5
\$\begingroup\$

I've been doing some architectural photography and indoors with the Nikon 14-24 f2.8 lens.

Most of the time I can get my camera high enough to frame my shot and keep vertical lines vertical. But there are times when I want a low- or high-angle shot, so I end up using the new "Upright" tool in Lightroom to straighten my photos out in post.

I've never used or owned a tilt-shift lens before, so my question is how much of a difference is there between photos that have the perspective fixed in post as opposed to being shot with a tilt lens?

\$\endgroup\$
1
  • 3
    \$\begingroup\$ This video might be helpful. While not showing the difference between photos fixed in post and tilt-shift, it does show a nice example on how tilt-shift works in practice. \$\endgroup\$ Commented Aug 2, 2014 at 10:13

2 Answers 2

4
\$\begingroup\$

This is a variation of the question, "Should I capture the final image in-camera or create the image in Post Processing?".

The answer depends on your goals for the image. I use both approaches. I like the photographic challenges of capturing everything in camera. So, sometimes my set up looks like a Rube Goldberg contraption. On the other hand, in some cases I could only get the image in post processing. Remember, even Ansel Adams was adept at post processing; read about his efforts to burn or dodge parts of an image in his darkroom.

However, there are a couple of items to consider:

One consideration, is time. In general, I find that if I can capture the image in camera, I spend disproportionately more time outside photographing and less time in front of a computer.

Another consideration is use of the image. In this case, the tilt shift lens will minimize artifacts that might be created in post processing. Publishers may prefer to have the image without those post processing artifacts.

Also, take a look at Matt Grum's comment below - he adds a couple of additional, thoughtful considerations.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ I think there's more to it than just the philosophy of doing it in camera vs doing it in post. For example perspective correction in camera requires a tilt lens with a larger image circle, and sharpness can decrease rapidly at full shift, on the other hand software correction requires upsampling of parts of the image and cropping, both of which reduce detail. There are probably cases where one technique gives overall better results and vice versa. \$\endgroup\$
    – Matt Grum
    Commented Aug 3, 2014 at 13:45
  • \$\begingroup\$ @MattGrum - Good points - I edited my response to point out your thoughtful comment. Thanks. \$\endgroup\$
    – B Shaw
    Commented Aug 3, 2014 at 21:20
1
\$\begingroup\$

I have the Nikon 28 PC AIS. You can often find it used at a good price, mine was $100. The lens is a little soft at the corners and chromatic abberation increases the more you shift (easily fixed in post). However, it's not a bad lens at all and if you're primarily worried about converging verticals then it may be all you need.

I strongly prefer shifting with a lens to doing it in post as I find the artifacts unacceptable. If you are interested in architectural photography, you might also consider getting a view camera which will give you MUCH more in terms of perspective correction, changing the plane of focus, etc. Also, you could probably find a view camera with lens for far less than a modern TS lens. However, if you need quick turn around or don't have access to a good lab then it's probably not what you want.

Additionally, if you have a camera rental store near you I would try the TS lens before you buy. You may find it to be an unnecessary expense.

\$\endgroup\$
4
  • \$\begingroup\$ The view camera idea is a good one, but really only for product shots (or anything else at a relatively long distance). The problem being that it's difficult to find lenses that are both short/wide enough and that have a large-enough image circle to allow for movements at anything like a reasonable price - as often as not, you'll wind up using a wide-angle 35mm-format T/S lens anyway for its increased image circle while taking advantage of the independent front and back movements on the view camera. (Back movements are so much handier than adjusting the whole camera.) \$\endgroup\$
    – user28116
    Commented Aug 4, 2014 at 12:15
  • \$\begingroup\$ If I'm not mistaken, the relationship between the tilts and shifts on the Nikon TS lenses are fixed which is another strike. The 90mm (equivalent to around 30mm on 135) Nikon SW f8 covers 8x10 and can be found easily for 400 or less used. It's also copal 0 which cuts down on weight. As long as you're willing to shop used I think you can find similar "deals" on shorter lengths. Totally agree on rear movements. \$\endgroup\$
    – moorej
    Commented Aug 4, 2014 at 17:41
  • \$\begingroup\$ It's equivalent to a 90mm lens if you want to use it with your DSLR; that rules it out for a lot of architecture. Like I said, a view camera great for other things, but you'll usually wind up finding it too long for the photos that drive a lot of people to the T/S lenses. (If it were easy, Schneider would offer an enormously expensive lens shorter than their 50.) \$\endgroup\$
    – user28116
    Commented Aug 4, 2014 at 20:41
  • \$\begingroup\$ @user28116 I was talking about shooting 4x5, I should have been more specific. \$\endgroup\$
    – moorej
    Commented Aug 4, 2014 at 20:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.