Misc LensTest

This test are collected from different places on the Internet.

I never did any resolution tests on my lenses. I don´t have the time... and I don´t need it. I do art, not scientific work... a lens "looks" sharp and contrasty to me or not... I evaluate by looking at the pictures... I own this lens since 1997 and earn money for living with it... As a pro, mostly in fashion/show-biz portraiture, I do nearly 90% of my work with my Hasselblad Medium Format cameras ( thinking about the new H1 system or Contax 645 system last days... :-)... ) and the remaining 10% with a Linhof Kardan LF beast or my F5/F90X/FM2 35mm "friends"... The reason why I wrote this: the AF Nikkor 85mm/1.4D IF is the lens, which can be used for shots made for Covers of Magazines (Elle... GQ... Harper´s Bazaar...) without any doubt. And I am used to Zeiss, Rodenstock and Schneider-Kreuznach lenses... :-)... It happened to me many times, that fellow photographers didn´t believe, that the Cover is not from my Hasselblad/Zeiss equipment... Celebrity shots made with AF 85mm/1.4D around f:4 and Elinchrom Style Studio Strobes are saleable and printable with no problems... and I often do Portraits with f:1.4 and the resulting smoothness is very pleasing, but the eyes ( or the one eye, which was used for focus ... :-) ... ) are still dangerously sharp. The Contrast is still very good and the coulours are simply beautiful. Slight, but visible vignetting occurs between f:1.4 and f:2, but I got used to that and make it a part of the working style... and this lens has unnoticeable linear distortion... if any. And one of the most important things: bokeh is very nice, maybe not as good as the best portrait lenses for Leicas, but definitely very good = smooth = not distracting! My only complains: why Nikon could not make the sunshade a bayonet-type instead of a "screw-in"? ... Every time I want to attach the lens cap "at the end of the day", I have to unscrew the shade... and this loosened the black painted metall front part of the lens barrel after few years! This is the only weak point on my otherwise sturdy AF Nikkor 85mm/1.4D IF... and there is some dirt behind the last glass element at the "end" of the lens, which is a plain one (but Multicoated), only for protection I presume, which can be easily removed and cleaned, thanks god ( or Nikon ... :-) I developed this rules for myself, when I am in full controll of the light: f:1.4 Female portraits, silky smooth, "dreamy", minimal depth of field, "hides" skin blemishes... f:2 Mostly I "switch" directly to f:2.8 ... :-) f:2.8 Male portraits, Everything with background out of focus but demanding superb sharpness and contrast... fashion on location... fashion stories... editorial work - fullsize DIN A4 ... f:4 Covers and "the entire page" images with controlled background in the studio... but still with little depth of field... f:5.6 Covers with controlled background in the studio... when maximum depth of field is needed... f:8 I use this aperture only for my private work, when largest possible depth of field is needed... like some architectural details, when the 35mm system is the only one I have at the moment with me... (travels)... I never use f:11 and f:16 on this lens due to expected limitations from diffraction... there are another lenses I use for such shots... not this "f:1.4" beauty... Of course, "breaking rules" sometimes creates extraordinary pictures... but these could be some basic rules... = for example: I would not normally shoot a male portrait at f:1.4... he would look too "soft/smooth"... :-) The sharpness/resolution of this lens around f:2.8-4 is already so good even in the corners, that you can "cut-off" the hair the way used in fashion photography (... :-) ...) getting eyes of the model so close to the border of the image with no fear the perceived final sharpness of the whole photo will suffer at all, enabling pretty weird compositions and still be able to print the final image big as it gets from a 35mm format... Sometimes I mix Medium format with this lens on a F5 during a session and I can "join" them in a 4-double-page-spread story without an apparent difference in the images clearly showing: "This is 120 and this is 35mm"... it looks pretty "uniform" but allows me to do dynamic shots I never get with my "slow" Hasselblad bodies... This lens is really exceptional and with no doubt among the best lenses Nikon ever made for the F mount... If you shoot portraits in 35mm with a Nikon SLR, You should definitely have a look at this lens together with AF Nikkor 105mm/2D DC, 135mm/2D DC, 180mm/2.8D ED... :-) ... a real pro lens, comparable to portrait lenses around 85mm by Leica and Zeiss (Contax)... All the best! Paul (ps6190@web.de)

Date: Tue, 4 Jun 2002 15:51:19 +0200 (CEST) From: Olle Bjernulf Subject: Evaluation of 28/2 Ais vs 28/2.8 Ais I have both lenses and did a test. Genernal sharpness/contrast: Both lenses are good in center wide open - about equal. 28/2.8 is a little little better in center at 5.6 - 11. Corners are not good until stopped down 2-3 stops on both lenses. Corners are about equal on both lenses, maybe a little bit better on 28/2. At F22 the performance is not so good on both lenses. The differnece between F22 and F16 is much greater than the difference between the lenses in center at F5.6 - F11. Light fall-off, both lenses: wide open: severe light fall-off, unacceptable one stop down: very little two stops down: gone Distortion: very low on 28/2, almost non-existing on 28/2.8 Conclusion: None of the lenses is really useful wide open or at F22. 28/2 is very good at F2.8 so it is still one stop faster. The performance at F4 - F16 is so similar that I wouldn't pick one lens over the other. I will keep the 28/2 and sell the 28/2.8 mainly due to brighter viewfinder. Olle bjernulf Uppsala, Sweden

