17-35/2.8 AFS Nikkor vs. Wide Angle Primes

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI, 20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison is done at f/8 in the corner of a D1x frame. At this aperture and image location the major differences between these lenses is the amount of color fringing present. In every case the zoom lens has less color fringing. The zoom also has less distortion, and I'll have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Interesting test Brian.

Click to expand...

What about the mid section and other f-settings?

http://www.julienlanda.nl
--Julien

Julien:

On-axis at f/8, all the lenses performed essentially identically on a D1x. In the intermediate image region between the center and corner of a D1x frame there were really no surprises, the color fringing is very nearly a linear function of radial distance in all cases, so halfway to the corner it will be half as bad as it is in the corner.

At wider apertures things get a bit more complex. At f/2.8 the zoom performs extremely well at 35mm, and in fact is superior to my 35/1.4 prime at that aperture. However, at shorter focal lengths the zoom has reduced contrast wide open, although resolution is still very high. The contrast picks up rapidly as you stop down.

I was most interested in performance at f/8 because that is where I tend to shoot wide angle lenses most often.

Brian

Julien Landa said:

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Interesting test Brian.

Click to expand...

What about the mid section and other f-settings?

http://www.julienlanda.nl

--
Julien

Click to expand...

The hot ticket for a large-aperture wide angle is without a doubt the 28mm f/2.8 AIS prime, and the 24mm f/2.8D is slightly better than the 17-35 AFS, but otherwise the zoom really is the best thing out there. It is so close to the quality of the 24/2.8D that it's tough discerning the differences and as long as you shoot f/5.6 and smaller there is little compromise in using that zoom, and a lot of compositional flexibility. As you mentioned, for work at larger apertures it is best to stay near the long end of the zoom. The sweet spot, focal-length wise, is 22-35mm and the best results are achieved between f/5.6 and f/11. Even though I don't shoot WA anywhere near as often as I shoot telephoto, I'd not want to give up the 17-35 AFS. It's a great tool.

Ron

brian said:

At wider apertures things get a bit more complex. At f/2.8 the
zoom performs extremely well at 35mm, and in fact is superior to my
35/1.4 prime at that aperture. However, at shorter focal lengths
the zoom has reduced contrast wide open, although resolution is
still very high. The contrast picks up rapidly as you stop down.

I was most interested in performance at f/8 because that is where I
tend to shoot wide angle lenses most often.

Brian

Julien Landa said:

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Interesting test Brian.

Click to expand...

What about the mid section and other f-settings?

http://www.julienlanda.nl

--
Julien

Click to expand...

Click to expand...

--Ron Reznick http://digital-images.nethttp://trapagon.com

brian said:

Julien:
On-axis at f/8, all the lenses performed essentially identically on
a D1x. In the intermediate image region between the center and
corner of a D1x frame there were really no surprises, the color
fringing is very nearly a linear function of radial distance in all
cases, so halfway to the corner it will be half as bad as it is in
the corner.

At wider apertures things get a bit more complex. At f/2.8 the
zoom performs extremely well at 35mm, and in fact is superior to my
35/1.4 prime at that aperture. However, at shorter focal lengths
the zoom has reduced contrast wide open, although resolution is
still very high. The contrast picks up rapidly as you stop down.

I was most interested in performance at f/8 because that is where I
tend to shoot wide angle lenses most often.

Brian

Thank you,Brian

Click to expand...

--Julien

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

Click to expand...

