20mm for landscape

[David H. Hartman]

If you shoot across multiple platforms, the phrase 1:1 can become meaningless.

At 1:1 a bee that is 2.0 cm long in life will [be] 2.0 cm long on 36x24 and 2.0 cm long on 6x6 and still 2.0 cm long on 4x5. 1:1 means exactly the same thing on all formats.

Dave Hartman

[JKoerner007]

At 1:1 a bee that is 2.0 cm long in life will 2.0 cm long on 36x24 and 2.0 cm long on 6x6 and still 2.0 cm long on 4x5. 1:1 means exactly the same thing on all formats.

Dave Hartman

Point missed, but I think I am at fault for my prior wording.

1) I realize it's not the bee that changes, my point was it's the framing.

2) I also wasn't speaking of 6x6 or 4x5 ... no one uses these for bee or macro photography.

I understand that 1:1 magnification = 1:1 magnification; that's a tautology.

So, yes, the reproduction ratio remains the same.

The point that is continuously being missed is 1:1 on a 36mm sensor might be too small ... and on another sensor it might be too large.

I can see how I worded it wrong previously, so let me re-phrase:

Let's say I am trying frame a 16mm subject with a 2:1 lens.

At its closest distance, a 2:1 lens reproduces 18mm of subject across a 36mm sensor, doubling the reproduction ratio.
So a 2:1 lens would be perfect for trying to capture a tight shot of 16mm subject (if I shoot FF).
The 16mm subject has its reproduction ratio doubled, to 32mm, which fits nicely on my 36mm FF sensor.

However, if I tried to use that same 2:1 lens on a crop 1.5x DX sensor, this means I will be trying to place 32mm of augmented subject on my 24mm sensor ... so some part of the subject will be clipped off.

When shooting macro, I can't just think 1:1 or 2:1 magnification; I have to think of 1) the size of my subject, 2) the magnification, and 3) the size of my sensor.

Saying 1:1 (by itself) is meaningless for macro composition.

These 3 questions have to be asked:

• What size subject is being considered for the macro shot? (4mm? 12mm? 30mm?)
• What is the reproduction ratio of the lens?
• What size sensor am I using?

If I know a 20mm AI-S lens offers 3.4x magnification, reversed, this means a 10 mm subject will be magnified to 34mm, which will fit on a 36mm FX sensor.

That same 10mm subject will magnify to the same 34mm using the 20mm AI-S on a DX ... but 34mm won't fit on a 24mm DX sensor.

If I know the 20mm AI-S offers 3.4x magnification, reversed, this means (if I am shooting a DX) that sensor can only handle ~7mm (with that 3.4x multiplier) before the reproduction ratio exceeds a 24mm sensor.

[Per Inge Oestmoen]

I never heard an earnest praise for the Samyangs except when it comes to price performance relation esp for the fisheye

If I was in the situation I'd sure get the 20mm/f=1.8G

I do not have a Samyang, but I have tried the Samyang 14mm 2.8 once. That was enough; The Samyang is by no means a brilliant performer and its mechanical construction was rather unconvincing - and that was an understatement. It was not very sharp in the center at 2.8 and the corners were so-so or worse. I believe that any lens has to be judged from its absolute performance. The Samyangs are inexpensive, but unfortunately the low price does little to improve the optical and mechanical quality.

[Per Inge Oestmoen]

Thank you ... and, yes indeed, they are

Yes, exactly. In fact, the second image (the fly) was taken with the 28mm; the first was with the 20mm. Effectively, AI-S lenses (reversed) provide the following magnification beyond a standard 1x macro:

• A 50mm f/1.2 AI-S lens reverses to:   1.1x ( = 32.7mm edge-to-edge sensor coverage)
• A 35mm f/1.4 AI-S lens reverses to:   1.8x ( = 20.0mm edge-to-edge sensor coverage)
• A 28mm f/2.8 AI-S lens reverses to:   2.1x ( = 17.1mm edge-to-edge sensor coverage)
• A 24mm f/2.8 AI-S lens reverses to:   2.6x ( = 13.9mm edge-to-edge sensor coverage)
• A 20mm f/2.8 AI-S lens reverses to:   3.4x ( = 10.6mm edge-to-edge sensor coverage)

[...]

