MicrobeHunter.com Microscopy Forum

Hi Folks,
I have a modest Trinocular OMAX with 6 M camera and ToupView. I have done some good stacks but wish for more accuracy in the focus process. Unfortunately, my microscope has no hash marks or scale on the focus knobs. Are round, calibrated stick on labels available somewhere? Will they even help?
Beyond that, I'm handy with electronics if I can find or build a highly precise motorized focus drive? If I can use it on my telescopes all the better. DIY or used would be great.
So, how do you handle incremental focus? What's more, does taking more layers sort of make super accurate focus stepping less important?
I sure appreciate this forum and thank you for reading my post!
Take good care,
Jimmy

JKelley -- you could fit a larger fine focus knob over the existing one and print out and glue on a series of graduations on the periphery. If you know someone with a lathe this would be a quick project. Swap the time for some bit of electronic wizardly they might want?

Might even fit the new and larger cylindrical over-knob with a couple of nylon tipped setscrews so you can take easily remove it if needed.

Depending on how your fine focus knob goes on, might also be easy enough to simply replace it with a larger one.

Hardest part is figuring things out (line spacing) so that the graduations wrap to leave the last graduation about right -- or just not worry about it if a single turn gets you enough depth of field. You can also print out the graduations and then machine the diameter (circumference) of your new knob to match.

Another way to print up graduations (besides a printer) is to get something like a Brother P Touch label maker. Print a vertical line a zillion times . . . You can buy all sorts of label stock and it sticks and holds up pretty well.

Engraving or stamping the lines is also pretty easy for someone with a metal shop.

3D printer is another possibility if you know someone with one. Turning from solid aluminum or an engineered plastic would likely be faster and easier and provide a better looking result.

Given that your focus mechanism will have some backlash, be sure to take that out when you start a stack and keep incrementing in the same direction until you're done.

Another possibility is to fit something like a red dot laser sight (they can be very small) to the end of the fine focus. Even with double stick tape, hot glue, or a drilled and tapped hole. Play that laser dot on a distant surface and a couple microns of fine focus movement might correspond to an inch or more on paper yardstick affixed to a distant wall and/or ceiling. For perfect results, it ought to be an arc -- but get far enough way and it won't matter much.

Others have drilled holes in the fine focus know and fixed spokes.

Stepper motor is another possibility, but with the same challenge of fitting the motor to the fine focus as with fitting a larger knob. Ideally you'd automatically trigger the camera after each step had settled, being mindful to eliminate sources of vibration in the stepper and camera.

This subject has been recently discussed:
viewtopic.php?f=12&t=6940&hilit=stepper
But, I think that the question : "what is the essential and sufficient number of single images to yield a satisfactory stack" has yet to be answered.
In addition, what may be important is that each image in the set is in focus (at least a part of the area), and not necessarily that they are evenly spaced. On the other hand, it is far easier to simply run an automated sequence of vertical stage movements and shooting at each stop.
I am certain that in the research literature it has been addressed, so all we need to do is locate the article and "translate" it to a practical working rule for hobbyists that stack for pleasure rather than business...

Hobbyst46 wrote: I am certain that in the research literature it has been addressed, so all we need to do is locate the article and "translate" it to a practical working rule for hobbyists that stack for pleasure rather than business...

My question to Zerene support is largely irrelevant, but Rik Littlfield's answer was concise:

On 14 Jul 2018, at 01:45, support@zerenesystems.com wrote:

Michael,

I do not anticipate any special problems from either the unequal spacing or the small number of frames. Unless you're doing synthetic stereo (probably not practical with only 3 images), Zerene Stacker does not make any assumptions about equal spacing. And while 3 frames is definitely on the short end of things, I have seen even 2-image stacks processed to good effect.

... Straight from the horse's mouth !

MichaelG.

Hi, I never use equally-spaced focus for my stack layers. I concentrate on focusing to features that I notice coming into focus as I move the focus knob - regardless of their distance from the focal-plane before them. I also tend very much to keep my number of frames (layers) to a minimum, definitely finding that 'less is more'.

