Amazing, this sort of subject fascinates me, thank you for posting and for the the Flicker links.
I have the Laowa 25mm f/2.8, 2.5x - 5x lens on my wanted list, I tried it at UK Digital, the day after I bought my D800, I was stunned by the power of the lens. I took hand held, a photograph of the tip of a ball point pen and could see the reflections of the shop interior in the tiny 0.7mm ball!
Please can you tell me the stacking increments or the total depth of the stack in the coloured sand shots, I think you quoted 121/124 shots for the sand. I am trying to grasp the scale and dimensions of the images. I am planning to make my own stacking rail, I have many ideas but no experience of micro stacking. I am thinking hydraulics, I was going to make electronics and stepper motor rail like stackshot but it all seems very flimsy, I would rather something solid and heavy like a big hydraulic ram and a tiny metering pump of some sort to provide the motivation.
Edited for clarity
The Laowa 2.5-5x is a solid lens which is extremely good, especially for the price. I have an in depth review here:
https://macrocosmosblog.wordpress.com/2018/12/23/laowa-venus-optics-25mm-f-2-8-ultra-macro-lens-long-term-review/ Here's a list of typical increments I use:
1x: 50um (Printing-Nikkor 105mm)
2x: 30um (Rayfact 95mm)
2.5-5x: 15um - 30um (Laowa 25mm)
5x: 3um (Mit M Plan 5x HR)
10x: 3um (Mit M Plan 10x)
20x: 1um (Mit M Plan 20x)
50x: 0.5um (Mit M Plan 50x)
1-20x I use the stackshot, for 50x, I do it manually.
For the sand, it's the 2x so I use 30um. If you visit my flickr page and navigate to the description, I have the increment used listed. As for the size of the sand, it's pretty much the same as rice grains. I've attached it.
The Stackshot is not flimsy at all, it's far from flimsy actually. The only downside of the stackshot would be the 2um resolution, hefty pricetag, and fair amounts of backlash and wobble due to the lead screw. By wobble, I mean as the stack progresses, you will see the image move. The Stackshot controller is able to counter this wobble however, it's really no big deal.
How are you going to implement the hydraulics though? Are you sure a hydraulic ram is able to handle increments in microns whilst having predictable backlash, such as lead screw rails?
Moreover, that platform you're talking about seems to be vertically translated, which means your camera will be kept stationary, right? The photographical interest goes onto the platform. In this case, it's a mess.
Your lighting must go onto the platform too. As the subject travels along the Z axis, if lighting stays stationary, it's going to change the result. If you use small speedlights or constant lighting, this might not be an issue, but studio strobes, I don't think it will work.
Then is the ram able to travel back to the precise point you set it at after the stack? For stacking, I have liveview on, I adjust the rail till everything is out of focus, then adjust it till something is in focus, mark that. I will then adjust further into the subject until nothing seems to be in focus again, which is marked too. The rail then translates back to my original point, the first marker. Due to backlash, it will travel a bit further or stop too early, which stackshot compensates for. In optomechanics, we call this "Bidirectional Repeatability". High end rails score <2um, which corresponds to just two exposures for 20x stacking, no big deal, leave enough room.
Obviously, one can just calculate the depth needed and put the position A at where there's only a tiny amount of subject interest in focus, then let the rail go forward said amount. I am working on a system that allows this. Doing so will decrease the amount of shots needed for a stack, which allows an ultimately faster workflow. Wish stackshot is able to give feedback on the length travelled, which is entirely possible in my opinion.
Amazing optical quality and impeccably executed stacking, I'd say. Just a personal wish one could have a broader magnification range than just 0.5 and 2X.
Agree with Robert (seapy) that the Stackshot rail, although superficially robust, lacks sophistication and is in some respects "flimsy". The latter in particular applies to the camera platform and the awkward manner in which a camera mounts onto it. There are also play and unfortunate lateral movement in the rail during its operation.
The Laowa 25/2.8 (2.5-5X) is certainly not an APO design yet factoring in its low price and very decent optical quality, a bargain if one requires this magnification range.
Thanks! The PN95 should do alright with a macro attachment, I've yet to try though.
The movement is fundamental to all lead screw rails. The Stackshot is expensive already. Higher end rails from companies such as THK and Thorlabs are priced in the 2000+ range, and that's without a dedicated controller. An entire package where everything is sorted would cost over $5000. The good thing is that these rails can be bought on the used market for cheap prices, and adapted to the Stackshot controller. So putting everything in perspective, the Stackshot, along with the Wemacro rail are decent. The Wemacro especially has a resolution of 0.925, not as much backlash, and it's cheaper. The ideal off the shelf solution would be to mount the wemacro rail onto the stackshot controller, or DIY something based off Arduinos and surplus THK rails.
The mounting of the stackshot is indeed stupid and flimsy. They justify this by claiming the majority of their customers do not use the arca-swiss system, which they also said such customers should use it. Stackshot offers an add-on:
https://www.cognisys-inc.com/store/adapter-plate.html One must also get a clamp to go with it, so that's an extra $150 just to get a reasonable mounting solution. Another issue is the elevated centre of gravity, becomes an annoyance since my setup is vertical.
Birna, thank you for your valued comments, my opinion wasn't based on experience of using the stacking device but on my impression of it's apparently minimalistic construction.
The magnification involved with these lenses is so great we are moving into a different realm from the normal definition of rigid. This calls for mass and solid construction. I just need to understand what scale of movement is required to achieve optimum stacking increment distance. I have looked at charts but this is real world, charts are just numbers and convey very little beyond raw data.
I have a hydraulic puller which achieves ten tons of force with finger pressure by multiplying the size of the pump piston by the size of the pressing piston, also obviously the relative distances travelled, which is what I am interested in here.
Maintaining perfectly straight movement during travel is vital as I see it, accompanied by a solid camera mount and perfect repeatable alignment.
Optimal system will have:
1. No backlash
2. No lateral wobble (impossible)
3. Bidirectional repeatability of 0 (impossible)
4. High vertical loads
5. Sophisticated controller (Stackshot 3x is very comprehensive)
For minimal backlash and lateral wobble, you can get linear translation stages featuring micrometers, Newport make good ones. I've made a simple tutorial on how to convert one:
https://www.youtube.com/watch?v=UHSsmG0JaVg This is a purely manual solution though. Piezoelectric actuators can make it automatic, but I've never tried that. Those adjusters are also very expensive and fragile, which means the used one could just be dead.
This rail also should have negligible wobble, but the price and vertical load ratings aren't the best:
https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=3423 I don't think the hydraulic ram idea will work to be honest, better off finding a used stackshot 3x controller (can be had for $250-$300) and DIY a high precision THK rail to control.