We agree again ... which is why I shoot only AI-S lenses (with the exception of a super-telephoto). They're lighter, just as good IMO, and much-much more versatile

I must say that my 20mm 1.8 AF-S is superior to my 20mm 2.8 AIS, but as already mentioned this is relevant only in the corners and at two largest apertures. They are both eminently capable of being used for everything from informal snapshots to the most demanding tasks within commercial photography.

We fully agree that many of the manual Nikkors are true gems, and if one wants to use this macro capability their usefulness suddenly is above that of the newer designs unless one absolutely must have autofocus.

I have been a Canon user since 2002, and the sole reason why I chose to buy D750 was Nikon's admirable backward compatibility which enabled me to use my manual Nikkors - mostly acquired second hand - with a modern digital body. Perhaps I should have saved up for the D810, but after all the D3X also has 24 MP and if memory serves it received very high praise as a landscape camera when it hit the market. In my case, the high-quality manual Nikkors were the direct reason why I decided to add the D750 to my toolbox. My experience since then tends to justify that decision.

[Bjørn Rørslett]

Whether it's 1:1 or 5:1 is immaterial.  The magnification is the principal determinant of capture parameters, not the angle of coverage.

Isn't it about time to return to the actual topic of this thread?

[JKoerner007]

I must say that my 20mm 1.8 AF-S is superior to my 20mm 2.8 AIS, but as already mentioned this is relevant only in the corners and at two largest apertures. They are both eminently capable of being used for everything from informal snapshots to the most demanding tasks within commercial photography.

At f/8 (which is where I shoot most landscapes) the 20 AI-S is pretty sharp. Colors are very nice too.

At f/4, where I shoot most macro-stacks, it is very sharp in the center, good colors also, and the corners really don't matter.

We fully agree that many of the manual Nikkors are true gems, and if one wants to use this macro capability their usefulness suddenly is above that of the newer designs unless one absolutely must have autofocus.

Indeed. For my purposes, I have no need for AF on a wide lens; only a tripod.

I have been a Canon user since 2002, and the sole reason why I chose to buy D750 was Nikon's admirable backward compatibility which enabled me to use my manual Nikkors - mostly acquired second hand - with a modern digital body. Perhaps I should have saved up for the D810, but after all the D3X also has 24 MP and if memory serves it received very high praise as a landscape camera when it hit the market. In my case, the high-quality manual Nikkors were the direct reason why I decided to add the D750 to my toolbox. My experience since then tends to justify that decision.

We have a similar background, as my first years were with Canon also. Unlike you, the AI-S world was a delightful surprise for me ... and now I am a Zeiss/Cosina fan because they still make all-metal lenses (although Zeiss' switch to the Milvus style is a disappointment, because of the rubber focus ring). The fact I was able to get a Zeiss 15mm Distagon, and the Apo-Sonnar at tremendous discount was a nice bonus.

Since using AI-S lenses, and since shooting the classic Zeiss, I am now interested in the pre-AI Nikkors, because they have the scalloped-metal focus rings, which is just a lot classier/nicer, IMO.

[pluton]

I can personally recommend the Zeiss 21/2.8 as a "landscape 20".  It's imaging performance towers above the 20mm Nikkors I have used (20/3.5 Ais, 20/2.8 Ais) except maybe the 21/4 mirror lockup Nikkor.   It is larger and heavier than the Nikkors, however.

[Per Inge Oestmoen]

I can personally recommend the Zeiss 21/2.8 as a "landscape 20".  It's imaging performance towers above the 20mm Nikkors I have used (20/3.5 Ais, 20/2.8 Ais) except maybe the 21/4 mirror lockup Nikkor.   It is larger and heavier than the Nikkors, however.