Date: Fri, 8 Jun 2001 13:35:43 -0500 From: Larry Kopitnik <kopitnil@marketingcomm.com> Subject: 45 f/2.8 P Nikkor: It's a Gem I like normal focal length lenses. For years I shot with just an FM and 50 f/1.4 Nikkor, and when that's all you use, you learn to pre-visualize scenes in that perspective. Even today, when using a 35-70 f/2.8D Nikkor, I find myself shooting it around 50 mm more often than any other focal length. But several years ago I sold my last 50 mm lens. Closed down a bit it was sharp (one photographer who shot with it and the latest $2000 Leica-R 50 f/1.4 told me that he found the Nikkor to be sharper). But other characteristics of photos it delivers -- compromises the lens designers made to achieve that excellent sharpness at mid apertures -- displeased me: harsh, not gradual, tonal gradations; harsh out-of-focus areas, often with doubled lines I found distracting; an inability to hold detail in shadow areas. These are not important aspects of a photograph to many photographers, and for them the 50 mm Nikkors (I've used the 1.2, 1.4 and 1.8 lenses) are outstanding tools. But they caused me to move to the 35-70 f/2.8 and 35 f/1.4 Nikkors (the 60 mm Micro-Nikkor is also very good). So I was excited to read that Nikon was releasing the 45 mm f/2.8 P Nikkor. Tessar optical designs, which this lens is, typically render the look of photos I prefer while maintaining good sharpness. Additionally, a small, light lens to carry around on the camera is something I'm always looking for. When I saw B&H had the lens available, I ordered one. When you take the lens out of the box, the first thing apparent is how small it is. Put it on a camera body, and it extends just 5/8" from the lens mount -- about half the depth of the 50 mm f/1.8 AF Nikkor. It will easily slip into a pants pocket. Its mechanical quality is superb, with metal construction, and all numbers and words engraved, not just printed on. The filter threads are metal (not even the super big and expensive 80-200 f/2.8 AFS Nikkor has that). Manual focus feel is perfect, and quickly makes one remember how loose the manual focus of so many AF lenses, particularly around this focal length, is. The only aperture scale is the one used for ADR (aperture direct readout); the lens isn't big enough for two scales, like other Nikkors. Aperture and focus rings are ribbed to be easily grasped, though grabbing a focus ring so close to the lens mount takes a little practice (close due to the lens' tiny size). It's a matte silver color. The lens comes with a silver-rimmed coated protection filter, a silver outside but black inside lens hood, silver colored front and rear lens caps (the rear one is really more of a light grey), and a soft case (a pouch with a drawstring). The lens cap is specially designed, with different latches from other Nikon lens caps, so it can either be used on the lens without the hood or it can attach over the lens hood. So there's no excuse not to always use the hood with this lens. The hood is unusual in that it bows inward, like an inverted bowl. Attach the filter, the hood (with a rim only about as deep as a filter) and the lens cap, and you've added 1/2" to the of the lens, almost doubling its depth. The elements are inset deeply enough and the hood adds enough additional protection that a protection filter seems unnecessary. The lens has the contacts to convey information to Nikon electronic cameras and flashes. Using it on an F100, the aperture numbers show up in the viewfinder. Matrix metering works fine (though per the included instruction sheet, it does not send "D" distance information). The SB-28 auto-zooms (if you have it set to auto-zoom) to 35 mm. Using the 45 on a camera with ADR, like the FM2 or F5, is a little surprising the first time you look through the viewfinder: Instead of seeing the usual white numbers against a black background, a rectangle of black numbers against a silver background is reflected in. But put it on an FM2 and hang the combination around your neck and, especially after using an F100 with 35-70 f/2.8, and you'll barely feel anything there. I've had a chance to shoot three rolls in the week I've had the lens, two in the field and one test roll comparing the 45 to other lenses. Photos from the field look exceptionally sharp, and hold shadow detail to an extent few other Nikkor I've tried are capable of doing. Along with that, it holds a beautiful range of tonal gradations. Out of focus areas are smooth, not harsh as with the 50 mm Nikkors and not just neutral as with the 35 f/1.4 and 35-70 f/2.8 Nikkors. As expected with a Tessar-based design, those qualities I found objectionable in photos made with 50 mm Nikkors are not even hinted at in photos made with this lens. Flare is well controlled (as it should be with just four elements). The lens focuses to 1.5 feet (as opposed to the 2.6 feet of the old 45 GN Nikkor, from which this lens is derived), making close shots possible. The close performance was excellent, with sharp in-focus portions and beautifully blurred out-of-focus backgrounds. Indoors, I was able to shoot a test roll comparing it to a 50 mm f/1.8 AF Nikkor, a 35 mm f/1.4 AI-S Nikkor, and a 35-70 mm f/2.8 AF Nikkor. I set an F100 up in front of a bookcase full of record albums and video tapes, spines showing, with an SB-28 providing light (the SB-28 was set to Matrix TTL flash for each lens except the 35 f/1.4, for which it was set to center-weighted TTL flash). I shot each lens twice at f/2.8, f/4, f/5.6 and f/8 on Ecktachrome 100 slide film. The resulting slides were examined with a 10x Schneider loupe. Let's be clear on one point: This is not a scientific test. But it has always proved to be good enough to tell me what I want to know about the relative quality of lenses. Photos with the 45 show minimal pincushion distortion, nothing objectionable. Comparing photos taken with the four lenses at f/2.8, the 45 was clearly the sharpest lens, resolving the greatest amount of detail over the entire frame. It surprised me that the 45 even wide open is sharper than two prime lenses closed down. Second sharpest at f/2.8 was the 35 f/1.4, third was the 50 f/1.8. The 45 and 35 are particularly good in the corners, while the 50 was somewhat worse there. Fourth, as expected with a zoom wide open, was the 35-70 f/2.8. There was a fairly noticeable gap between the sharpness delivered from one lens to the next. At f/4, the difference between lenses narrows, though I'd still rate the sharpness from one lens to the next in the same order. At f/5.6, the level of detail delivered by the 45 remains a step ahead of the other lenses. It was difficult to discern a difference between the sharpness levels of the 35 f/1.4 and the 50 f/1.8 here. The 35-70 f/2.8, while very good, was a step behind the prime lenses. By f/8, there may be an advantage to the 45, but it is slight. The 45, 35 and 50 are nearly equal in terms of level of detail rendered on film. Here, proper exposure will have more impact on the detail captured than which lens is used. The 35-70 is a small step behind the primes lenses, but here, too, the difference is small. The 45 mm is not a lens for everyone. If you need a faster normal lens, you may be better off the 35 mm or one of the 50 mm Nikkors (I'm keeping my 35 f/1.4 for those times I do need a faster lens). If you need a cheaper lens, the 50 f/1.8 is your better choice (note that by the time you add a hood and filter to the 50 f/1.4, accessories the 45 comes with, its price is comparable to the 45). If you need a lens that focuses itself, the 45 is not right for you. But if you want a small and light lens to carry with you, which is optically and mechanically unsurpassed, the 45 f2.8 P Nikkor is a perfect choice. Larry

