Tricky close-up focusing problem for the gurus...

[Ian Leake]

I'm hoping that one or more of our LF gurus can help me work out a tricky focusing problem I'm struggling with. It's a bit complex to explain but here goes.

I want to shoot multiple photos of a subject that can be used to build a composite image. Kind of like some of the detail photos in my gallery, but overlapping more. Say, for example, that I have an arm and a leg in view, and the arm is some way in front of the leg. If I took two photos, each showing bits of the arm and bits of the leg, then obviously it's essential that both the arm and the leg (near object and far object) have the same magnification in each photo. Otherwise when the two photos are adjacent to each other they'll look wierd.

Now my problem is that while maintaining the same magnification I also want to change the point of focus. So in one photo I may focus on the leg, then in the other I may focus on the arm. In the attempts I've had to date I've shot at about 60% life size on a 8x10 with a 476mm lens (so bellows are about 760mm). It's not practical to just stop down and hope for the best so I need to refocus somehow. In one attempt last week everything seemed to just work, but in my second attempt today nothing worked. And unfortunately I can't remember how I did it the first time.

Does anyone know how I could do this? Assuming what I want to do is actually optically possible, I guess the options are focusing with the front standard, the back standard, moving the camera, or more likely a combination of these.

Many thanks in advance for any help you can offer.

Ian.

 

[David A. Goldfarb]

If you want the magnification to remain constant, then the bellows extension needs to remain constant, and you need to move the whole camera to compose and focus. You could do it roughly with the camera on a stand or tripod with a dolly, then move the camera on the tripod block (assuming you are using a monorail) for finer focus, and then adjust the rear standard slightly for the final adjustment if necessary.

If all the photos are going to be taken in the same shoot, then I'd just set the focus and move the camera. If you need to use the camera for other things before finishing the whole project you could mark the front and rear standard positions on the rail with a grease pencil or measure the distance between them and make a note of it.

 

[Ian Leake]

If you want the magnification to remain constant, then the bellows extension needs to remain constant, and you need to move the whole camera to compose and focus.

Thanks David, this is what I've done for some time, and it works fine when there's quite a gap between the photos. But when the photos actually overlap something else seems to happen.

If, for example, I focus on the far object and then move the camera backwards to focus on the near object, the magnification of the far object (which is now out of focus) changes as does the magnification of the near object (which is now in focus).

The magnification at the point of focus hasn't changed because the bellows are the same (and the distance between the subject and the lens rear nodal point is the same too), but when the two images are laid next to each other nothing aligns properly.

 

[Nick Zentena]

The bellows disc could help with this.

http://www.salzgeber.at/disc/disc.pdf

Basically pick a magnification and make sure you get the same result each time.

I guess you could just put a constant object in the scene. Measure it. Then remove it.

 

[Nick Zentena]

Wait you want to distant objects in the same photo to have the same magnification?

 

[Ian Leake]

Wait you want to distant objects in the same photo to have the same magnification?

Yes, between two photos. So if the near object was at 60% and the far object was at 55% in the first photo, then both stay at that magnification in the second.

 

[David A. Goldfarb]

So, are you trying to do a sort of multi-perspectival Hockneyesque bodyscape? If that's the case, then maybe the alignment shouldn't be perfect.

Or are you trying to do something more like a big image spliced together from multiple negatives shot in an evenly spaced grid?

Or is it something else?

 

[David A. Goldfarb]

Yes, between two photos. So if the near object was at 60% and the far object was at 55% in the first photo, then both stay at that magnification in the second.

Hmmm... I think in that case, you would want to keep the camera in one place and refocus using the rear standard, but mark the groundglass or measure the size of the objects on the groundglass to be sure.

 

[Sirius Glass]

I am not sure that I understand to complete situation.

• Is the camera stationary for all photos? [doesn't sound like it]
• Is the camera rotating a round a stationary point? [ again doesn't sound like it]
• Is the camera moving laterally?
  • If so, can you set the models up to be close to equidistant from the lateral line that the camera will move along?

I am thinking that if I know more about how you are shooting, I might be able to help.

Steve

 

[Ian Leake]

I am not sure that I understand to complete situation.