I have experienced that the Zeiss lenses are good, but not otherworldly good. I can testify that the Nikkor 20mm 1.8G ED AF-S is superior to the Zeiss 21mm 2.8 in the corners and equal in the center. In other words, the Nikkor 20mm 1.8-version is cheaper and optically better than the Zeiss 21 mm. However, when compared to the Nikkor 20mm 2.8 AIS the Zeiss is clearly sharper in the corners and evenly sharp over the whole frame. But the Zeiss is not better in the center than the 20/2.8 AIS and with both at 2.8. When stopped down to 5.6 the Nikkor 20/2.8 AIS is up to the task also in the corners and almost equal to the Zeiss 21 mm.

There is also a quirk: With my Zeiss lenses I regularly have to dial in +1 exposure compensation since the camera underexposes with one stop particularly in dark environments. That is with Canon versions, I do not know whether such is the case with the Nikon versions but assume that it is. This is inconvenient.

The Zeiss 15mm 2.8 Distagon is extremely well suited for architecture and interiors. I have never seen any better lens for such purposes. It shows impressively little light fall-off in the corners, nice colors and is razor sharp at close and middle distances. However, at infinity the things change a bit. It seems optimized for shorter distances and is not the best performer at infinity. I have found that the Canon 16-35 4.0 IS is clearly superior to the Zeiss 15mm when both are at infinity and the Canon is at 4.0 and the Zeiss at 2.8.

I also had the opportunity to test the Canon 16-35mm 2.8 III against the Zeiss 15mm 2.8 at different distances and with both at 2.8. It turned out that the Canon performed better at infinity, but that it also had more light fall-off with darker corners. At middle and short distances they were pretty equal, but the Zeiss 15 mm remained better in the corners.

[Per Inge Oestmoen]

At f/8 (which is where I shoot most landscapes) the 20 AI-S is pretty sharp. Colors are very nice too.

At f/4, where I shoot most macro-stacks, it is very sharp in the center, good colors also, and the corners really don't matter.

I would say that the 20/2.8 AIS is very sharp in the center from 2.8 and very good in the corners from 5.6.

[JKoerner007]

The Zeiss 15mm 2.8 Distagon is extremely well suited for architecture and interiors. I have never seen any better lens for such purposes. It shows impressively little light fall-off in the corners, nice colors and is razor sharp at close and middle distances. However, at infinity the things change a bit. It seems optimized for shorter distances and is not the best performer at infinity. I have found that the Canon 16-35 4.0 IS is clearly superior to the Zeiss 15mm when both are at infinity and the Canon is at 4.0 and the Zeiss at 2.8.

I also had the opportunity to test the Canon 16-35mm 2.8 III against the Zeiss 15mm 2.8 at different distances and with both at 2.8. It turned out that the Canon performed better at infinity, but that it also had more light fall-off with darker corners. At middle and short distances they were pretty equal, but the Zeiss 15 mm remained better in the corners.

Have to agree with you: the main purpose for my Zeiss 15mm is for indoor photography/scene investigations (when I need "the whole room" on the image).

I seldom use it for landscape because its extreme width makes you feel "too far away" ... and the 20mm is the lens I go to more often. (Not to mention the latter is lighter.)

I've never tested the 15mm formally, but I think you're right in the sense the extreme detail and micro-contrast are found up-close, to mid-distance, where the Zeiss 15mm excels.

Never tried the Canon 16-35 III, but (according to LenScore) the Zeiss 15mm beats it pretty handily in most respects:

RESOLVING POWER
Zeiss - 1078
Canon - 871

CONTRAST
Zeiss - 1180
Canon - 809

COLOR
Zeiss - 921
Canon - 910

BOKEH
Zeiss - 975
Canon - 747

STAR
Zeiss - 941
Canon - 765

DISTORTION
Zeiss - 934
Canon - 761

FALLOFF
Zeiss - 789
Canon - 615

FLARE
Zeiss - 864
Canon - 730

LaCA
Zeiss - 889
Canon - 764

LoCA
Zeiss - 973
Canon - 963

TOTAL SCORE
Zeiss - 974
Canon - 804

They don't really go into whether this is close or far; however, the likelihood is these are lab results and, for this reason, measured at a comparatively-close distance.