Date: Tue, 28 Nov 2000 13:18:55 -0600 From: Larry Kopitnik <kopitnil@marketingcomm.com> Subject: [NIKON] 24-85 f/2.8-4D IF Nikkor Review If a lens could be designated the Swiss army knife of lenses -- that is, one that includes a tool for nearly any situation in one convenient package -- the 24-85 f/2.8-4D IF Nikkor would be it. It covers a wide and extremely useful focal length range. It has the fastest aperture range of any wide-ratio zoom. It's reasonably small and light. It includes 1:2 macro, accessible by flipping a switch. The front does not rotate when either zooming or focusing (making use of a polarizing filter easier). The lens is internal focus, so it does not extend when focused. This tries to be the one lens that does it all. And it's a lens I really want to like. My normal "carry-everywhere" kit includes 24 f/2, 35 f/1.4 and 85 f/1.4D Nikkors. I'm not looking to replace those lenses -- they're an outstanding available light kit - -- but to have one lens that covers the focal lengths I shoot most during those times I don't need the extra lens speed, would be a terrific convenience. I bought the 24-85 looking for it to replace my 35-70 f/2.8D, a lens of limited range but excellent optical quality. For the record, the 24-85 is comparable in size, weight, and construction to the 24-120 zoom. It comes with a lens hood. First point to note on the 24-85 is that it exhibits significant barrel distortion at the 24 mm end, changing to pincushion at the 85 mm end. Around 50 mm distortion is minimized, but at no point is the lens distortion-free. I owned the 24-120 lens for about a year (too slow for the type of photography I prefer, so I kept the 35-70 instead). It's been awhile since I had that lens, and I've not done a side-by-side comparison, but distortion in the 24-85 is comparable to what I recall with the 24-120. I did evaluate it side-by-side with the 35-70 f/2.8, a well-corrected but not perfect lens. From 35 mm to 70 mm, distortion with the 24-85 is much more prominent than with the 35-70 f/2.8. This is not a surprise. A wide ratio, wide-to-tele zoom is going to have distortion. If you're going to shoot architecture, the 24-85 is not the lens to use. The 24-85 also flares easily in flare-prone situations. Flare is my principal complaint with the 35-70 f/2.8; with its 15 elements and wide aperture, there's a lot of glass in it for light to bounce off of. The 24-85 also has 15 elements and a wide aperture (and a much larger front element and a wider angle), so the fact that it flares is not unexpected. If anything, with the bigger and more protective lens hood it comes with, the 24-85 seems a bit better protected from flare than the 35-70 f/2.8. (An aside: I've tried to make the large lens hood for the 28-105 fit the 35-70 f/2.8, to give the 35-70 better flare protection. It appears that the 28-105 hood will not vignette on the 35-70, but its bayonet is different, so the hood will not stay in place.) There is noticeable light fall-off at the corners of the 24-85 at 24 mm. This improves when the lens is stopped down and when zooming to a longer focal length. Again, not unexpected, but something to be aware of in use. The 24-120 Nikkor is actually a slower lens than its markings. Popular Photography's tests show the 24-120 to really be f/3.8 - f/6.3, or a third stop slower at the long end than it is labeled. Since one of the advantages of the 24-85 is its faster aperture range, it was important to me to see how accurate that aperture is. I spot metered with my F100 a grey card with the 24-85 set to 24 mm wide open (marked f/2.8) and 85 mm wide open (marked f/4). Then I metered that same spot with my 24 mm f/2 closed down to f/2.8 and my 85 mm f/1.4D closed to f/4. Surprisingly, the the 24-85 mm at 24 mm was faster than the 24 mm prime closed to f/2.8. At the 85 mm end, readings between the two lenses were the same. Of course, the F100 only displays readings within 1/3 stop. But, assuming the prime lenses I was comparing against were accurate when stopped down, it seems the 24-85 is a bit faster than f/2.8 at the wide end and at least within 1/3 stop of f/4 at its tele end. Very impressive. I also wanted to check the actual focal range, which often differs from the lens markings. For example, depending on which source you believe (Pop Photo or Photodo), the 28-105 Nikkor is really 28-99 mm or 29-100 mm. Regardless, it doesn't reach 105 mm. The only test I had available was to compare the field of view to prime lenses of like focal length. Hardly an exact science, but it would give me some idea of focal length accuracy. The 24-85 shows nearly the exact same field of view at 24 mm as my 24 mm f/2 Nikkor. Assuming my 24 mm Nikkor is accurate, the 24-85 really is 24 mm -- or close enough to make little difference -- at its wide end. However, at 85 mm, the 24-85 showed a noticeably wider field than my 85 mm f/1.4D Nikkor. My best guess is that the 24-85 is really only about 80 mm at its long end. Additionally, internal focus zooms actually become a shorter focal length as they are focused closer. I often use my 85 mm lens close to its 3 foot minimum focus distance, so I wanted to see how much focal length the 24-85 loses at about 3 feet. Compared to the 85 f/1.4D, there is a significant difference in the field of view. I would guess the 24-85 at its long end is actually about a 75 mm lens when focused to 3 feet (maybe a little less, but definitely greater than 70 mm). I shoot with a couple of FM2s in addition to my F100, so how a lens handles manually is important to me. Most Nikon zooms hold focus throughout their focal range, and are most accurately focused at the long end, then zoomed back to the desired framing. The 24-120 Nikkor, however, is varifocal; its focus point changes as it is zoomed. This is noted in its instruction sheet. That lens could only be focused after zooming (not a problem on an autofocus camera, but a major inconvenience when used manually). The instructions to the 24-85 note that it only changes focus when zooming in the macro range. Since the macro range can be locked out, that is a big advantage over the 24-120. But the focus throw is extraordinarily short for the long end. It is comparable to the focus throw of a 24 mm prime lens. But when zoomed to 85 mm, an eyelash-width move of the focus ring is a significant change in focus. I keep the original screen, with split image and microprism focus aids, in one of my FM2s, and the grid screen, with no focus aids, in the other. When using the FM2s, I'll generally frame the image and focus on the ground glass portion of the screen. But the 24-85 has such a hair-trigger focus throw that I would not trust that I've achieved accurate focus without using a focus aid. I would use the 24-85 only on my FM2 with the standard screen. With that focus aid, the 24-85 can be manually focused accurately, albeit slowly and with care. This is not a lens which anyone is going to use for manual follow-focus. So far, so good. Not a perfect lens, one with compromises (as expected), but perfectly useable if it delivers sharp images. So does it deliver sharp images? No, it does not. At least, not wide open, when examining slides under a 10x Schneider loupe. Closed down a couple stops images are sharp and crisp. But what's the point of having a fast lens if it must be closed down to deliver acceptable images? Wide open images will likely look fine if all you ever view is a 4" x 6" print. However, when examined with that loupe, image softness wide open was, for me, unacceptable. My test was this: An open vertical display case of record albums (old 33 rpm records), with their spines showing. This provided a mostly flat area with plenty of variation and type to help in evaluating sharpness. With an F100 on a tripod, I made two shots at each marked focal length (24 mm, 28 mm, 35 mm, 50 mm, 70 mm and 85 mm), one wide open and one closed down two stops. I used flash so camera shake would be irrelevant. I shot on Velvia, a fine-grained film. I used the camera's autofocus, focusing on the album spines. Let's note two point here. One, this is not a scientific test. I make no pretensions of it being so. Anyone who cares to call the test flawed I will not argue with. But it gives me enough information to evaluate sharpness for my own uses and purposes. Second, one could reasonably make the argument that the autofocus might be a bit off, given the short focus throw, and manually focusing would be more accurate. But I would counter that a lens which cannot autofocus accurately with my F100 is as useless as a lens which is optically soft. So either way, the test tells me enough to determine whether this is a lens I'll keep. As previously noted, distortion was evident at all focal lengths. On every shot, the shelves of the display case bowed outward or inward noticeably. But more importantly to me, on every shot wide open, at all focal lengths, type on the album spines was soft. It was softest at the wide end and grew increasingly soft towards the corners of the image, becoming illegible in the corners at 24 mm. The image improved as the lens was zoomed out, but at no focal length would I call the image when shot wide open to be sharp and crisp. Closed down two stops was another story. Now all images are very sharp, very crisp. Except at the extreme corners at 24 mm, all type - -- even the smallest on the albums -- is perfectly legible when examined under that 10x Schneider loupe. And even the corners at 24 mm are significantly improved. I've shot two other rolls of slide film with this lens, not on a tripod, but in everyday situations. All confirm what I found in my test shots. No shot made with the lens wide open has any point which, when examined with a 10x loupe, I would judge to be sharp. However, every photo made with the lens closed down two stops is sharp and holds an excellent level of detail. Two images of a brick wall are among the most telling: Wide open, the outline of the bricks is soft. Two stops down, every pore in the bricks is captured. I just played with the macro feature and did not shoot valid enough test shots to reasonably evaluate its optical quality. My conclusion? Closed down two stops, the 24-85 f/2.8-4D IF Nikkor is a good all-purpose lens. But used wide open, it delivers an unacceptably soft image. As I write this, I've just come back from my dealer. I've returned the 24-85 and gotten my money back on it. The 35-70 f/2.8D remains my mid-range zoom of choice. Because optical quality is more important to me than a wide zoom range. Larry

