Focus Stacking Rail.

[Jim Ford]

A few months ago I mentioned that I was working on an automatic 'focus stacking' system, and showed a prototype.

Problems with image stacking

The final version is now complete(ish!) and 'ready for action'

The system operates from an external battery (currently dry-cells, but I have a lithium-ion pack I intend to use), and works as follows:

1) On power-up, the controller announces the current software version, drives the carriage to its intial position and waits for user input.
2) Using lens focussing, moving the rail, or 'inching' the rail carriage with the remote controller, the nearest element of the subject is brought into sharp focus (this is where 'live view' at 10x magnification really scores).
3) Using the remote control the lens aperture is input.
4) The user is then prompted for an actual dimension of the subject eg height or width, and this value input.
5) Next, the user inputs the same dimension above, but of the image. Again, live-view is invaluable for this, as it can be determined by meaurement on the screen.
6) The depth of the subject from front to back is next input.
7) The 'pupil ratio' is input. This is the ratio of the lens entrance and exit pupils, and determines the effective lens aperture, which in turn influences depth of field, Until I can determine otherwise, I'm currently using a factor of '1' for this with a simple lens setup. With more complicated setups eg reversed lens, extension tubes/bellows, I'll need to look more closely at this. In fact, I may do it entirely differently!
8) The controller software then calculates and displays the depth of field and number of images to take, prompting the user to accept the values.
9) On accepting the values, the controller operates the camera shutter and moves to the next position without user intervention, until the complete stack is taken. This operation takes place quite quickly. A stack of 15 images with a DOF of 0.85mm took 20 seconds to complete.
10) Having completed the stack, the controller then drives the carriage back to it's initial starting position.

Description and images to follow.

[Jim Ford]

Stacking Rail 1



1) Base - 8mm 'Tufnol'.
Tufnol is an engineering thermoplasic composite, chosen for its dimensional stability and ease of working with hand tools (although it tends to blunt them!)
2) Stepper Motor - 200 steps per revolution. Salvaged from an Epson Printer.
3) Limit Switch. To reset microcontroller if extreme of travel reached.
4) Bearer - one of two. 10mm Acetal - chosen for it's dimensional stability and good sliding qualities (though difficult to file owing to it's slipperiness!).
5) End Plate. 1/4 inch Tufnol (It was to hand!)
6) Lead Screw Nut. 2 thicknesses of 10mm Acetal, allowing for 'shimming' to take up any wear in use.
7) Guide Rail. 10mm A2 stainless steel rod.
8) 9 Pin 'D' Socket. For connection to controller.
9) Carriage. With 'QR' camera mounting.
10) Lead Screw. 10mm 1.5mm pitch A2 stainless steel studding, with 'rolled thread' giving a good finish, which was further improved with metal polish. With this pitch lead screw, a pulley ratio of 1:1 and a 200 steps per revolution stepper motor, this gives a smallest carriage step size of 0.0075mm.

More to follow.

[Jim Ford]

Stacking Rail 2




1) Microswitch (behind) . For initialising carriage position
2) Bearing (behind cap). 10mm O.D. ball race - same other end of lead screw.
3) Camera Socket. For shutter release cable.

More to follow

[Jim Ford]

Stacking Rail 3



1) Microswitch. For initialising carriage position
2) Driven Pulley - 25 teeth. A pulley system was chosen rather than direct drive initially on the prototype, for experimentation of motors and ratios. The system was carried over to this version. It added considerably to the overall cost of the project. There are 'fors and againsts' over a direct drive system.
3) Drive Belt.
4) Drive Pulley - 25 teeth.
5) Tripod Mounting Threaded Holes. 1/4 and 3/8.

More to follow

[Jim Ford]

Stacking Rail 4



1) IR Controller Sensor. For input of data with remote controller.
2) 'Hard' reset Push-Button.
3) Programming Port. For internal 'Arduino' microcontroller.
4) 9 Pin 'D' Socket. For cable connection to rail.
5) LCD.

More to follow.

[Jim Ford]

Stacking Rail 5



1) Power Socket: 6-9V.
2) On-Off Switch.

More to follow

[Jim Ford]

Stacking Rail 6



Showing whole system and waiting for input of data.

More to follow

[Jim Ford]

Test image, showing sharp cut-off of focus sharpness either side of stack. This is the whole point of stacking - to have the interesting elements of a subject all in focus, but render those outside 'soft'.

Stack data in image.



Software used to combine the images: 'CombineZM'.

Jim

[nightshade]

Very nicely executed bit of kit, congratulations!

[Lancashire Lad]

Nice one Jim!

I can see that this bit of kit might have great potential for "home studio" work.

I'm still not convinced though, that such a device would stand up to the rigours of "on-site" use (weatherproofing / dirt & grit ingress, etc. etc.).

Especially when I consider some of the wet & muddy situations I find myself (and more to the point, my camera kit), in when doing focus stacks of fungi & myxomycetes.

Very much lookong forward to seeing some more "interesting" images.

Regards,
Mike.