Using Scheimpflug

[GFS]

Hi,

I am using Helicon extensively on a new book I'm working on and am using an Arca-Swiss view camera.

I am wondering what might be the best way to handle this sort of scenario, in terms of where I position the plane-of-focus using front element movements (tilt/swing)?



Supposing I have 2 or more vertical 'blocks', where I am seeing one or two edges as illustrated.

Question : Is it best for Helicon to have the plane-of-focus vertically (aligned with the front face of the blocks) or does it make no difference how this is shot?

Specifically, I am thinking about how Helicon deals with the blurred edges around sharp objects (halos). I am typically shooting around 30 to 60 images for this sort of shot.

Any help much appreciated!

[Stas Yatsenko]

That is a very interesting question! Indeed, the simplest approach is to shoot images with vertical plane of focus. The danger is that edge of the foreground block will be blurred optically and will mask details of the block behind it. Usually we advise to set smaller aperture to minimize this optical effect. With tilt/shift lens this can also be achieved by tilting plane of focus.


Though I am not sure about possible artefacts or distortions associated with this method. I would appreciate if you let us know about your experiments in this area.

[GFS]

Stas,

the result of my findings so far is my feature request : http://goo.gl/mFrOzk

Essentially it appears to me that the only way to deal with these kinds of scenarios is to use multiple renders (2 is most likely sufficient). B is giving me best results overall, with a pleasing look, but C is the only way to get the overlapping areas sharp. However, its inherent contrast is way to high for my work, so I can only use it to 'repair' the B images.

[Stas Yatsenko]

Is the contrast the only downside of method C? Have you tried to combine colors from method B (low pass) and details from method C (high pass)?

[GFS]

Is the contrast the only downside of method C?

I would say it's contrast that is the disadvantage. There seems to be a 1/4 tone to mid-tone increase in contrast that doesn't work so well for me.

Have you tried to combine colors from method B (low pass) and details from method C (high pass)?

How would I do this?

[Stas Yatsenko]

Can you send us this stack (low res is ok) for tests? Maybe we will find a ways to minimize contrast distortions.

low/high filters. I think this can be done as two layers in photoshop, but I can't tell you exactly how.

It should look like result = blurred_method_b+(method_c - blurred_method_c)

[GFS]

Ahh ... OK. I thought you meant I could combine them in HF.

Regarding the difference with the C method. I was looking at them again last night and it seems to increase grain in an ugly way. This could be because of increased 1/4 tone contrast, but you'll know more about that that me.

Essentially though, I can have areas of fairly flat colour, that look the same using B as the original files, but with C they get a sort of grainy texture.

[Stas Yatsenko]

Method C behaves differently on different stacks, that is why sending your stack is the best solution.