perfectly satisfied with my 90TE.

Jay: The size argument in favor of the Elmarit over the Summicron has never made sense to me. In the current black versions, the 90E is 90 grams lighter then the 90AA, 2mm shorter, and 7mm narrower at the front -- IMO, not a whole lot of savings on any dimension, but perhaps it is a notable difference in the bag or on the camera... By comparison, the 90TE is miniscule -- untill you mount its shade ;-)

-- Jack Flesher (jbflesher@msn.com), November 29, 2001.

Peter: The targets were shot at 50X the focal length, or 4.5 Meters.

-- Jack Flesher (jbflesher@msn.com), November 29, 2001.

Jack:

Nice review of the lenses! I have the pre-ASPH 90 Summicron and have always been impressed by the quality of the slides shot with it at full aperture. As a result, I have never seriously considered buying the 90 AA.

-- Muhammad Chishty (applemac97@aol.com), November 29, 2001.

Jack, the size difference between the 90/2.8 and 90/2 is more noticeable than the dry statistics would indicate. The difference in diameter means your fingers aren't blocking the rangefinder windows especially in portrait mode. The focus ring moves much easier on the 90/2.8. The E46 filter is common to the 28/2.8, 35/1.4ASPH, 50/1.4 and last-version 135/4 so unless you have a 21ASPH, 24 or 1st-version Tri-Elmar you don't need a set of the larger filters. Finally the 90/2's are nose-heavy whereas the 90/2.8 is better balanced on the M body. Since the performance from f/2.8 is virtually identical on all 3 lenses, it really boils down to whether your shooting style demands f/2. If not, then the excesses of the 90/2 lenses are for naught; if so, then they're inconsequential.

-- Jay (infinitydt@aol.com), November 29, 2001.

Muhammed: Astute of you to not rush into the newer-is-better trap -- obviously wisdom at work! (I'm learning!)

Jay: I agree. But, the 90Pre-A and current 90E currently sell for about the same amounts used -- so IMO it really boils down to a size vs speed (vs filter size) issue...

-- Jack Flesher (jbflesher@msn.com), November 29, 2001.

Jack, Thank you for your observations.I have resisted parting with my pre-aspheric 90mm 'Cron black heavy weight version for many years.I note your testing was carried out at 4.5 meters.For my part with closer focusing,head and shoulders shots,even at f5.6,a smoother rendition is seen.Would the compact pre asph. 90mm be sharper than mine in general or more specifically close up? When I'm in an f2 mood and need accurate metering the 90mm 'Cron head with viso. focus mount on the SL2 provides a satisfying setup. Regards, Sheridan Zantis.

-- Sheridan Zantis (albada60@hotmail.com), November 29, 2001.

I did some similar tests when I first got my M a few years back. I came to the conclusion that all the 90's I tried (including the 90 Rokkor) were exceptional, and the old cliche "splitting hairs" was appropriate. I first had the TE, but mine ended up with inner element trouble that caused it to flare like I was using a special effect filter. I then tried out an older style Elmarit and Summicron at the same time. The older 90 is just way too heavy for an M camera in my opinion. The older Elmarit was suprisingly good, and I would have probably been happy to stop there, but the one I found had been dropped and had a kink in the focusing, so I sent it back. I then got a deal on the current 90 f2.8, and was blown away by the image quality especially at f2.8 where it is unbelievably good. Flare control is excellent as well. I later got a 90 Rokkor with my CLE outfit, and I liked the tiny size, silky smooth focus, and excellent image quality, but found f4.0 to be too limiting an f stop on a focal length I seem to need to use at f2.8 often.

-- Andrew Schank (aschank@flash.net), November 29, 2001.

Where does the 90mm Voigtlander (that Erwin recently liked) fit in? At less than $400, is it maybe a good lens for someone who wants to see if he or she likes a 90mm on an M? Or maybe it's good for travel, at 260 grams (about the same as a 50mm Summicron). Erwin didn't say anything about bokeh, so I'm interested to hear about that. I'm tempted, since my longest lens is still a 50.