Very interesting test. But in one way not surprising. The 17-35/2.8 AFS Nikkor is designed specially to use with the digital SLR's. I guess the lens has a more retrofocus design than the wide angle primes of the past. So the light is coming more in a perpendicular way to the CCD than the wide angle lenses. You have to be in your mind that the color fringing is not a representation of the lens quality only, but for digital has to do with the character that light hits the CCD in an angle. An interesting message in the past about this subject you can see here:
http://www.dpreview.com/forums/read.asp?forum=1021&message=2165556
--Leon Obers

Hi Leon:

Actually, the degree of telecentricity, or perpendicularity of the off-axis ray bundles, of the lens has nothing to do with the appearance of color fringing. Consider this: the 55mm/2.8 Micro Nikkor is actually less telecentric than any of the wide angle primes in my test except for the 20mm/2.8, which has a slightly shorter exit pupil location. Nevertheless, the 55mm/2.8 displays absolutely no color fringing on a D1x, which is exactly what I would expect from having analyzed its optical performance by raytracing.

Therefore, any color fringing which appears in a D1x frame is due to the lens and not the sensor.

I remember reading the post you linked to, and although interesting it is mostly incorrect.

I do recall once seeing an image shot with a wide-angle Nikkor lens on a full-frame CCD attached to the Horseman Digiflex camera. There appeared to be no problems with this image, although I haven't been able to relocate it.

Many sensors do have a maximum acceptance angle, but in my experience what happens when you exceed this limit is you get vignetting, not color fringing.

Brian

Leon Obers said:

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

Click to expand...

Very interesting test. But in one way not surprising. The 17-35/2.8
AFS Nikkor is designed specially to use with the digital SLR's. I
guess the lens has a more retrofocus design than the wide angle
primes of the past. So the light is coming more in a perpendicular
way to the CCD than the wide angle lenses. You have to be in your
mind that the color fringing is not a representation of the lens
quality only, but for digital has to do with the character that
light hits the CCD in an angle. An interesting message in the past
about this subject you can see here:
http://www.dpreview.com/forums/read.asp?forum=1021&message=2165556

--
Leon Obers

Click to expand...

Hi Ron:

I've never tried either the 24/2.8 or 28/2.8, although I have heard speak highly of them as you do. One lens that really interests me for large-aperture wide angle work is the 28/1.4. According to raytracing analysis of the patent formula this lens should give very high contrast results at f/1.4, although I don't know of many people who have one. I think it might make a nice companion to the 58mm Noct-Nikkor for available light stuff. The 35/1.4 just doesn't do very well when used wide open.

Brian

Ron Reznick said:

Ron

brian said:

At wider apertures things get a bit more complex. At f/2.8 the
zoom performs extremely well at 35mm, and in fact is superior to my
35/1.4 prime at that aperture. However, at shorter focal lengths
the zoom has reduced contrast wide open, although resolution is
still very high. The contrast picks up rapidly as you stop down.

I was most interested in performance at f/8 because that is where I
tend to shoot wide angle lenses most often.

Brian

Julien Landa said:

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Interesting test Brian.

Click to expand...

What about the mid section and other f-settings?

http://www.julienlanda.nl

--
Julien

Click to expand...

Click to expand...

--
Ron Reznick
http://digital-images.net
http://trapagon.com

Click to expand...

This test only evaluates a few parameters of these lenses, while leaving some important ones out, namely:

• resistance to flare and ghosting
• image quality at infinity and close up (30cm)
• performance well stopped down and wide open
• performance on a full-frame 35mm

I mostly agree that the AF-S 17-35 is superior to many of the primes out there optically, especially at f/5.6-11, where its performance is consistent across the zoom range.

the points where the zoom stands out are

• convenience
• image quality WITHOUT FILTERS
• minimum focusing distance
• construction

the points where it is inferior to the primes in general are:

• put any filter (I mainly use polarizer on the D1x, since color correction filters are not needed any more) on it and you will be amazed by the bad ghosting (often green spot).
• incompatibility with K-1 ring. This is a big deal for me, since the K-1 ring significantly extends the possibilities of several primes that focus close by themselves (mailny AF 20/2.8, AIS 28/2.8)
• bulk, although it is less bulky than a set of primes.

the situations where I still prefer the primes:

• at 28mm, up to f/8 it des not come close to the quality of the AF 28/1.4, sharpness, coma. This lens ((the only nikkor who is superior to a Leica summilux IMHO) has no competition in quality. Believe it or not, it is worth its sky-high price.
• at 28mm, close-up, i is bettered significantly by the AIS 28/2.8, which incidentally can take the K-1 ring .
• at 28mm (still!), it is bettered at f/8 and beyond by my Schneider-Kreuznach 28/2.8 PC (a wonderful lens indeed since PC is invaluable at this focal length).
• at 28mm, the 28/2 AI has better flare/ghosting resistance and I find it sharper f/2-f/5.6.
• at 35mm, it is bettered in resolution at f/4-5.6 (not at f/8 and 2.8 though) by the 35/1.4 AIS IMHO
• at 24mm, I find the 24/2.8 AF-D slightly better in sharpness, although the zoom has zero barrel distortion there and the prime has little.

Don't get me wrong, I love this lens and use it much more often than the primes. However, I would not depart from the primes for the reasons mentioned above.

Philippe

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Click to expand...

Hi Philippe:

Don't worry, I don't think that I will take your comments the wrong way. Actually I appreciate them a great deal.

Its true, I've only done a very limited test that reflects some of the qualities that are most important to me. I think, though that since I did my testing at 100x it is very representative of performance from infinity down to about 10 feet or so. Aberrations don't change too rapidly with a change in magnification. True close-up distances are another matter of course. I can only state anecdotally that I've had very good results with the 17-35 shooting at these distances at f/8, particularly when shooting museum subjects in glass cases. I don't think that I'm in your league when it comes to shooting wide-angle closeups.

I had heard of the flare problem when a filter is attached, although I didn't realize it was as severe as you describe. I'll use your evidence to avoid buying a filter for this lens!

I've been very curious about the 28/1.4 for some time, and your real-world results confirm what I've learned from raytracing the patent example. Much appreciated.

I'll have to take a closer look at the 35/1.4 at f/4 and f/5.6. I've normally used this lens either at f/2.8 or f/8-f/16. It seems to have a bit of astigmatism at wider apertures that limit corner performance even though axially it is extremely sharp at f/2.8.

Thanks,
Brian

Philippe Rigault said:

• resistance to flare and ghosting
• image quality at infinity and close up (30cm)
• performance well stopped down and wide open
• performance on a full-frame 35mm

I mostly agree that the AF-S 17-35 is superior to many of the
primes out there optically, especially at f/5.6-11, where its
performance is consistent across the zoom range.

the points where the zoom stands out are

• convenience
• image quality WITHOUT FILTERS
• minimum focusing distance
• construction

the points where it is inferior to the primes in general are:

• put any filter (I mainly use polarizer on the D1x, since color

correction filters are not needed any more) on it and you will be
amazed by the bad ghosting (often green spot).

• incompatibility with K-1 ring. This is a big deal for me, since

the K-1 ring significantly extends the possibilities of several
primes that focus close by themselves (mailny AF 20/2.8, AIS 28/2.8)

• bulk, although it is less bulky than a set of primes.

the situations where I still prefer the primes:

• at 28mm, up to f/8 it des not come close to the quality of the AF

28/1.4, sharpness, coma. This lens ((the only nikkor who is
superior to a Leica summilux IMHO) has no competition in quality.
Believe it or not, it is worth its sky-high price.

• at 28mm, close-up, i is bettered significantly by the AIS 28/2.8,

which incidentally can take the K-1 ring .

• at 28mm (still!), it is bettered at f/8 and beyond by my

Schneider-Kreuznach 28/2.8 PC (a wonderful lens indeed since PC is
invaluable at this focal length).

• at 28mm, the 28/2 AI has better flare/ghosting resistance and I

find it sharper f/2-f/5.6.

• at 35mm, it is bettered in resolution at f/4-5.6 (not at f/8 and

2.8 though) by the 35/1.4 AIS IMHO

• at 24mm, I find the 24/2.8 AF-D slightly better in sharpness,

although the zoom has zero barrel distortion there and the prime
has little.