That said the context of my above remark is one of semi-transparent/translucent specimens sectioned very thinly as stained permanent slides. This is far different to a completely opaque and incident-lit specimen such as say a beetle etc. In this case of complete opacity and incident-lighting I would say that more layers and regular spacing were likely to give better perception of depth.

With stacking of translucent subjects the aim (mine anyway) is revelation of details that don't coincide in the X,Y plane. For details that do coincide in the X,Y plane but not in the Z plane the technique is one of 'optical sectioning', where the different levels of focus are not stacked but taken as separate images.

Opaque subjects of course don't have details coincident in the X,Y plane.

John B.

Note that anything not quite in focus will probably give you false colours from longitudinal chromatic aberrations. For Nikon and many others that would be purples. If there's something there, you probably want it in focus.
My old computer used to take one minute per frame to stack. The current one does one a second, so I don't worry about having too many. I do trim the ends, though.
(Using Zerene Stacker)

Thanks so much everyone. Nothing beats experienced users like you.
I appreciate your posts very much,
Jimmy

Coincidentally I'm just having bother with an Olympus focus knob which IS calibrated, but only by means of raised numbers and bars, so it's black on black.
I tried painting it, but it's not pretty.
So I'm printing a circle with 100, or 200 marks round it, sized to stick on one or other knob.
If you have photoshop in one version or another, you can Transform > rotate > by 3.6°
then when you hold the right combination of Ctrl-ALt-Shift T (may be one key out, I can't remember) it repeats the transformation, so you tap 100 times. Or spend time finding a more efficient way!

Here's one I stole - Ed French won't mind:

If you don't need the printed numbers ... This is very useful:
http://www.cgtk.co.uk/metalwork/reference/divider

MichaelG.

On the microscope, I make focus steps by eye judgement. I watch where the focus is now, where the next details will likely appear, and turn the wheel until I see them. Of course, I scout the subject before starting, so I know where to look. This works quite well in all magnifications (4x-100x). Stepsizes vary with magnification and aperture.

This said, I will be making an automated system based on a stepper motor. Note that normally these motors make 200 or 400 steps per full turn, which is not always enough. Microstepping, gears, or a belt drive can be used to make smaller steps. I will try a geared system. Gears don't have the best of reputations for the extra play that comes into the system. But I think (or hope...) that that won't be a problem as long as the moter is used one way.

Note that the hardware is only part of the job: you'll also have to write software for the controller. If you don't feel easy with that, you might consider using e.g. the stackshot controller to drive your motor.

iconoclastica wrote:This said, I will be making an automated system based on a stepper motor. Note that normally these motors make 200 or 400 steps per full turn, which is not always enough. Microstepping, gears, or a belt drive can be used to make smaller steps. I will try a geared system. Gears don't have the best of reputations for the extra play that comes into the system. But I think (or hope...) that that won't be a problem as long as the moter is used one way.
Note that the hardware is only part of the job: you'll also have to write software for the controller. If you don't feel easy

I would start with 200/400 steps/turn, on my microscope that means 2 or 1 micrometers/step, respectively.
Software is (IMO) easier done that hardware, not many commands are needed, but the initial position of the moving part ("zero point") must be defined, so a "reset" will always bring the stage back to the starting point.
And, of course, provided that the communication protocol with the camera is available... actually of this I doubt...

I find 1-2µm quite big steps, but of course all depend on your intended magnification. I quite agree on your stand on software, but i have been a software engineer all my life. Not everyone feels the same about it.
Triggering the camera is easy: with an optocoupler yout short the pc remote cable for about 0.1 sec. That's all there is to it. Can be done with a transistor too, but I prefer to have both circuits completely separated.

