Focus Stacking

Yes, there really are two ways to take the images for the stack. You can move the entire camera assembly (not changing the focus), which moves the zone of focus in the same direction as the rail moves. Or you can move the center of the focus zone without moving the camera.


Let's take your chess board with 3 pieces, it's what... 8" square. You have a piece on the A row, the D row and the H row. You want all three pieces PLUS the board to be in focus. So from zero to 8". And your camera can only put about 3" (3 rows) in sharp focus.

Shot 1 center of the focus band is the edge between A and B. So the sharp area runs from in front of the board out to maybe half of C. This means the piece on A and rows A and B are in focus, plus some of C.

Without changing the camera, you move the assembly on the rail so that Shot 2's center of the focus band is maybe 1/4 of the way into D. That means that edge between B/C and out to 3/4 of E are in focus.

Shot 3, middle F center, so front of D to back of G is sharp (do you begin to see the small overlaps that are critical to the process?)

Shot 4, between G/H is sharp, so some of F to past the edge of the board.

If you now take the in-focus parts of each of the 4 shots and blend it into one image - that's focus stacking.

You can also see why more shots aren't necessarily better. If you move the assemble 1" at a time instead of 2.5" (or so), you get multiple shots where the focus for specific areas is good. So somehow the software has to pick just one.


Now you can also automate this the other way as you said. The software actually handles the small differences in image size automatically. But it's not as simple so the results may not be as good. This what you are alluding to by "touching the piece". That basically tells the camera to center the sharp focus zone at that point. But since the camera is re-focusing, the lens moves and the area of the image is slightly different. It may only be a pixel or two larger/smaller, but it also means the edges (lines of the chess board) move and the software has to keep those aligned.


When I use my StackShot (focus rail) it automatically does the first - since I autofocus on the coin (which does an overall or average focus) and then the rail moves the pre-focused camera assembly from the point where the highest detail is in the sharp zone to the point where the lowest area (field) is in the sharp zone.


Helicon lets you cheat by highlighting what will actually be in focus with a blue overlay (I assume Zerene does something like this too). So you

* Auto focus
* Move the rail out until the last bits of blue fade and set one point
* Move the rail in until the last bits of blue fade and set the 2nd point

based on the distance between the points, select either a step size or # of shots (the software calculates the other) so I don't get an INSANE # of shots.



This one took a large # of shots for the reverse, because the focus had to move from the tiny bit of rim left between 7 and 9 o'clock and the flat surface revealed after the split.



But that's a special case.

Thanks. Of course it's legal to post links to other sides. Especially legit knowledge. Now if you want to post links to how to doctor coins and fool TPGs, the mods might take issue.

http://zerenesystems.com/cms/stacke...ngversusrail

You are right, that's a very interesting piece. Of the items they list, a coin is most similar to which item? That's probably the methods to use.

Helicon has a video on remote, here http://www.heliconsoft.com/helicons...con-remote/, but they don't show the option (you can see it on the right side just above the bottom edge of the frame) "Highlight focused area"... I'll try to post a screenshot next time I do some shooting...

So you are moving in say the X direction (where X Y are the two axis of the paper and Z the in/out one?

I haven't tried anything but Z, but you should be able to move in any straight line. I guess even something that was a diagonal. As long as there is enough commonality in the image for the software to stitch it together (like a panorama shot?).

watching these youtube videos. it is so amazing to see what a person can do with a camera a lens. the right gear. these images astound me. I realize I am only scratching the surface. stacking takes this so much farther. I am going to learn this. wow that is coin shooting at its best.

Rocky : Not sure how far you have got......I have the stack shot system.... It may seem expensive but its the most effective.....However when I purchased my stackshot it was far to early for my skill level at that time. What I believe I would have benefited from is buying an old microscope and cutting it down. The fine focus control of the microscope allows you to adjust the focus by very small steps. Had I started that way I would have gained a much better feel for stacking before I jumped to the stackshot.

I use a cut down microscope with an XY stage as the coin stage as well as the stack shot. The cut down microscope, as a Z stage, gives me much finer manual Focus control than the moving the camera on the copy stand column. Adding in an XY stage to it really helped with composition... What would cap it off would be a rotary element that way I would minimise handling of the coin.

Here is an example of what I purchased ( around the same price to)

http://www.ebay.com/itm/Olympus-CH-...AOSw5IJWdWUT


I just cut off the microscope head off my example. The example above is Ok but if you can find one with clear graduations on the focus knobs you will gain the ability to move the coin in Micrometer increments needed for stacking.

I have a Zeiss rotary (pol) stage, and it makes straightening the coin much easier than doing it in Photoshop afterwards. It does X-Y movement, too, and it's attached using a Zeiss WL focusing block, for stacking and fine focus. The disadvantage of the Zeiss WL focus block is that it's limited to 2mm fine-focus vertical travel. So far, this is enough for coins, though.