• Is the camera stationary for all photos? [doesn't sound like it]
• Is the camera rotating a round a stationary point? [ again doesn't sound like it]
• Is the camera moving laterally?
  • If so, can you set the models up to be close to equidistant from the lateral line that the camera will move along?

I am thinking that if I know more about how you are shooting, I might be able to help.

Steve

I've tried a number of different ways, but not systematically. I'm quite happy for the camera to move horizontally or vertically (which is what I usually do with the 4x5). But in the two attempts with the 8x10 I mentioned earlier, I tried to keep the tripod in the same place and pivot the camera on the head.

Hmmm... I think in that case, you would want to keep the camera in one place and refocus using the rear standard, but mark the groundglass or measure the size of the objects on the groundglass to be sure.

Intuitively this seems to be the right thing to do, but it didn't work today Of course today's difficulty may not have been due to optics at all. We tend to shoot each photograph several times to get everything right, and that means reposing repeatedly. Now although I know she hardly moved today it will certainly have complicated things.

What I should probably do is set up a test with static objects, get my tape measure out, and shoot some polaroids.

 

[Helen B]

Here's one interpretation.

You have a picture with an object at 60% magnification in perfect focus. Another object is out of focus, with a magnification of 55%. 476 mm lens.

The first shot:
The object at 60% is 1269 mm from the front node, the rear node is 761 mm from the film plane. The 55% object is 1384 mm from the front node.

To keep the same perspective, the lens must be kept in the same place (specifically the entrance pupil, if lens swing is involved) for the second shot.

The second shot:
The rear standard is brought forward to focus the far object, which is 1384 mm from the front node. The rear node is, therefore, 725 mm from the film plane. The magnification of this object is now 52.4%. The close object is 1269 mm from the front node, and the magnification is 57.1%.

The magnifications in the second shot are different from those in the first shot, but they are in the correct proportion. If the second shot is enlarged 1.05 times the enlargement of the first shot, the images should match as much as two images with different areas in focus can.

How does that sound?

Best,
Helen

 

[Ian Leake]

Here's one interpretation.

You have a picture with an object at 60% magnification in perfect focus. Another object is out of focus, with a magnification of 55%. 476 mm lens.

The first shot:
The object at 60% is 1269 mm from the front node, the rear node is 761 mm from the film plane. The 55% object is 1384 mm from the front node.

To keep the same perspective, the lens must be kept in the same place (specifically the entrance pupil, if lens swing is involved) for the second shot.

The second shot:
The rear standard is brought forward to focus the far object, which is 1384 mm from the front node. The rear node is, therefore, 725 mm from the film plane. The magnification of this object is now 52.4%. The close object is 1269 mm from the front node, and the magnification is 57.1%.

The magnifications in the second shot are different from those in the first shot, but they are in the correct proportion. If the second shot is enlarged 1.05 times the enlargement of the first shot, the images should match as much as two images with different areas in focus can.

How does that sound?

Best,
Helen

Thanks Helen, I forgot to say that I contact print so enlargement is out, but you've got me there anyway [with a bit more help from AA's The Camera ].

What's now clear is that I have to sacrifice either magnification or perspective.

Using the same numbers that you used Helen, if I sacrifice perspective by moving the camera (and changing the bellows) but keeping the same magnification (and now focusing on the far object) then I end up with a lens to far subject distance of 1,341 mm. It can be calculated that the two subjects are 116 mm apart, so the camera has to move forward by 44 mm. The magnification of the front object (now out of focus) is still 60%.

My first conclusion is that as I can't enlarge, I'll have to accept the change in perspective. Obviously that will work better for some compositions than for others. My second conclusion is that I need to use the longest lens I can in order to minimise the effect of changed perspective.

My next task is to create a printed ready reckoner so I can work this out quickly for different setups rather than having to resort to paper, pencil and mathematics...

Thanks again for your help everyone.

Ian.

 

[nicolai]

How about usin one object as a size reference, like a film holder? Have the model hold one, use photo tape to mark its size on the ground glass, and you can match it on subsequent shots? (Or put frosted scotch tape on the GG and mark it with a pencil, whatever's the least intrusive.

I'm not sure I'm fully awake yet, but it *seems* like it would work...