I have also pondered using a motorized solution that I wired up somehow... I am just afraid I might beef it and it makes my focus knob fly off or something

I have a little clicker made out of a bit of an aluminium soft-drink can (the stuff is not as sharp as you might at first think.)
I cut mine into an L-shape ( a bit like a piece of angle iron) with a pair of scissors and have one flat side of it attached flat to the microscope body and the end of other flat face pressed again the focussing knob.

This thing makes a surprising loud click as each moulded line on the knob's surface flicks it.

You do need to turn the knob slowly otherwise you will end up doing a 'click and a half' rather than a click.
Not very sophisticated .. I push it away from the focussing knob with a finger when racking the microscope down.

On my vertical set up I use a motorized rail. On the microscope for now I do not even look at the tick marks, just try make small steps; I plan to motorize it soon...

Chas wrote: ↑

Mon Mar 29, 2021 9:33 pm

You do need to turn the knob slowly otherwise you will end up doing a 'click and a half' rather than a click.
Not very sophisticated .. I push it away from the focussing knob with a finger when racking the microscope down.
clicker 640 480.jpg

Best focussing knob mod ever..

set to take a pic every second and turn focus slooooowly

That rules chas

I started focus stacking over 15 years ago doing terrestrial photos with a DSLR and adjusting increments manually. I used Combine ZM which was about all there was then. Now I use a simular approach with a DSLR on my microscope. For each image I do a couple of “practice runs” to find the two endpoints. I watch the image on a large LCD monitor. I always go from the counter-clockwise endpoint to the clockwise endpoint. That way I don’t get confused as to which way I’m turning the fine focus knob. The farther the distance between endpoints the more steps I use, it’s strictly seat of the pants. I don’t think it matters that each increment is equal to all the others in terms of microns or degrees of knob rotation. If Zerene is putting out something goofy look thru all the images and find the one that’s not right and delete it. Then run Zerene again.

Digital imaging people say that all digital images require some amount of sharpening, it’s the nature of the beast. I’m sure most of us do that. Recently artificial intelligence algorithms have greatly increased the capability of sharpening software. This seems especially true with focus stacked images. I’m not plugging anyone but if you want to try increasing the quality of your stacked images you might investigate a trial version of something like Topaz Sharpen AI. I’m sure there are others as well.

I was trying to do a stack with an immersion lens on a tilted old microscope and came up with this:
It is a disk magnet joined to the fine focus knob and with a junior hacksaw blade as the lever... the magnet-hacksaw blade connection acts as a bit of a slip-clutch, so theoretically one can hold the fine focus knob on the other side, push the hacksaw blade forward, and so keep going: Pulling the lever backwards with a finger with the side of your hand resting on the table, does work really well, but gives not much more than one quarter of a turn of the fine focus.
Unfortunately It was hard to hold the focus in the same place when sliding the hacksaw blade forward again.


..so I thought about using a stronger 'clicker' as a ratchet mechanism attached to the focus knob on other side .. it didnt quite work as a ratchet but..
I found that using a cylindrical magnet and a washer to trap a bit of (random ) plastic allows a good load to be put on the knurling of the fine focus knob and get a really nice tick.
The magnet is attached to the stand with a bit of thick double-sided peel-off/ reuseable tape ... it holds itself directly onto stands made of iron.
Here is a video the ticker bit on a Watson bactil
Edit: oops that video hosting site was nasty ...I will find another.
[In short: the ticker ticks roughly 3 times for each engraved division ]
As far as attaching the magnet to the fine focus knob, goes, I found something random that push-fitted onto the knob and hot-melted the disk magnet onto that.

Here is a stack using the ticker on the Cooke Troughton Simms: -Not a slide that I made, I hasten to add

I’m with mrsonchus. I never use preset focus lines. You can’t count on that distance actually focusing on anything but protoplasm.

I recently experimented with putting a stepper motor onto an old AO35 microscope with non-plan objectives and got very promising results.

I talk about it briefly in this post and a couple below it.
viewtopic.php?p=117716#p117716

My intent is to return to this, but I have a different design I to try using a rubber drive wheel instead of gears. I have thus far not had any good luck finding a drive wheel.