Image content, a city view as test object for lens+camera comparisons

[Bernard Delley]

How much image detail can I get with a camera lens combo. This is an important question for somebody like me, who does  a good part of photography while traveling in one or another way. If possible, it should cover the equivalent range from 24mm to a little beyond 100mm for me as a minimum. And it can be done with a single 5x zoom lens. For many years this was realized with the D7000 and the 16-85mm f/3.5-5.6 lens. Lately it has been the D7200 with the 16-80mm f.2.8-4 lens. I was always willing to carry moderately more for a more capable camera body (7xxx), but it seemed impossible to gain significantly in FX without a big step up in bulk and weight.
I have a convenient popular city view point nearby, which I can reach in a 300mm steep uphill hike. Despite a webcam, you never really know how good the sighting conditions really are, when they seem good enough. I never could carry all my gear up in a go, so I limit it to a backpack with maximum of about 8kg of gear.
Recently, the Z 24-120mm f/4 lens appeared and sure should be a step up in image detail at a small step up in weight. So I start the comparison with pictures between these two combos taken on the same day. Both takes freehand, the D7200 16-80 at 38mm f/5.6 and the Z7II 24-120 at 50mm f/4. Both images are out of the camera full size jpg (basic,standard) extracted from the the raw files. As the day was very slightly hazy I upped contrast and in PS +20 and clarity +20 and possibly minor exposure, WB adjustment. (first is D7200 6000x4000 second is Z&II  8256x5504) aha I have to reduce size to stay safely in the bounds (4000, lets see how that looks.  My first impression is image detail the same. Inconsistency in angle of view and WB difference remain the most prominent differences.

[John Geerts]

Interesting Bernard.  Have you ever tested the combination with the  Nikon D3x and the 105/1.4E or 58/1.4G? 

[Bernard Delley]

thanks John. I stop posting material in this thread. I cannot put up full size images. And the announced second image could not be put up in hours even at the reduced size of the first one. I give up. I have a lot of such test images over the years with D800 and D850 to test and compare the lenses primarily. A selection of 23 images was up in my Zurich gallery at DPReview in 100% and 50% reproduction size. Perhaps I can find a site which allows to post 6.5 MB jpg files of 45Mpix  8865x5504.

[Frank Fremerey]

Hey, Bernhard! You can always upload to a sharing site like Google Drive or Microsoft One Drive, then link here. But you can also share your findings by adding 100% crops of significant details at 100% quality.

For a consitent side-by-side-test you might want to consider using a tripod. In my experience handholding technique plus aerial perspective can be more important for detail rendering than lens quality.

So shots taken in stark sunlight shortly after a good rainfall with a tripod should be a good reference.

[Bernard Delley]

Hi Frank,
there are 23 such "test" images posted already at
https://www.dpreview.com/galleries/0440825497/albums/zurich
we will see how long that survives as an archive.

most of the time I really hate forum discussions that link to info on other sites. But in the present case, be it.

These are not seriously reproducible tests. The sighting conditions and cloud overcast changes. You may detect wiggly architectural lines due to air 'turbulence' with focal length as short as 100mm for sure and maybe for shorter FL.

I take these images free hand or with a monopod, suitable for a usually crowded place. The exposure times are quite short under my preferred sighting conditions. Really serious reproducible tests are my MTF characterizations, you may have seen the thread.  MTF is somewhat abstract, and maybe hard to relate for some. The city views were done to see hot that shows in a realistic practical imaging setting. The MTF insights do show clearly, especially also the sharpness loss for some lenses or TC combos towards the far side edges of the image.

[Birna Rørslett]

NG has a size limit for attachments to posts. Currently this is 2MB for a single file and a total of 10MB for all attachments to any post. There is no de facto pixel number limit, but it hardly makes sense to post jpgs of large files at ridiculously low jpg quality just to get within the max. size limit.

If you wish to show differences, making nice 100% crops is the way to go. Just remember to make them devoid of any additional sharpening and preferably not more than 2000pix on the major axis. This way the forum software will not clobber the displayed image if you right-click and select 'opening in a new window'.

[Bernard Delley]

I totally understand, 2000 pix long side is plenty for most documentation and also for an image with striking content and composition.

As  comparison pictures are already posted at this https://www.dpreview.com/galleries/0440825497/albums/zurich album some in 100% some in 50% full image there is no immediate need to repost them. The latest addition is 100% images for D7200 16-80 at 16mm 38mm and 80mm  f/5.6 viewable at 6000x4000 and Z7II 24-120 at 24mm 50mm and 120mm f/4 viewable at 8856 x 5504 so one can set expectations right for big size prints.

For my recent trip into Grand Canyon I opted for D7200 and AF-S 16-80mm f/2.8-4 (+ AF-P 70-300mm f/4.5-5.6 FX) for my annual calendar 420mm wide. The 24 Mpix images look really good, limited by the print quality of the calendar.  The lesser image quality of an image done with Olympus TG-6 4000x3000 12 Mpix in bright daylight is quickly seen by the astute viewer. But, for some viewpoints it is safer to use a waterproof camera.

below a more or less classical Nankoweap view in 1920 pix wide: