Easily transportable self-contained stacking setup

[rjlittlefield]

This is a contribution to the list of stacking systems that have been posted out recently. I put it together a couple of weeks ago to take to a small regional meeting of the Lepidopterists Society. Back in July, I gave a presentation on focus stacking at the national meeting, but that was all PowerPoint and stereo prints. For the regional meeting, I decided to take a working rig and shoot some stacks onsite.



One of my criteria in assembling the system was that it be reproducible and fairly easy to assemble and use. Except for the gray brackets, everything you see here is either currently available as off-the-shelf equipment, or it's pre-production prototype that soon will be.

I took enough stuff to cover seamlessly from infinity focus to 10X on sensor. It all fit comfortably in a backpack and a laptop bag.

Listing the parts that you see here:

• Focus stepping: StackShot rail with a USB interface.
• Camera: Canon T1i
• Lens: Canon MP-E 65mm
• Illumination: Canon Speedlite 580EX II with Optek diffuser, mounted on Manfrotto 819-1 hydrostatic arm; Petzl LED headlamp for focus and framing.
• Specimen positioning: Adorama 2-axis rail
• Computer & software: laptop running Zerene Stacker and Canon EOS Utility
• Stereo viewers: Wheatstone Mini-Scopes

Also contained in other bags were a 105 mm macro lens, and a Nikon CFI 10X NA 0.25 infinity objective plus telephoto to match.

There are, of course, a couple of aspects of this system that you haven't seen before. You can see them on the laptop screen and sitting next to it: Zerene Stacker controlling the StackShot rail while simultaneously stacking.

As one of the developers, of course I'm a biased observer. On the other hand, as an active focus stacker, I can truthfully say this is pretty cool.

There's a small control panel that basically mimics the coarse and fine focus knobs on a microscope. Turn the knobs by running the mouse, and the camera moves up and down while you watch the LiveView display in the EOS Utility window. Click a couple of buttons to set the start and end of the stack, one more to start shooting, then stand back and watch while the system automatically shoots the stack and simultaneously generates a PMax result.

Before anybody asks, let me volunteer that, no, I can't quote an exact date when the pre-production stuff will be released.

Hope you find this interesting...

--Rik

Edit: correct spelling error in lens name

[Craig Gerard]

I like it

For current StackShot owners here is a handy link:
support@cognisys-inc.com

For current Zerene Stacker license holders here is a handy link:
support@zerenesystems.com

Before anybody asks, let me volunteer that, no, I can't quote an exact date when the pre-production stuff will be released.

No doubt, many will have specific questions in the meantime



Craig

[Bob^3]

So you gone went and done it---three click push button stacking! I figured when I read you had the StackShot it was just a matter of time before you added code to ZS to control the thing.

And I won't ask your forbidden question, or the other 50 that come instantly to mind yet.

Very cool stuff, Rik!

[Peter M. Macdonald]

All you needt to add will be RAW support ...

No, more seriously, this looks very promising and intersting.

What are you using to support the Adorama rail? Sort of looks metalic, but I am sure that I can see some woodgrain. I like wood. I can cut that! Metal is a whole different ballgame.

Best,

Peter

[morfa]

The Stackshot suddenly jumped up a good couple of notches on my wish list – exciting stuff Rik!

In the 40x stack posted here you say "170 frames at 0.001 mm avg". Could you tell us a little more about the precision of the virtual fine focus knob in combination with the Stackshot hardware? Do you have any more (real world) data on the frequency and magnitude of deviations from this step size?

How about using using the Stackshot controller with a different stepper motor + microscope focusing block – would this work with Zerene as well?

[AndrewC]

Nice integration - so will the ZS "ingester" at least be released soon ?

[AndrewC]

Not having Canon gear (and not at all wanting to detract from this lovely set up), a question - does the MPE-65 have internal focussing ? If so, how would it compare to automate a stack by using external control of the IF rather than moving with locked focus ? Strikes me that that would yield a much more portable set up and eliminate some hardware ? Run it all out of a small netbook and you've got a great field rig - I've just spent another afternoon shooting fungi

[Bob^3]

If so, how would it compare to automate a stack by using external control of the IF rather than moving with locked focus ?

That exact functionality has been available for some time in a number of existing software packages (at least 6 I'm aware of) using the Canon and Nikon SDKs. Three that come to mind are Helicon Remote (comes with Helicon Focus), DSLR Remote Pro (Breeze Systems) and a new one for Nikon only called ControlMyNikon for $10 (used to be called TetherScript- thread HERE).

I also don't own the MP-E 65. So I can't say how well these programs function with it. But I have used them with a variety of Nikon bodies and lenses. They do work well and can make a portable package for field work to capture stacks for close-ups and landscapes, but with limitations:

Only work with Live View capable cameras where the manufacturer has offered a SDK.

Many don't function well on limited hardware like Netbooks.

Depending on the lens, the focus step size can be limited.

But perhaps the biggest limitation is the requirment for an autofocus lens, so you can't use all those super quality MF lenses and microscope objectives.

Personally, while this capability would be a nice addition to ZS, I'm much more interested in the ability to control stepper motor driven stages as is shown here. I am not aware of any program that can do this to date. I know Helicon has been working on it for some time---but I haven't seen any results. So it looks like ZS will be the first to implement this.

[AndrewC]

..a new one for Nikon only called ControlMyNikon for $10 (used to be called TetherScript- thread ...

Trying not to go off on a tangent here, but I see ControlMyNikon still hasn't solved the issue with M-Up mode.

Rik: does the software platform you're developing cope with M-Up on Nikons or haven't you be able to test that ? Everything else I've come across, and tried to code myself, using SDKs and the basic Windows media interfaces seems to choke when presented with M-Up. Even Nikon's own software last time I tried it.

Bob: my point was that while this is undoubtedly impressive and a big step forward for an easy to setup desktop system, it would be really nice if it ported to a minimalist system you could use in the field. It is pretty straightforward to pull a Liveview feed from certain cameras, take a single shot, upload an image through USB and also drive a stepper motor. What is really nice IMHO is what Rik has done here to pull together the various SDKs and tie them together with what I call an "ingester" feed to ZS.

[Peter M. Macdonald]

Andrew,

There is no focus ring on the MP-E 65! There is no autofocus either. The only onboard adjustment is the zoom mechanism. This makes a very large adjustment to the length of the lens as it operates. At 5x the lens is more than twice the length that it is at 1x.

The only way to focus this lens is to move it and the camera relative to the subject - just like Rik is doing.

Peter

[Bob^3]

I see ControlMyNikon still hasn't solved the issue with M-Up mode.

I have read that the control of the Nikon M-up sequence is baked into the low-level firmware and is not available to the SDK or to outside control. I keep hoping for a firmware update that will address this poor implementation. But if it is hardware limited, it might only be fixed with a new body design.

Bob: my point was that while this is undoubtedly impressive and a big step forward for an easy to setup desktop system, it would be really nice if it ported to a minimalist system you could use in the field.

I would guess the key question here is what is the minimum hardware requirement to run ZS---will it run on a Netbook? Also, the stage and controller could be made smaller and lighter for more portable field use.

The Stackshot suddenly jumped up a good couple of notches on my wish list – exciting stuff Rik!

Me too, John, but not so much for the stage as for the controller unit. I already own several very high precision ball-screw stages that are smaller (lower profile) and lighter than the StackShot rail, purchased at low cost on eBay. What is harder to find is a small battery powered stepper controller to run the stage. I have a couple Parker Compumotor controller/drivers that work well, but they lack the useful LCD display on the StackShot controller. The last time I looked, Cognisys did not list the controller as a separate purchase (but I haven't tried to contact them on this).

Rik, I'm wondering if the ZS control functions could also send commands to other controllers via usb, RS-232 or network ports? This would make the interface much more generic and cater to us die-hard hackers who like to roll our own custom setups for specific applications.

[rjlittlefield]

Thanks for the interest, comments, and questions!

A few quick answers...

Peter, all the gray stuff is painted wood. It's by far my most common construction material. Usually I don't bother to paint it, but I figured it should look good for a public showing.

Morfa, there have been some improvements to the firmware that allow Zerene Stacker to control the rail much more accurately than the standalone controller will. Using ZS and the prototype firmware, I have direct measurements of the unloaded rail doing sustained movement at less than 1.3 microns per largest microstep. This compares to 0.5 microns per average microstep. The remaining deviations from ideal behavior are repeatable and periodic, but it's not clear whether they can be further reduced. The stack you ask about that was shot at average 1 micron steps actually varied between 0.6 and 1.9 microns, as determined from the earlier direct measurements. I was pleased to see that there were no focus bands despite the high magnification and use of just the rail, no microscope focus block. If a focus block is used with a separate stepper, then yes, the improvements apply to that use as well. I have not taken measurements in that case, but from just watching the knob the maximum movement is much smaller than in my previous report.

Andrew, the current plan is that all the capabilities described here will be released at one time as an extension to Professional Edition only. I will contact you offline about getting a beta version. Bob's comments about the MP-E 65 and other issues about internal focus are right on target so I won't repeat those. I will add that while using the SDKs to drive a camera directly is "on the list", it's not near the top. What ZS is doing right now does not involve the camera SDKs. ZS talks to the StackShot controller directly, relies on the StackShot controller to trigger the camera, and does on-the-fly ingest of images by watching for them to appear in some specified directory. In the system as pictured, that's a Canon EOS Utility window showing at upper right, and it's EOS Utility that downloads images from the camera as they are shot. Exactly how that strategy works with a Nikon and different software remains to be seen because my D5000 hasn't arrived yet. Regarding mirror-up, at this point ZS relies on the StackShot to handle that as correctly as possible. It works by mimicking a remote shutter release, so I'm presuming there's no problem.

Bob, the current implementation talks only to StackShot, but the design is fairly modular. About the controller, I'm pretty sure you could get that separately even though it's not currently listed that way.

--Rik

[elf]

Is the ingestion of images tied to StackShot? I really hope not. While the StackShot works well as fixed step mover, it can't do focus steps by changing the bellows draw.

[Bob^3]

Is the ingestion of images tied to StackShot?

Elf, I don't think that is what Rik is saying. It appears that the (free?) Canon EOS utility is handling the transfer of images from the camera to a specified folder on the computer. ZS simply monitors that folder for new files to be added to the stack for processing.

One issue I can see with Nikon cameras is that I don't know if there is a free utility which can perform the transfer function automatically (at least, not from Nikon). I think Camera Control Pro can, but it's certainly not free (or inexpensive). I happen to own CCP, but this could be an issue for many members.

Edit: Nikon ViewNX (free) can download images from the camera through the USB port, but not sure it can do this automatically. So ZS would need to request the transfer through ViewNX, if that is possible? Perhaps, that's the way ZS is working with the EOS utility?

While the StackShot works well as fixed step mover, it can't do focus steps by changing the bellows draw.

I think it could if a stepper motor were coupled to the front standard to move the lens, which if I remember correctly is linear with focus depth? But if the rear standard (with camera attached) was moved by the stepper, a non-linear step series might be best? I wouldn't think this would be too difficult to implement in the StackShot firmware, for new builds---but might be problematic for existing owners who might need this feature.

[AndrewC]

.. I will add that while using the SDKs to drive a camera directly is "on the list", it's not near the top....
--Rik

There aren't that many attractions for communicating through the SDK unless perhaps you want to drive aperture, shutter speed or lens focus remotely. It does give you the option of one less wire - if you have a USB connection for image transfer you could also fire the shutter through that.