Naptown Gaijin

Because a tilt shift lens has a changeable focal length as projected on to the sensor, I do not see how a photographer can use it for precise photogrammetry. I used to take photos of objects for use by people who would analyze the photo forensuration, the measuring of an object by calculating focal length, distance to object, etc. You have probably seen crime photos where an object of known length is included in a photo, usually a ruler or yard stick. By analyzing the length of the yardstick, you can, within parameters evaluate the densions of an object of hitherto unknown size. I used a 50mm f1.4 lens for photos of vehicles, and we could tell the size of the vehicle within small tolerances. Don't see how you can do that with a tilt shift lens as you are adjusting it so that the side of the building look straight.
In other words, you are making it look good, not taking a photo than can be used for measurent or mapmaking or intelligence analysis. Zoom lenses have the same problem, in that the precise focal length is unknown, except when you are at either end, though they can be used for a close approximation when the photo is taken at either end of the zoom range. There is always some distortion in ANY lens, the difference is in degree. Space telecopes have less, cheap binoculars have more.... same with camera lenses. Thisay be one reason why pros carry a bunch of fixed lenses rather than one long zoom....less distortion, in addition other advantages.
The OP needs an accurate method of reording densions for an excavation project; itmcould be for a nuclear power plant containment vessel, who know? In such a case, I am confident a tilt shift lens would not be used when precise meaaurents are needed.

spencehg1

We are doing a wide variety of archaeology photogrammetry for the Forest Service. Check out some of our models at Sketchfab.com/FOFSedona. We primarily use Nikons and Agisoft Photoscan. You have many options. We have successfully modeled narrow deep, inaccessible trenches and holes with a Gopro on a pole. For the vast majority of your work just have a reasonable camera and lens leaning towards the wide angle range. We use both 10 and 24mp camera bodies. It is going to be rare if ever that you will see a difference in the final result. Once you get the hang of the photo acquisition with even lighting and good overlap you will get consistent good results. There is a learning curve for both acquisition and processing but it comes quickly. The biggest cost we have had is that we do some large very complex models and the software, regardless of processing time (we have a 12 processor old workstation) is a RAM hog. Basic processor may take a bit longer but you need the RAM for any significant number of input photos.
Bottom line, start out with whatever cameras you have and worry about the actual acquisition and processing.

JohnSwanda Loc: San Francisco

Converging lines are not distortion. They are actually what the camera sees. Our eye compensates when we look at something like tall buildings, so we don't notice it. The tilt/shift lens distorts the photo into having parallel lines which is more pleasing to the eye, but it is still distorting the photo. That distortion will affect the accuracy of photos used for photogrammetry. Do you even know what photogrammetry is?

dsmeltz Loc: Philadelphia

I agree that a tilt/shift would be great for this use, but the OP expressed the existence of budget restraints. One thing T/S lenses are not is affordable. I believe a lot of the Photogrammetry programs mange to work around distortion through the number of shots involved.

aflundi Loc: Albuquerque, NM

Maybe we are thinking of different scenarios.

There are certainly many types of photogrammetry and many different types of scenes to be measured, which ultimately may require different methods for optimal use, but if you've cleared an area and made it reasonably flat, having a flat, undistorted image would allow easy measurements.

For instance, if you placed a rectangular frame with a meter stick on each end and then shifted a TS-lens so that the meter sticks on each end of the frame filled the appropriate side of the camera frame, you'd have an undistorted image that would make measurements not only simpler, but trial. In addition, by applying the correct tilt to the lens, you'd have better focus of the area. No distortion, no out-of-focus parts of the image.

dsmeltz Loc: Philadelphia

My understanding of Photogrammetry programs (which I admit quite is limited) is that you are either imputing information regarding points in each photograph that are common/related from photo to photo or the program is automatically analyzing the photographs and identifying these points. It then looks at relative differences from photograph to photograph to triangulate how they relate in real space. I am not sure that that process would be either helped or hindered by the use of a T/S lens. I think as long as the lens is constant, any quality lens would still produce data on relative points that would be accurate.

Toment Loc: FL, IL

There’s an app on my IPhone called “measure”
Take a look...

Architect1776 Loc: In my mind

Interesting subject.
For buildings, especially interiors I use a scanner that does the whole interior.
I get the correct dimensions of the building, all the components etc.
Of course it is all on the computer and not prints and don't know how prints work in this case but the scanner camera is very good for this and creates an accurate to scale model with correct dimensions.

lamiaceae Loc: San Luis Obispo County, CA

You did not state photogrammety of what. An archaeological site, a building construction site, ecological? For anything scientific where you need low distortion and image "accuracy". I would think you would want to use a Prime Normal lens! Say a 50mm Lens for a Full Frame (or Nikon FX) camera; or a 35mm Lens for a Crop Factor / APS-C (or Nikon DX) camera. I used to work in a museum and we pretty much used various "primes" or "symmetrical" lenses for copying art work and objects. We shot film in 35mm, 2 1/4", 4x5", and 8x10" sizes! Yes, this was some years ago!

Architect1776 Loc: In my mind

He said trenches.

dsmeltz Loc: Philadelphia

The second sentence says "an excavation"

robertjsmith

Any good sturdy camera will do what you want beyond what the eye can see, you just have to practice and get used to the techniques involved which can be very different than what most people are accustomed to.
Bob

lamiaceae Loc: San Luis Obispo County, CA

Both terms can be used in both fields. My Father-In-Law was an architect, civil engineer, and building contractor. And I worked on an archaeological dig in a college class. And did ecological field studies for a biology class. Again the terms are used across fields.

In any case asking here is probably not the best place for specific info on a technical application. There should be industry or field standards and SOPs for what ever he is doing. The equipment needed should be "spelled out".