Date: Sun, 11 Feb 2001 15:52:01 +0100 From: "Walter Freiberger" Subject: [NIKON] 24-85 impressions Hi list, and all those interested in the 24-85 lens I just got back some films I shot with the new Nikon 24-85 lens, and I think I let you know some of my impressions. I didn't do any scientific testing, I just compared it in real life situations to the other lenses I have in this range. Film used was Velvia rated at 50 ASA. The other lenses used were the 20/2.8 AF-D, 28/2.8 AF-D, 50/1.8 AIS, 85/1.8 AF and a 24-50 AF-D zoom. The results were projected with a Leica colorplan lens. The first point to mention is the colour rendition of this lens as it is completely different from that of the prime lenses: The pictures look like they were underexposed and/or taken through a warming filter and show a quite pronounced touch of yellow. This goes for all FL's and all apertures, it becomes a bit better when stopping down to f/8 or f/11 but not much. Knowing that Velvia is quite sensitive to underexposure and tolerant to overexposure, I overexposed the next shots by 2/3 stops and got better results. The pictures begin to look a bit overexposed though. An overexposure of 1/2 stop would be just right I guess. Even if overexposure helps, the color rendition of this lens remains a problem, my cheap old 24-50 zoom is definitely *much* brighter, not to mention the prime lenses. The lens contrast wise behaves better with E100VS. Results look not too bad. The difference in brightness to the 24-50 becomes smaller but is still very visible. The lens reaches its maximum sharpness at f-stops between f/8 and f/11 at all FL's except at 85mm were it reaches quite a good sharpness at f5.6 also when compared to the 85mm prime at f5.6. But generally the sharpness is nowhere near to prime lens sharpness. It is about as sharp as the 24-50, eventually a bit sharper at wide apertures, and a bit less sharp at f11. Flare behaviour of this lens is good, as long as you use the lens hood. With the sun in the picture it shows only little ghosting, but a bit more than the 20mm lens. With the sun just outside the picture it shows a bit more ghosting, but less than the 20mm prime (with hood) and the 28mm (without hood). Colour rendition in counterlight shots is horrible by the way, with a pronounced yellow touch (also see above). I didn't test the lens for distorsions. Handling is good with a nice zoom ring feeling. Size and balance are very conveniant on the F90X. The hammered finish is nice. The lens hood can be reverse mounted and looks quite cool. AF is very fast both on the F90X and F80. MF is critical at the long end only, but still not impossible. To sum it up: This is a nice lens with a perfect zoom range and a cool looking lens hood... A positive point is also the sharpness at the long end at relatively wide apertures which may distinguish it from the cheaper 28-105 lens. But: Its basic sharpness is still not breath taking and the colour balance for Velvia film becomes only supportable when you overexpose it by 1/2 or 2/3 stops. I guess these problems don't exist for print film though for which it is certainly a great lens. For Velvia, for which it seemed to be ideal given its fast aperture, it is a bit problematic. I'll try it again with E100VS and also with K64 film which may fit this lens better . Regards Walter

Subject: Tests of Nikon Telephotos/Converters (long) Date: Sun, 17 Nov 1996 08:35:39 GMT From: jplater@netcom.com (Jay Plater) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Newsgroups: rec.photo.equipment.35mm, rec.photo.misc, rec.photo.technique.nature Tests of Nikon Lens/Teleconverter Combinations by Jay Plater People are continually posting questions about various lens and teleconverter combinations. I've had a chance to test many of the Nikkor telephotos with quite a few converters using a standard methodology. Following is a list of the lenses and converters tested. If you don't see your lens on the list, just send it to me and I'll add it! Lenses tested: Nikon 180mm f2.8 ED-IF AF-N Nikon 300mm f4 ED-IF AF Nikon 500mm f4 ED-IF AIS-P Nikon 600mm f5.6 ED-IF AIS (not ÒNÓ) Converters tested: Nikon TC-14B Tamron 1.4X AF Nikon TC-16A Nikon TC-301 Tamron 2X AF (7-element) OVERALL RESULTS* Alone TC14B TC16A Tam1.4 TC301 Tam2X Overall ----- ----- ----- ------ ----- ----- ------- 180mm f2.8 5/4? No v(4)/v v/v(2.5) No 2.5/4 4.5 300mm f4 4/4.5 3.5/4 v/3 v/2.5 3.5/3.5 1.5/2 4.0 500mm f4 4.5/5 4.5/4.5 v/4 v(4)/v 3/3 NT 4.5 600mm f5.6 3/3.5 2 NT NT 1.5/2 NT 3.0 *(subjective 0-5, 5 best, a/b= wide open/one stop down, v(#)=unacceptable due to vignetting (#) center of image, No=combo won't fit together, NT=Not tested) Descriptive Results: Nikon 180mm f2.8 ED-IF AF-N Is it just me? In repeated tests with various converters this lens seemed to do better wide open than one stop down. This is a fantastic lens with great optics and good ergonomics. Unfortunately, I haven't yet found a converter that works with it. I imagine a TC14A. It's a shame this lens vignettes with a TC16A, because it was great in the center. The TC16A was supposed to be OK for use with the AIS 180 f2.8. Could use a focus limiter. Nikon 300mm f4 ED-IF AF A good lens, but be careful what you mate it with. Yes it's good by itself, but my tests of performance with the TC-14B were disappointing. I would try to avoid shooting this combo wide open. I even tested this one twice to double check that it wasn't operator error. On the other hand, I was surprised at how well it does with the TC 301. Fine ergonomics on this lens; a removable tripod mount would be nice. Nikon 500mm f4 ED-IF AIS-P The optics in this lens are simply incredible, if only ergonomics were as good. This and the 180 are the sharpest lenses I've tested, and I notice almost no degradation of quality with the TC-14B. This explains why the lens is sometimes called the "Nikon 500-700 zoom." Performance is acceptable but unimpressive with the TC-301. The combo seemed sharp enough, but color suffers. I also tested "stacked" converters: 500 + TC14B + Tamron 1.4. Result is v/2.5. With the Tamron 1.4 alone, centers were very sharp but it was still vignetting significantly a stop down. Incredibly hard to avoid camera shake at this focal length. In all of my 2X tests, results were better wide open than stopped down-this must be due to shake. Considering shake is such a problem under even ideal conditions, I would shoot the 2X combo wide open. Ergonomics and build quality are disappointing in my lens. Focus feel is poor. Nikon 600mm f5.6 ED-IF AIS This is a fine and very sharp lens; I have made numerous wonderful photos with it. But it gets blown away by the other lenses tested. This just shows how great Nikon's lenses are--the worst of the bunch is still a fine lens. Performance is noticeably worse with teleconverters. Perhaps shake was playing a roll here. I think this lens is too long and (comparatively) slow to have much luck with converters. This was the older version with lens hood built in, but I understand that the optical formula hasn't changed. The ergonomics were wonderful. Focus feel was perfect, silky smooth, much better than the 500. But considering the optical results, I'd spend the $5,000 for a long lens on something else. Converters: TC-14B Well it appears that everything that's said about this converter is true. With a well-matched lens, it's hard to see any image degradation. This converter is great with the 500f4, but just OK with the 300f4 wide open. I wonder how a TC-14C would do . . . Due to a protruding front element, this converter won't anything shorter than 300mm (but it *does* fit the 80-200 f2.8 EDIF). TC-16A It's a shame that this converter won't work with telephotos due to vignetting, because the optics are just a bit inferior to the TC14B. And it would be nice to have an intermediate extender between 1.4X and 2X. The centers of the vignetted slides were usually very sharp, making me wish there was a "TC16B" TC301 While this converter can do the job, my results show that it's best to avoid using it if possible. I think that color will really suffer with this converter along with sharpness,but I can't be definitive because I was using B&W targets. Results were pretty good with the 300f4, and acceptable with the 500. Like the TC-14B, it won't fit lenses shorter than 300mm. Tamron 1.4X and 2X Cheap converters, lousy images. That's about the conclusion. The 1.4X, which is actually stronger than 1.4 (maybe about 1.5), vignettes on everything I tested. And the centers weren't sharp either. The 2X doesn't vignette as much, but still delivers poor images. As these are usually regarded as the best of the off-brand converters, I would conclude that one should buy a converter from the company who made the camera. Notes on testing methodology: Tests were conducted using a Nikon F4, Arca ballhead, and Bogen tripod (unextended) on firm ground. Tests were performed under various shooting parameters, but always included a series with the mirror locked up. Additional stabilization included a Kirk enterprises bean bag placed on the lens, an electronic cable release, and in the case of the 500f4 and TC301 combo, a Kirk Enterprises support designed for this purpose. While I'm sure a more stable setup could be designed in a lab, stability was at least as good as one could possibly expect to have in the field under ideal circumstances. Tests were performed outdoors. Exposures were made on Fujichrome Velvia film. Shutter speeds seldom fell below 1/125. Each frame was shot, re-focused, and shot again. The best of the multiple images were used for comparison. (Very seldom was there any variation, which lends credence to the results.) The test target was a glossy Le Camera ad from Shutterbug Magazine with very small, sharp text that had been linotronically printed. The page was mounted on board to ensure flatness. The lenses were somewhere in the middle or close end of their focus ranges, but none were at or near their minimum focusing distance (the 180 was fairly close-in). The resulting slides have about a hundred rows and five columns of *very* small black and white text. Slides were labeled and viewed on a light table with dual apochromatic 15X loupes, which offered ample magnification to compare sharpness. Since test targets were black and white, color rendition could not be tested. I did notice varying degrees of color shift, however. Please feel free to send comments. Jay Plater

These were shot using David Jacobson's PostScript 1951 USAF type test targets printed on 600 dpi LaserWriter. The camera was an F4s placed at 40 X the focal length of the lens being tested on a sturdy Tiltall tripod (except for the 24mm/2, which was shot at 25 X focal length). Mirror lockup was used after focusing. The film was Kodachrome 25, which was evaluated on Zeiss or Olympus microscopes at 100X total magnification. I use these results as a guide for picking the best F-stop when highest sharpness is the primary consideration. These are the results for my lenses and may not match others, but may give you a rough guide as to the quality of each lens. While doing the tests, I found that I could generally focus slightly better by eye than by using autofocus or the electronic rangefinder. I also found that filters (even Nikon L37c UV filters) cut resolution about 5-10 lpmm, especially a problem with the with the 20mm/2.8. The values are given for each f-stop in lines per mm on the film for center and corners (which are generally averages of the four corners). Also given is the diffraction limit estimate for each f-stop in lpmm, which represents the theoretical maximum limit of resolution for that aperture. A value of at least 40 lpmm center and corner will allow a 10x enlargement that most people will consider sharp when viewed at about a foot (30.5 cm) viewing distance. AF-Nikkor 20mm/2.8 AIs-Nikkor 24mm/2.0 f-stop: Center: Corner: D-limit: f-stop: Center: Corner: D-limit: 2.8 50 34 530 2 47 38 743 4 57 38 371 2.8 53 40 530 5.6 71 43 265 4 60 49 371 8 90 46 186 5.6 67 58 265 11 80 49 135 8 67 60 186 16 71 50 93 11 67 58 135 22 50 40 68 16 67 53 93 22 60 45 68 AF-Nikkor 35mm/2.0 AF-Micro Nikkor 60mm/2.8 f-stop: Center: Corner: D-limit: f-stop: Center: Corner: D-limit: 2 50 38 743 2.8 50 45 530 2.8 57 39 530 4 63 50 371 4 71 43 371 5.6 63 63 265 5.6 90 45 265 8 83 83 186 8 90 47 186 11 71 71 135 11 90 62 135 16 67 63 93 16 80 67 93 22 57 50 68 22 63 58 68 32 42 42 46 AF-Nikkor 85mm/1.8 Nikkor-P 105mm/2.5 (type 2) f-stop: Center: Corner: D-limit: f-stop: Center: Corner: D-limit: 1.8 50 38 825 2.5 50 30 594 2 50 40 743 2.8 62 32 530 2.8 71 53 530 4 62 38 371 4 80 57 371 5.6 83 43 265 5.6 80 58 265 8 83 54 186 8 71 71 186 11 83 56 135 11 71 67 135 16 63 58 93 16 71 59 93 22 50 45 68 AF-Micro Nikkor 105mm/2.8 AF-Nikkor 180mm/2.8 ED-IF "n" f-stop: Center: Corner: D-limit: f-stop: Center: Corner: D-limit: 2.8 36 36 530 2.8 57 44 530 4 45 40 371 4 63 49 371 5.6 57 45 265 5.6 63 54 265 8 63 50 186 8 71 62 186 11 83 63 135 11 63 62 135 16 63 63 93 16 63 57 93 22 57 50 68 22 57 49 68 32 41 41 46 Sincerely, Steve Keirstead Biology Teaching Labs, Harvard University Man having created the concept of God the Creator, Found himself unsatisfied. For despite the proven pragmatic value of this image, through which the fine arts of music and literature, of architecture, painting and sculpture, together with the less fine arts of murder, thievery and general human exploitation, had been carried to new heights, there was still something unfulfilled: the impulse of curiosity in man was still hungry. - Paul Strand, from "Photography and the New God"

From jouve@cmapx.polytechnique.fr Tue Jan 17 11:03:41 MST 1995 From: Francois Jouve Newsgroups: rec.photo.advanced,rec.photo.misc Subject: Personnal lens tests Date: 17 Jan 1995 17:04:34 GMT Organization: Acces regional Ile-de-France (Univ. Paris VI/VII) - France Lines: 58 I made some personnal tests on my lenses. It may be interresting for adicts of this group: Technicaly, I used a lens testing chart sold by "chasseur d'images", a tripod and a flash (to avoid body vibrations), and Pan F Plus film developped in Perceptol (--> 40 Iso). I didn't have any rolls of Tech Pan available at this moment but I think PanF is good enough to resolve 100 lines/mm (this is the limit of the chart). I didn't test objectively the contrast of the results, which is known to be very important in actual photography. I didn't test either color rendering, flare and distortion. Resolution Resolution f-stop at center at periphery (lines/mm) (lines/mm) Nikon AI 2.8/24 2.8 50 33 Nikon AI 2.8/24 8 90 48 Nikon AI 1.8/50 2.8 60 40 Nikon AI 1.8/50 8 90 62 Nikon AI 2.5/105 2.8 50 31 (poor contrast) Nikon AI 2.5/105 8 65 48 Nikon AI 4/200 4 50 < 30 Nikon AI 4/200 8 55 35 Nikon AF-ED 2.8/180 2.8 60 35 Nikon AF-ED 2.8/180 8 65 40 Summicron M 2/35 2.8 70 43 (poor contrast) Summicron M 2/35 8 80 50 Summicron M 2/90 2.8 70 38 Summicron M 2/90 8 70 38 (=2.8 --> bizarre, maybe a manipulation error) (Minox 35 GT) Minotar 2.8/35 2.8 unreadable (bad focus) Minotar 2.8/35 8 40 50 As you can see, the lens's resolution isn't enough to juge it completely. For example, my old 4/200 always gave me acceptable results, but no more than acceptable. Slides made with my new 2.8/180 are incredibly sharp compared to the old one, but the figures of the two lens are quite close... Hope it will be usefull F.J. Results have been watched through my old toy microscope (x100), which is not very sharp, but sharp enough to see the emulsion grain.