Don't get me wrong, I love this lens and use it much more often
than the primes. However, I would not depart from the primes for
the reasons mentioned above.

Philippe

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Click to expand...

Click to expand...

brian said:

Actually, the degree of telecentricity, or perpendicularity of the
off-axis ray bundles, of the lens has nothing to do with the
appearance of color fringing.

Click to expand...

I don't know. In my experiance it can have several errors. E.g. in past I had a PC Nikkor 28 mm shift lens. Using some kinds of film with a bad anti-"halo" correction, you did see "double" contours when shifted the lens in a high degree. But using another film, the error was not there (or at least very limited).

Last year I had a Fuji S1. It has a bad green cast in the left corner. The green cast was more heavy using wide-angle lenses and especially when using a PC Nikkor shifted in a high degree. It seems if the green cast is added or multiplied. See:
http://home.iae.nl/users/lobers/Fuji_S1/felberg/felberg_1.htm
Or the full story (not updated anymore).
http://home.iae.nl/users/lobers/Fuji_S1/index.htm
But not of those pictures do give more color fringing.

Could it be that there also is a difference between the CCD solutions from one brand or another?

Member said:

Consider this: the 55mm/2.8 Micro
Nikkor is actually less telecentric than any of the wide angle
primes in my test

Click to expand...

Is it?

I think the 55mm/2.8 Micro Nikkor is a less "retrofocus" lense than a wide angle lens. But maybe has less to do with exit pupil location (I don't know).

It should be a nice test if you could use an old Nikkor-O 21 mm / f 4 to see if it should have much or less fringing by its design. Unfortunately you can not use it on a D1X, because you can not lock-up the mirror for shooting.
http://www.nikon.co.jp/main/eng/d-archives/lens/nikkor_s/21-4_e.htm

Member said:

I do recall once seeing an image shot with a wide-angle Nikkor lens
on a full-frame CCD attached to the Horseman Digiflex camera.
There appeared to be no problems with this image, although I
haven't been able to relocate it.

Click to expand...

What I say earlier, I think different brands make different CCD's. What could be a problem for one, maybe it is not a problem for another. I did see examples of a 14 mm Nikkor used on a D1X. The corners had very less noticable sharpness than the center. I don't know if it is by the lens or the CCD.
Do read some remarkable other test in this:
http://www.naturfotograf.com/lens_wide.html

Nevertheless it is interesting to see your testresults and to exchange our thoughts about it.--Leon Obers

Hi Leon:

Perhaps the S1 sensor had some sort of problem, although as you point out a generalized green cast is quite different from color fringing, which is a very localized radially symmetrical aberration.

I've also got a 28mm PC lens, and it does suffer from a considerable amount of color fringing, especially in the extreme outer parts of the field that are exposed only when you shift the lens. I'm not quite sure what you mean by "double contours", however. My disappointment with this lens is one of the things that prompted me to get into high resolution stitched mosaics. As I recall, you were the person who actually alerted me to Max Lyons' website, which really got me going.

Telecentricity is completely determined by the exit pupil position. In a perfectly telecentric lens the exit pupil is infinitely far from the image plane. In a simple biconvex lens the exit pupil is one focal length away from the image plane. In reversed telephoto lenses the exit pupil is typically (but not necessarily) farther from the image plane than the focal length. The exit pupil locations for a few Nikkor primes are as follows (bigger number means more telecentric):

15mm f/3.5 AI: 65.7mm
20mm f/2.8 AFD: 53.9mm
24mm f/2 AIS: 57.1mm
28mm f/2 AIS: 61.9mm
35mm f/1.4 AIS: 73.6mm
55mm f/2.8 Micro AIS: 57.1mm

I calculated these numbers from the patented lens prescriptions. As you can see, the 55mm/2.8 micro Nikkor is actually more telecentric than the 20mm/2.8, and equal to the 24mm/2.0. Thus, as far as the CCD is concerned the 55/2.8 and 24/2.0 are identical when they are stopped down to f/8. Except that the 24/2.0 has tons of color fringing and the 55/2.8 has none. This is to be expected from the optical design of these lenses.