[chambeshi]

I can personally recommend the Zeiss 21/2.8 as a "landscape 20".  It's imaging performance towers above the 20mm Nikkors I have used (20/3.5 Ais, 20/2.8 Ais) except maybe the 21/4 mirror lockup Nikkor.   It is larger and heavier than the Nikkors, however.

agreed :-)

Zeiss 21mm Distagon f2.8 on Nikon D500 ISO 100, f/16

[chambeshi]

The Zeiss 15mm 2.8 Distagon is extremely well suited for architecture and interiors. I have never seen any better lens for such purposes. It shows impressively little light fall-off in the corners, nice colors and is razor sharp at close and middle distances. However, at infinity the things change a bit. It seems optimized for shorter distances and is not the best performer at infinity. I have found that the Canon 16-35 4.0 IS is clearly superior to the Zeiss 15mm when both are at infinity and the Canon is at 4.0 and the Zeiss at 2.8.

I also had the opportunity to test the Canon 16-35mm 2.8 III against the Zeiss 15mm 2.8 at different distances and with both at 2.8. It turned out that the Canon performed better at infinity, but that it also had more light fall-off with darker corners. At middle and short distances they were pretty equal, but the Zeiss 15 mm remained better in the corners.

Interesting. What optical attributes would cause such discrepancies with distance in the IQ of close-medium vs near infinity subjects ? Are on IQ determinants not invariant in optics of this quality?

thanks

woody

[Bjørn Rørslett]

Optical performance across the entire focusing range is never invariant. The designers maximise performance according to set criteria, so the fact a lens is better for near or far distance comes as little surprise.

[Per Inge Oestmoen]

I've never tested the 15mm formally, but I think you're right in the sense the extreme detail and micro-contrast are found up-close, to mid-distance, where the Zeiss 15mm excels.

Never tried the Canon 16-35 III, but (according to LenScore) the Zeiss 15mm beats it pretty handily in most respects:

[...]

TOTAL SCORE
Zeiss - 974
Canon - 804

They don't really go into whether this is close or far; however, the likelihood is these are lab results and, for this reason, measured at a comparatively-close distance.

In fact, I almost returned my Zeiss 15mm 2.8, until I discovered that it was indeed brilliant at short and medium subject distances. Now when I have understood that it is superior at the distances for which it is optimized I am content with that since architecture and indoor photo is something that deserves a dedicated lens.

However, it must be stated that the Zeiss Distagon 15mm is not superior when it comes to infinity. In that respect, the numbers that seem to illustrate that the Zeiss is generally superior are misleading. I recommend the Zeiss, but there are better alternatives for landscape and subjects at infinity.

I and two friends tried the Nikkor 14-24 2.8 AF-S, the Zeiss Distagon 15mm 2.8 and the Canon 16-35mm 4.0 IS together against a landscape at infinity. All three lenses were set at their max aperture. That means the Canon was at 4.0 whereas the Zeiss and the Nikkor were at 2.8.

The Zeiss gave the weakest rendering of detail, then came the Nikkor, and the Canon 16-35 4.0 IS was superior.

I concluded that the days when Canon users need to use the Nikkor 14-24 in order to get a true UWA zoom are gone. The Canon 16-35mm 2.8 III is also an excellent performer at infinity, but I am afraid that it shows significantly more light fall-off in the corners than the Nikkor 14-24 which is admirable in that department. It would have been interesting to have the Canon 16-35mm 2.8 III together with the others above.

[JKoerner007]

... next page