From alj@u.washington.edu Tue Mar 21 09:46:03 MST 1995 From: alj@u.washington.edu (Allen Johnson) Newsgroups: rec.photo.advanced Subject: aerial image tests - Sigma 70-300 Date: 21 Mar 1995 05:06:03 GMT Organization: University of Washington I tested some more lenses using microscopic examination of aerial images. This time the orientation was horizontal - no mirrors. We tested a number of Nikon and Nikon compatible long glass, and the results were in general that the center resolution was quite good for prime quality lenses - we could see ~320lpm at best aperature (generally -1 stop from max). The test conditions were the Charles Sliecher target at distance = .18x(focal length in mm)feet i.e. .02X magnification. The target was illuminated by a 300 watt lamp ~1ft away. In more detail, I tested the Sigma 70-300mm f4.5-5.6 APO AF macro zoom for Nikon, and found: at 70mm ~300lpm at center, ~120lpm at corner at 200mm ~200lpm at center, ~120 at corner at 300mm 130lpm at center, ~130 at corner The Nikon 50mm gave 320lpm at center, 240 at corner The Sigma 28-70 2.8 gave 280 at center, 240 at corner at 70mm The Nikon 28-70 3.5-4.5 gave at 70 280 center, 120 corner The Nikon 200 macro gave 240 center, 240 corner The Nikon 300 2.8 gave ~220 center, 160 corner. Note that there are still some problems with the test method - there is imperfect matching between the pupil of the lens with the pupil of the microscope, particularly at the corner and at lower focal length. These numbers are to be used as internal references only - comparing them to on film resolutions suggest that the lens per-se for high quality glass is not the limiting factor. I will mention that at 320lpm my microscope is starting to run out of resolution. The numbers obtained suggest that prime lenses are better in the corners, that top dollar zooms are almost as good, and that the Sigma goes softer at 300mm and in the corners. I rather suspect that the Sigma 70-300 will show softer photographs than the equivalent prime lenses. How much? Don't know.... yet. Hope these are of use. Allen JOHNSON@CHEME.WASHINGTON.EDU

From alj@u.washington.edu Wed Jan 11 16:10:15 MST 1995 Article: 8067 of rec.photo.advanced From: alj@u.washington.edu (Allen Johnson) Newsgroups: rec.photo.advanced Subject: lens tests - mostly Nikon Date: 9 Jan 1995 03:32:19 GMT Organization: University of Washington Lines: 103 Another in my attempts to figure out how to think about /measure optical equipment.... the caveat is up front - data is only good to 10% if that and is compromised by getting the measurement done in my spare time.... I spent some of this weekend testing more lenses - this time using an inverted Unitron trinocular microscope with 10X .25 objective and 10WF eyepieces to examine the image formed from photographic lenses observing a test chart (described in my earlier posts as the CS chart) lit with a ~300watt quartz halogen bulb with 2" reflector ~4 ft away. It was possible to work with f# up to 11. These numbers will probably surprise and/or confuse you - they are not to be compared with almost any of the resolution numbers that you find in magazines. They are close to the projection method results I posted sometime back on the Nikon micro lenses. The lesson I learned from doing this was that in most cases with high quality lenses your results are technique limited. Also, that "lpm"is not a generally useful measure of optical performance. The chart was used at two different distances - 6'9" and 10'6" - I will note the distance in the results. We examined visually the black target in the center and the edge - 1st caveat is that the center was well lit, and the edge was not. Thus, the edge resolutions are lower bounds. Further, we did not religiously reset the flange - objective distance to be correct - this means that for the zooms and IF lenses we were not working exactly at design point. We correct the observed spatial frequency to image plane lines/mm. With the above caveats, at 6.75' lens f# center corner comments 50mm 1.8 1.8 240 210 Nikon correction factor 1.5 4 240+ 8 195 195 11 135 28-70 2.8 2.8 160 @70 sigma c.factor 1 5.6 160 160 @70 5.6 240 210 @50mm c. factor 1.5 5.6 320 241 @28mm c. factor 2.7 28-70 3.5-4.5 5.6 160 140 @70 Nikon 5.6 240 210 @50 5.6 324 243 @28 28mm 2.8 5.6 294 214 Nikon 24mm 5.6 281 219 Nikon c. factor 3.125 20mm 2.8 5.6 262 225 Nikon c. factor 3.75 at 10.5' 105mm f2.8d 2.8 178 178 Nikon c. factor 1.11 5.6 178 156 105mm f2.8AIS 2.8 178 133 Nikon 5.6 178 156 70-210 2.8 2.8 267 200 Sigma @70mm c factor 1.67 5.6 267 217 @70mm 2.8 178 178 @50mm 5.6 178 178 @50mm 2.8 80 80 @210mm (ran out of target res.) 5.6 80 80 @210mm (ran out of target res.) Some discussion: Note that for most of the longer focal lengths (longer focal lengths here means >28mm), the maximum resolution of the chart, 160 lpm (for the technically inclined, this is a standard 1"=1mm at distance chart...) was too low to distinguish performance. However, the performance found was generally much better than ever seen on film - why? 1. visual threshold (lpm) is a rather useless number in intermediary stages - it only makes sense on slides or prints, because you will be looking at them with you eyes. Further, film resolution and lens resolution are only the start of picture resolution - you must interpose the film at the image plane, over all of the image plane - thus perhaps f8 works best because you then have more depth of field at the film plane. Further, vacuum backs, disk film and other methods of making film a particular geometry might not be bad ideas... lastly, film has non-zero thickness. More discussion of film geometry can be had in "Camera Technology: The Dark Side of the Lens" by Goldberg. Some observations - most of the lenses showed color fringing at lowest f# - gone by 5.6. It did not change lpm generally - one could still see lines in the fringes. The edge measurements were refocused - we did not try to make the chart exactly horizontal so it would have been useless not to refocus. Next go around, we will probably set up a large distance to the chart, as well as try to estimate a "useable contrast" limit rather than the visual limit. If I have time, money, and a job I would really like to put a CCD on my 'scope and get quantitative numbers - I would like to fourier transform a step response, and get MTF that way... neat. For further thoughts on the correct way of measuring optical elements I would suggest finding the r.p.a MTF discussion in either the lens FAQ/tutorial or in past posts. questions to JOHNSON@CHEME.WASHINGTON.EDU Allen (interesting, no?)