Brian

Leon Obers said:

brian said:

Actually, the degree of telecentricity, or perpendicularity of the
off-axis ray bundles, of the lens has nothing to do with the
appearance of color fringing.

Click to expand...

I don't know. In my experiance it can have several errors. E.g. in
past I had a PC Nikkor 28 mm shift lens. Using some kinds of film
with a bad anti-"halo" correction, you did see "double" contours
when shifted the lens in a high degree. But using another film, the
error was not there (or at least very limited).
Last year I had a Fuji S1. It has a bad green cast in the left
corner. The green cast was more heavy using wide-angle lenses and
especially when using a PC Nikkor shifted in a high degree. It
seems if the green cast is added or multiplied. See:
http://home.iae.nl/users/lobers/Fuji_S1/felberg/felberg_1.htm
Or the full story (not updated anymore).
http://home.iae.nl/users/lobers/Fuji_S1/index.htm
But not of those pictures do give more color fringing.
Could it be that there also is a difference between the CCD
solutions from one brand or another?

Member said:

Consider this: the 55mm/2.8 Micro
Nikkor is actually less telecentric than any of the wide angle
primes in my test

Click to expand...

Is it?
I think the 55mm/2.8 Micro Nikkor is a less "retrofocus" lense than
a wide angle lens. But maybe has less to do with exit pupil
location (I don't know).
It should be a nice test if you could use an old Nikkor-O 21 mm / f
4 to see if it should have much or less fringing by its design.
Unfortunately you can not use it on a D1X, because you can not
lock-up the mirror for shooting.
http://www.nikon.co.jp/main/eng/d-archives/lens/nikkor_s/21-4_e.htm

Member said:

I do recall once seeing an image shot with a wide-angle Nikkor lens
on a full-frame CCD attached to the Horseman Digiflex camera.
There appeared to be no problems with this image, although I
haven't been able to relocate it.

Click to expand...

What I say earlier, I think different brands make different CCD's.
What could be a problem for one, maybe it is not a problem for
another. I did see examples of a 14 mm Nikkor used on a D1X. The
corners had very less noticable sharpness than the center. I don't
know if it is by the lens or the CCD.
Do read some remarkable other test in this:
http://www.naturfotograf.com/lens_wide.html

Nevertheless it is interesting to see your testresults and to
exchange our thoughts about it.
--
Leon Obers

Click to expand...

--J. Brian Caldwellwww.caldwellphotographic.com

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Click to expand...

The "corners" you see with the D1x aren't anywhere near the actual corners of the 35mm frame, so I would be careful with that. It would also be appropriate to retitle it "..as tested with Nikon D1x"

Regards,

Matt
--Matt [email protected]

If you really are planning on shooting wide and slow (e.g. hand-held work in serious low-light) then the 28/1.4D at $1600 or so is something to look at. Honestly, it is not as sharp at f/2.8 as the 28mm f/2.8 AIS is, and the main reason for using it is the lack of coma in the edges and the fact that you can get reasonable shutter speeds when it's really pretty dark. Reasonable shutter speeds with a 28mm are pretty slow, you know.

Otherwise, the 28/2.8 AIS (manual focus) lens is sharper than anything going at f/2.8. Straight edges and lines too, and the 28/1.4 doesn't do that. Way, way cheaper for a final incentive. No AF though...

Ron

brian said:

Brian

Ron Reznick said:

Ron

brian said:

At wider apertures things get a bit more complex. At f/2.8 the
zoom performs extremely well at 35mm, and in fact is superior to my
35/1.4 prime at that aperture. However, at shorter focal lengths
the zoom has reduced contrast wide open, although resolution is
still very high. The contrast picks up rapidly as you stop down.

I was most interested in performance at f/8 because that is where I
tend to shoot wide angle lenses most often.