-- John Morris (jtmorris@slb.com), November 29, 2001.

Jack; thanks for your 90´s tests, I liked more than Put´s, and I am a 90/2 preasph owner and less user, but I realy like the size and shape of this lens, and of course is so glad to know it is a such performer, I have never had the oportunity to compare to another one, well to my 90/4 collapsible, but that doesn´t count, they are so diferent. and now the only time I use the 90 is when I drive, to have something longer to reach images I can´t aproach any other way.

-- r watson (al1231234@hotmail.com), November 29, 2001.

John: I have never tested the 90 VC, but Paul did a while back. As I recall, it did not fair all that well wide open. Also, for $400 you can find 90TE's in decent shape. If you can find a 90TE that is clean (ie; little internal dust, and has none of the dreaded Leica fungus or element separation) then I think it would be a *much* wiser purchase than the VC lens, IMO. Note: the 90TE is known to be flare- prone, but I have not had any problems yet with mine yet.

-- Jack Flesher (jbflesher@msn.com), November 29, 2001.

Roberto:

Thank you for the kind compliment! My tests are no where near as scientific or technical as Erwin's, but I think they work for those of us who use our Leicas in the real-world!

Happy to share this information with all on the forum!

PS: If anyone has a burning desire to try out a 90APO, contact me off- line -- I have a LNIB one for sale ;-)

Cheers,

-- Jack Flesher (jbflesher@msn.com), November 29, 2001.

My test of the Colander APO-Lanthar is invalid. I've discovered that my example has serious focussing errors. Once I get it back from the shop I'll re-test it. I love the size and the construction, and would dearly love a lighter travel companion than the Summicron.

-- Paul Chefurka (paul@chefurka.com), November 29, 2001.

Jack- Add my own thanks for the tests. That is really why we put this site on our "Favorites" list. What I would find interesting would be tests of old camera lenses, like the 44mm Ektar of the Kodak Signet 35; the f2 Ektar of the Kodak Bantam Special; the f3.5 Ektar of early Kodak Retinas; The Argus C-3's Cintar f3.5. My understanding is that APO lenses correct for three aberations and that the next level would be the "Superchromat," which corrects for 4 aberations. I asked the Leica chief in NJ if we might on e day see a Superchromat, but he did not answer the letter. You can imagine a firm such as Leica trying to answer all the questions that appear on a site like this!

-- Frank Horn (owlhoot45@hotmail.com), November 29, 2001.

For Frank,

Strictly speaking, apochromatic lenses correct for only one type of aberration: ie. chromatic aberration. A lens is considered to be apochromatic if all three primary colors of light (red, blue, and green) are brought to focus at the same place. Achromats bring two of the three primary colors to the same focus.

Apochromatic correction does not necessarily imply perfect correction for other (non-chromatic) errors, such as spherical aberration, distortion, coma. However, Leica's stated policy was that they would not call a lens "apochromatic" unless it met several criteria: 1) correctly focussing all three primary colors of light; 2) a high degree of correction of the secondary spectrum; and 3)a very high degree of correction of the non-chromatic errors such as spherical aberration.

I guess to do this, they had to utilize an aspherical surface for the 90/2.0 Apo-Summicron-ASPH, whereas the 135/3.4-Apo-Telyt, with its more modest aperture, did not require an aspherical surface.

I'm not sure what a "superchromat" means, unless it is utilized to denote a lens in which the primary and secondary (magenta, cyan, yellow) spectrum are all brought to the same focus.

-- Eliot (erosen@lij.edu), November 29, 2001.

For Frank, Eliot and other interested parties: (slightly off-topic)

Carl Zeiss' definition of superachromat.

For Jack: so when you get you R, are you going to wait for the R- version of the 90 SAA to come out?

-- Andy Piper (apidens@denver.infi.net), November 29, 2001.

Andy:

Not likely! Think I'll jump on one of those junkie 80 f1.4's...

-- Jack Flesher (jbflesher@msn.com), November 30, 2001.