From: alj@u.washington.edu (Allen Johnson) Subject: Nikon lens test results - comments on mirror lockup I did some lens testing recently, and thought I'd post the results. These tests were to examine Nikon micro lenses both at normal distances (.04X) and at micro distances (.5-6X). For the micro work I used a resolution target sold by Four Designs which is a film with Air Force like targets densely packed on a 4cmx4cm grid, with features down to 320 lpm. The normal test was done with an improved (field is 18% gray, 93% viewfinder guide lines, focusing target, 4 color wide line resolution targets over the field, high quality print job [important! some of the commercial targets aren't good past group 3...], etc.) target designed and printed by a friend and associate. The first target is the FD target, the second is the CS target. The results are below - some discussion.... the resolution step is 10 lpm - and so the numbers should be considered ambiguous to that degree. The FA was operated with pseudo-mirror lockup - still shake was observed. These results are best compared with itself - I didn't do it the same way that PP might, though I will mention that my numbers and MP's numbers were close. I used a light tripod for the 0.04X work, and a fairly heavy micro tripod on a large light table for the micro work. I might have been able to get better numbers using bean bags, heavy tripods, stuff like that but I think this is more representative as to how I (and many other people) make photos. So, the numbers are relative in any case. (rather than reformating this... the CS target has 10 lpm resolution step, the FD target is much finer - correction in proofreading.....) CS target, 0.04X, technidol LC, Techpan @ ASA 30, Nikon FA, 1/30 & 1/4 sec f5.6 & f16 Lens Res. center edge notes Lens Res. center edge notes 105f2.8m 70(50) 60(50) 5.6 ... 60(60) 60(40) 16 ... 60(50) 50(40) 5.6 w/hood 60f2.8m 80(70) 60(60) 5.6 ... 70(50) 60(50) 16 55f2.8m 80(60) 70(60) 5.6 ... 70(60) 70(60) 16 28-70D 60(50) 60(50) 5.6 55mm ... 60(60) 50(40) 16 55mm ... 60(50) 60(60) 5.6 70mm ... 60(60) 50(50) 16 70mm 50f1.8 90(70) 90(60) 5.6 ... 70(60) 60(60) 16 All lenses latest versions AF except 55mm f2.8 micro, all Nikon brand. At .1X 55mm and 60mm are obviously sharper than the 105mm f2.8 micro. The 50mm f1.8 AF was remarkably good. The red targets were the lowest contrast, the blue and green targets were of intermediate contrast, and the black/white targets (measured) were the highest contrast. Generally, one or two steps of frequency down gave usable contrast. Hood did not help. The resolution in parenthesis were the horizontal bars, which show the effect of residual shutter shake, whereas the vertical bars did not. FD targets, .5X, as above but various shutter speeds 105f2.8m 45(32) 28(22) 5.6 ... 50(50) 50(45) 16 60f2.8m 72(72) 56(40) 5.6 ... 56(56) 56(56) 16 55f2.8m 80(70) 64(56) 5.6 ... 50(50) 56(50) 16 At .5X, the 105<60<55. The 55 is only slightly better than the 60. Camera shake is less of a problem than with the above tests. FD targets, 2X (i.e. at 1X with tokina 2x teleconverter) 105f2.8m ... 22(20) 16 60f2.8m 35(28) 5.6 ... 25(25) 16 I also experimented with reversed lenses (the 50mm) and extension - it seems as if the basic resolution of the lens predicts the behavior of the combination - about the same film resolution, magnified by the teleconverter. I will note if one reverses a 50mm lens as a supplemental lens, one must use it at full aperture, or risk severe vignetting. My conclusion? IF you are film resolution limited teleconverters make sense - else enlarge in the darkroom. at 1X the 105 gives about 40 lpm. Hope this all is of use. If there are questions I can be reached at the address below... Allen JOHNSON@cheme.washington.edu (I plan to test these using AF - is the 50mm so good because the focus was brighter/better? Is the AF 105 worse than the MF 105? Is the shake with a N90 of equal magnitude? (possible....). My coworker tested some Leicas, a T2, a T4, some other P&S cameras to get an idea of which to get - I'm trying to get them to put the results on the net.... )

From: jacobson@cello.hpl.hp.com (David Jacobson) Subject: Re: Nikon lens test results - comments on mirror lockup In article rma@clockwise.mh.att.com (Robert M Atkins) writes: >If you are shooting distant targets (esp.outdoors) atmospheric effects >*could* degrade resolution at lower shutter speeds. If you look at >the aerial image under these conditions you can see it "shimmer" >(astronomers are well aware of this effect). You can see it over a >10 yard path on a bad day. We were using the large (7in?) RIT alphanumeric test targets. The longest distance was 17.4 feet. (All the magnifications were 1/24.3 and the longest lens was 210mm. Well, we actually tried a 500mm reflex Nikkor, but at higher magnification.) This was indoors inside Bill Tyler's house with a uniform temperature. It was this spring and the outdoor temperature was comfortable, so no heating or air conditioning was operating. Although you didn't ask, here is some more info. We used a sturdy tripod and an F2 with MLU. The film was Tech Pan develped in HC-110 dilution D, I believe, for 8 or 10 minutes. I don't have my notes handy and without them I can't remember all details, including the exposure we used. >My resolution numbers too have ranged from around 30 lp/mm (edge of >a poor lens or zoom +TC) to 95+ lp/mm (very good prime lens at >optimum aperture). > >Bob Atkins Well, Bill and I decided to just give out the numbers we got. These are still preliminary. I've only checked one corner, not all 4 etc. (It is very tedious reading lens resolution charts through a microscope.) We also have reason to believe we didn't get the magnifications perfect. Please don't take these very seriously and please don't start any wars with them. Note that there are several cases where the edge does better than the center. This makes me suspicious. Note also that deciding what patter is legible is *very* subjective. The RIT patterns give 3 character sequences, but there are only a few of them, and it begins to get half easy to guess. "Let's see this one looks like A9S... but they don't use that pattern. Must be A5S." I geneally read them through the microscope while Bill compared with the "answer book". The ratio from one pattern to the next is the 6th root of 2. We used 5 RIT test charts: one in the center and one in each corner. But each test chart has 4 patterns. So one is the extreme corner and one a fair amount inside. We have data on all the inside ones, but not all extreme ones. Lens f-num center corner extreme ---- 70-210 f/4-5.6 AF@200 5.6 24 27 17 70-210 f/4-5.6 AF@200 11 48 27 24 70-210 f/4-5.6 AF@135 5.6 43 34 70-210 f/4-5.6 AF@200 11 55 43 70-210 f/4-5.6 AF@85 4 49 39 70-210 f/4-5.6 AF@85 5.6 55 49 70-210 f/4-5.6 AF@85 11 55 61 200 4 21 24 <= no MLU 200 4 27 34 200 5.6 39 24 200 11 49 43 500 reflex 8 24 135 2.8 31 43 135 5.6 39 49 135 11 55 55 85 2 38 48 85 4.5 55 61 85 11 55 77 60 MicroNikkor 2.8 55 61 39 60 MicroNikkor 5.6 55 77 69 60 MicroNikkor 11 61 61 61 60 MicroNikkor 16 55 55 61 55 MicroNikkor 3.5 61 69 55 MicroNikkor 5.6 61 86 55 MicroNikkor 11 55 61 24 2.8 61 55 24 5.6 61 55 55 24 11 61 43 61 24 22 55 55 55 Sigma 24 2.8 55 34 22 Sigma 24 5.6 77 54 13 < recording error? Sigma 24 11 61 54 39 Sigma 24 22 43 43 39 Bill conjectures that his F2 has a fairly large focus offset, ie there is considerable difference between the length of the optical path to the film and the optical path to the focusing screen. As someone else suggested, we need to check this out before believing any of this. -- David Jacobson