Brian

Julien Landa said:

brian said:

I recently put up a page comparing the 17-35/AFS to the 15/3.5 AI,
20/2.8 AFD, 24/2.0 AIS, 28/2.0 AIS, and 35/1.4 AIS. The comparison
is done at f/8 in the corner of a D1x frame. At this aperture and
image location the major differences between these lenses is the
amount of color fringing present. In every case the zoom lens has
less color fringing. The zoom also has less distortion, and I'll
have more on that later.

http://www.caldwellphotographic.com/LensTesting1735vsPrimesAtf8.html

Brian

Interesting test Brian.

Click to expand...

What about the mid section and other f-settings?

http://www.julienlanda.nl

--
Julien

Click to expand...

Click to expand...

--
Ron Reznick
http://digital-images.net
http://trapagon.com

Click to expand...

Click to expand...

--Ron Reznick http://digital-images.nethttp://trapagon.com

Matt Chroust said:

The "corners" you see with the D1x aren't anywhere near the actual
corners of the 35mm frame, so I would be careful with that. It
would also be appropriate to retitle it "..as tested with Nikon D1x"

Regards,

Matt

Click to expand...

Matt:
I dislike needlessly long titles.

Suspecting the worst is not always the wisest thing. In many 35mm format lenses the color fringing is a linear function of radial distance from the image center out to about 12-15mm. After that the color fringing actually improves in many instances.

Besides, the 24x36mm format is getting to be pretty much irrelevant. Especially in this forum.

Brian
--J. Brian Caldwellwww.caldwellphotographic.com

I liked this comparison very much.

Is there somebody,who could do it with the AFS 80-200 lens.

http://www.julienlanda.nl--Julien

Ron Reznick said:

If you really are planning on shooting wide and slow (e.g.
hand-held work in serious low-light) then the 28/1.4D at $1600 or
so is something to look at. Honestly, it is not as sharp at f/2.8
as the 28mm f/2.8 AIS is, and the main reason for using it is the
lack of coma in the edges and the fact that you can get reasonable
shutter speeds when it's really pretty dark. Reasonable shutter
speeds with a 28mm are pretty slow, you know.

Otherwise, the 28/2.8 AIS (manual focus) lens is sharper than
anything going at f/2.8. Straight edges and lines too, and the
28/1.4 doesn't do that. Way, way cheaper for a final incentive. No
AF though...

Ron

Click to expand...

Ron, have you actually used the 28/1.4?

With all due respect, I find that the 28/2.8 AIS is nowhere near the quality of the 28/1.4 at f/2.8 to f/5.6, at all distances. Sharpness and coma don't come close, bokeh of the 28/2.8 is so-so to be polite (like all nikon primes having 7-blade not rounded diaphragms) whereas the 28/1.4 bokeh is absolutely first class. The result is that the images obtained with the 28/1.4, in addition to being truly sharper, have much more snap and have a unique quality. At 2.8, I prefer also the 17-35 to the 28/2.8 AIS by far. Both the 28/1.4 and the 17-35 are fine portrait and low light lenses, whereas I would never think of using my 28/2.8 AIS to do close portraits, as well as low light because it has ghosts easily.

The 28/2.8 AIS is a very good lens closed down (5.6-16) and at distances up to 6 feet, and adding extension retains its superb image quality. So it is the perfect lens for flower wide perspectives and product shooting. Wide open and at infinity, I have found nothing special about this lens.

Best regards,

Philippe

I did the same test with a Kodak DCS 760 and a AFS 17-35 on the zoom lens side and a 2.8 20mm Nikon, 2.0 35mm Nikon prime lens on the other side.
My results are equal to the test with the D1x.
The 17-35mm zoom is a superb lens if the light is in your back.
(I think better than the primes).

