Tilt Shift Lens vs Focus Stacking

[dtech]

I've been attempting to make 3d models of small bugs and rocks and am running into a depth of field issue.  There is no option to get close and carry depth of field throughout the object and still staying around f16.

Is anyone using a tilt shift lens? 

Are there any other solutions? 

Is focus stacking time efficient?  Seems like a ton of work to focus stack every image.

Does Agisoft process your focus stack automagically?

Would love to hear your thoughts?

[Kiesel]

I have read about successful use of tilt-shift lenses in photogrammetry, unfortunately I can't remember where.

Atleast I have found this paper about calibration of tilt-shift lenses:

http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLI-B5/99/2016/isprs-archives-XLI-B5-99-2016.pdf

Karsten

[dtech]

Thanks.  Thats an interesting article.  Im about to rent a lens to test it and will give my results here as well. 

[dtech]

I tested out a 45mm tilt shift lens on a d810 and ran some tests and the tilt shift still lacks the ability to carry depth of field through even a moderately long object.  It did improve it but not good enough to justify the difficulty.

I then tested Helicon Focus and Remote which is a fabulous solution if you are shooting objects for just photography but for photogrammetry I couldn't imagine the headache of stacking every angle.  100 images would become 500 then back to 100.

Is anyone doing this?

Is depth of field an issue for anyone else out there?

Thanks and I look forward to your comments.

[Kiesel]

Hi dtech,

you have some options:

1. Instead of using a full frame Camera, which is known for low depth of field, use a crop camera with more depth of field,  a MFT-camera with even more depth of field or a good point and shot camera with again more depth of field.

2. You can use your tilted photos and mask out unsharp parts. It can be done automatically or semi automatically. (search the forum, there was a discussion about this last year).

3. Perhaps you can even stack your tilted photos (but I have never heard someone is doing this).

Is depth of field an issue for anyone else out there?

Yes, for everyone who is shooting makro photos for photogrammetry!

I then tested Helicon Focus and Remote which is a fabulous solution if you are shooting objects for just photography but for photogrammetry I couldn't imagine the headache of stacking every angle.  100 images would become 500 then back to 100.

Is anyone doing this?

Yes, I have read of someone. Indeed a lot of work, needs some automation.

Karsten

[ptmcinerney]

I work in the field of archaeological photogrammetric survey and I have used the 45mm Nikkor PC-E tilt shift giving excellent results regarding depth of field. Modest adjustment to the swing/tilt angle of the lens results in a dramatic change to the plane of focus and effective DoF along the axis of the object. Erica's article cited above makes an excellent argument for the use of these lenses in photographic acquisition in cultural heritage work. Another important issue is the calibration of the lens. Making a tilt adjustment means that the the principal axis of the lens moves relative to the sensor plane and the optical profile of the lens will need to be calibrated after this movement. There is still research to be done in validating the use of these lenses for metrology purposes in actual field work but things are looking good!

http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLI-B5/99/2016/isprs-archives-XLI-B5-99-2016.pdf

Patrick

[Kiesel]

I came across a paper that might be interesting to extend DOF, but it's not about tilt-shift-lenses:

"An automated device for the digitization and 3D
modelling of insects, combining extended-depth-of-field
and all-side multi-view imaging"

https://pdfs.semanticscholar.org/8afa/56e3a65d276c04e1e2f8396feacb36d5befa.pdf

Regards

[Mike25660]

Focus stacking can be a lot of work. It takes some patience but it can be somewhat automated. Anyway, it will work with photogrammetry projects.

https://www.jpaleontologicaltechniques.org/pasta3/JPT%20N18/Bulletin.html