Date: Fri, 21 Jan 2000 18:20:32 -0500 From: "Christophe Heyman" Subject: [NIKON] AF-S 17-35 f/2.8 IF-ED vs. primes - LONG! First of all, this is a long message... don't say I didn't warn you...;-) As promised, here's a 'condensed' report of the CDI test of this lens + comparison with some corresponding primes (also based on CDI reports). I do understand that not all of you care for all these numbers and 'scientific' tests, and favor a personal practical test instead. In that case, simply disregard this whole message - no hard feelings ;-) CDI always shows graphs, but these obviously don't translate very well in a plain text message, so I've converted the bar graphs to numbers. Scale goes from 1 to 20 (it basically goes to 24, but I don't recall having seen any lens come close to that except for some Leica lenses): 4 = weak 8 = average 12 = good 16 = very good 20 = extraordinary CDI always shows test results as combination of sharpness and contrast, both for center and corners, at full opening, 1stop down, 2stops down and 3 stops down. Also, vignetting is shown, expressed in f-stops difference between center and corners, for all f-stops where it occurs (normally wide open + 1 f-stop down). 0.7 to 0.5 f-stops is considered visible, 0.4 to 0.3 f-stops slightly visible. Distortion is also shown. Everything over 0.7% is very visible, down to 0.5% is visible, down to 0.2% is slightly visible, and down to 0 not visible. Having said all that, here are the results: 17mm: distortion: 0.9% barrel 2.8 center = 9 2.8 corners = 5 2.8 vignetting = 0.7 f 4 center = 12 4 corners = 9 4 vignetting = 0.4 f 5.6 center = 16 5.6 corners = 14 8 center = 19 8 corners = 16 20mm: distortion: 0.5% barrel 2.8 center = 9 2.8 corners = 7 2.8 vignetting = 0.5 f 4 center = 11 4 corners = 9 4 vignetting = 0.3 f 5.6 center = 15 5.6 corners = 13 8 center = 18 8 corners = 15 24mm distortion: 0 2.8 vignetting = 0.5 f 2.8 center = 8 2.8 corners = 7 4 vignetting = 0.3 f 4 center = 14 4 corners = 11 5.6 center = 15 5.6 corners = 13 8 center = 18 8 corners = 15 28mm distortion: 0 2.8 vignetting = 0.4 f 2.8 center = 8 2.8 corners = 6 4 center = 10 4 corners = 8 5.6 center = 13 5.6 corners = 11 8 center = 16 8 corners = 14 35mm distortion: 0.4 % pincushion 2.8 vignetting = 0.5 f 2.8 center = 8 2.8 corners = 6 4 vignetting = 0.3 f 4 center = 11 4 corners = 9 5.6 center = 13 5.6 corners = 11 8 center = 16 8 corners = 14 CDI Comments: Performance increases considerably by closing down - 2 f-stops needed for very good performance. Distortion very well controlled, even at 17mm where it still is not excessive. Very well controlled at all other focal lengths. Vignetting remains within acceptable limits. Color rendition is very neutral. My feelings based on the results (mind you, I have not tested this lens myself - I wish I could...): Results in the 17 to 24mm range are very good, almost comparable to the primes. From 24 to 35mm, performance drops considerably compared to the primes. I would consider this more a 17-24mm than a 17-35mm zoom, given the performance. You really must stop down 2 f-stops to get very good performance (as with e.g. the 28-70 AFS and to a somewhat lesser degree with the 80-200 AFS). Having said all that, I'm still drooling... On paper, this lens seems to have a completely different optical behaviour than the 20-35 AFD (also from a CDI test): the 20-35 gets weaker towards 20mm, whereas the 17-35 gets weaker towards the other end. At 20mm, the AFS is better, at 24mm both are very comparable. At 28 and 35mm, the AFD is better. The AFD is overall better in the corners. Also, the 28mm setting compared to the 28mm setting of the 28-70 AFS: 28-70 AFS better overall, although it vignettes worse - 0.7 f wide open - and distortion is worse as well - 0.9% barrel). Now for the PRIMES: No 17mm... AF 20mm f/2.8 D: distortion: 0.5% barrel 2.8 center = 11 2.8 corners = 7 2.8 vignetting = 0.7 f 4 center = 13 4 corners = 11 4 vignetting = 0.4 f 5.6 center = 15 5.6 corners = 13 8 center = 18 8 corners = 13 Color rendition: neutral My comment: Compared to the AFS zoom, this lens better than zoom between 2.8 and 4, and about equally good at 5.6 and beyond. Vignetting is less with the zoom (!). One more personal comment here: I compared this specific test with a previous CDI test of the non-D version of this lens: the non-D was better at all f-stops except f/8. Vignetting and distortion were slightly better. Amazing, since the only difference is supposed to be the addition of the D. Sample difference or degrading quality?? (I noticed the same thing when comparing several CDI tests over time of the 180 f/2.8). AF 24mm f/2.8D: distortion: 0.4% barrel 2.8 center = 16 2.8 corners = 8 2.8 vignetting = 0.6 f 4 center = 16 4 corners = 12 4 vignetting = 0.3 f 5.6 center = 20 5.6 corners = 16 8 center = 20 8 corners = 16 Color rendition: slightly warm My comment: Compared to the AFS zoom, this lens is better at all f-stops. Vignetting is about equal. Strangely enough, colors are slightly warm. AF 35mm f/2 D (NOT f/2.8): distortion: 0.4% barrel 2 center = 12 2 corners = 6 2 vignetting = 0.6 f 2.8 center = 16 2.8 corners = 9 2.8 vignetting = 0.3 f 4 center = 19 4 corners = 11 5.6 center = 18 5.6 corners = 14 Color rendition: neutral My comment: compared to the AFS zoom, this lens is considerably better at all f-stops. Mind you, this is an f/2 lens, so not completely comparable. That's it! Lots of numbers, food for thought etc. Just to make sure y'all understand: I do not own or use any of the above lenses, I merely report what CDI found during their tests + I added a few personal comments (again, based on these test results only, not on personal findings). Happy reading, Christophe.

Date: Sat, 27 Jan 2001 22:37:37 -0500 From: "Christophe Heyman" Subject: [NIKON] Nikon 85mm f/1.4 AF-D vs. Nikkor 85mm f/1.4 AIS I posted these test results yesterday on the rec.photo.equipment.35mm newsgroup (someone asked whether the AF was as good as the AIS), but I thought you'd be interested in this as well, so here they are. According to the Chasseur d'Images tests, they are both a very good lenses. Below are the numbers (AIS first, then AF-D) . Contrast/Sharpness: (for reference: 0 = minimum, 24= maximum 4 = "weak" 8= "Average" 12 = "Good" 16= "Very good" 20 = "Excellent") (in the numbers below, the first one is the score for the center of the AIS, the second number is the score for the corners of the AIS, the 3rd number is for the center of the AF, the 4th number is for the corners of the AF) f1.4: 8,4 11,9 f2: 12,8 (no score for the AF) f2.8: 16,16 13,12 f4: 16,16 15,13 f5.6: (no score for the AIS) 17,15 light fall-off: 0.6 EV at f1.4 (both lenses) 0.3 EV at f2 (both lenses) none at f2.8 (both lenses) Distortion: very slight pincushion distortion (almost unnoticeable) Color rendition: Slightly warm for the AIS, neutral for the AF One side note: the older tests of Chasseur d'Images apparently did not use ratings other than 4,8,12,16,20, whereas in the newer tests, they also use the ratings in-between, so it might be hard to compare the numbers. Also, for the AF-D, they made the remark that although the numbers at full opening might not look too good at first sight, the lens is really exceptional, in the center as well as in the corners. Evidently, this comparison is only an indication. I've never used either one, so I can't comment on practical results. Hope this helps. Christophe.