If you put a lamp, the sun, a window or anything else like that in your picture you will get completly different results.
The lens flare is a big problem for this lens.

brian said:

I've also got a 28mm PC lens, and it does suffer from a
considerable amount of color fringing, especially in the extreme
outer parts of the field that are exposed only when you shift the
lens. I'm not quite sure what you mean by "double contours",
however.

Click to expand...

Light comes oblique to the film when shifted, goes trough the emulsion, is reflected back by the film layer (reflected up again) by a bad anti halo, comes to another place within the emulsion. So tiny detail contours you see twice.

Member said:

My disappointment with this lens is one of the things
that prompted me to get into high resolution stitched mosaics.

Click to expand...

I think just taking a normal wide angel and correcting the image in perspective afterwards, the result is better than directly with the Nikkor shift lens.

Member said:

As
I recall, you were the person who actually alerted me to Max Lyons'
website, which really got me going.

Click to expand...

Oh, that's nice to hear (again). I know somebody was asking several months ago.

Member said:

Telecentricity is completely determined by the exit pupil position.

Click to expand...

------ etc.

Interesting stuff. I copied the text to my PC to study afterwards. It is a bit complex to me, so I want to have it for later use. Thanks for all of this.
--Leon Obers

Philippe, I have only done a short test with the 28/1.4D. I have however shot the 28/2.8 AIS (0.2m min. focusing distance version) lens quite a bit. It is possible that the 28/1.4 lens I shot at Samy's Camera against my 17-35 at f/2.8 and f/4 (with a few shots at f/1.4 to check the quality) was flawed. I did not rent the lens for a long-term test. Maybe I should some day, but based on the results I got I wasn't all that interested.

My particular sample of the 28/2.8 AIS is quite sharp at short-to-medium distance at f/2.8. I have shot it in museums at f/2.8 and gotten excellent results (see image below). I have also shot it against the 17-35 AFS that I own and prefer it for close-to-medium distance work at wider apertures.

Considering the fact that I have both the 17-35 AFS and the 28/2.8 AIS, it's not likely that I'll get a 28/1.4D, but if I have a chance to shoot one over an extended period, believe me I'll put it to every test I can think of.



Ron

Philippe Rigault said:

Ron Reznick said:

If you really are planning on shooting wide and slow (e.g.
hand-held work in serious low-light) then the 28/1.4D at $1600 or
so is something to look at. Honestly, it is not as sharp at f/2.8
as the 28mm f/2.8 AIS is, and the main reason for using it is the
lack of coma in the edges and the fact that you can get reasonable
shutter speeds when it's really pretty dark. Reasonable shutter
speeds with a 28mm are pretty slow, you know.

Otherwise, the 28/2.8 AIS (manual focus) lens is sharper than
anything going at f/2.8. Straight edges and lines too, and the
28/1.4 doesn't do that. Way, way cheaper for a final incentive. No
AF though...

Ron

Click to expand...

Ron, have you actually used the 28/1.4?
With all due respect, I find that the 28/2.8 AIS is nowhere near
the quality of the 28/1.4 at f/2.8 to f/5.6, at all distances.
Sharpness and coma don't come close, bokeh of the 28/2.8 is so-so
to be polite (like all nikon primes having 7-blade not rounded
diaphragms) whereas the 28/1.4 bokeh is absolutely first class. The
result is that the images obtained with the 28/1.4, in addition to
being truly sharper, have much more snap and have a unique quality.
At 2.8, I prefer also the 17-35 to the 28/2.8 AIS by far. Both the
28/1.4 and the 17-35 are fine portrait and low light lenses,
whereas I would never think of using my 28/2.8 AIS to do close
portraits, as well as low light because it has ghosts easily.

The 28/2.8 AIS is a very good lens closed down (5.6-16) and at
distances up to 6 feet, and adding extension retains its superb
image quality. So it is the perfect lens for flower wide
perspectives and product shooting. Wide open and at infinity, I
have found nothing special about this lens.

Best regards,

Philippe

Click to expand...

--Ron Reznick http://digital-images.nethttp://trapagon.com