DIY + AI = 0.0 x ZZZs

Greetings from the Bunker!

3 BH2 microscopes

It has been another busy Monday, getting ready for another busy week. But it has already been a very productive one. I am delighted to say that I am very close to finishing up the printing and ready to begin the final assembly of TEN cross polarization platforms. This feels like a very important step as it seems to have taken an age to get through all this printing. Everything has gone quite smoothly and I managed to find the time to write a detailed instruction manual for the crystal viewer. I would expect to have this batch mailed out by the end of the week, after which both my printer and I will be taking a short break before getting started on the next batch. I thought you might be interested in looking over the instruction sheet so I am posting that separately. This sheet would be very helpful to anyone who is planning to print their own viewer.

10 cross polarizers, in the making

To say I underestimated the time and cost of making these things would be a massive understatement - my printer has gone through three spools of filament in a solid week of round the clock printing. The sudden addition of huge import fees (tariff penalties charged directly back to the customer) caused a tripling of my costs for all the parts I order from China - I can’t afford to buy the same parts from a US supplier because they cost over five times as much here. Anyway, it got real expensive, real quick. But I am a man of my word and the folks that ordered this first batch of viewers will have their orders fulfilled  for the price promised.

the end product

You would think that, as the only supplier of self contained cross polarizing birefringence viewing and photography platforms on the planet (that I am aware of), I should be able to make and sell the things to make a little profit. It is testament to my limited business acumen (very limited) that the completion of my first round of sales will leave me squarely in the red - it normally takes me months to get into debt with a business venture! The viewers have turned out very nice, with all the upgrades. If you are on the list to get one, you will not be disappointed.

Former BH2 microscopes - ready to paint

The news for those awaiting completion of their Olympus microscope conversions is even better! There were two major problems hindering the completion of the four machines sitting on my desk. The first was that 3/4 of the focus blocks had badly damaged fine focus drive systems, with bent focus shafts and stripped gears. I had thought this problem was fixed when I ordered some replacement steel tubing (from China) and started printing the tiny gears that needed to be replaced. But the tubes turned out to be  4.0mm in diameter, not the 3.97mm (5/32”) that I needed - I would not have thought that an extra 0.03mm (30 microns) would have been enough to prevent a steel shaft from passing through a 40 year old bushing - but it is.

from China - $7 for 6

I found some tubing of the correct size but its walls were either too thin or two thick to properly tap them for the 3mm screw that secures them into the focus knobs. I broke all my 3mm taps trying. I could not drill out the thicker walled tubes either. None of my small bits could handle the hardened steel of the tubes. Just as I was getting desperate, the last order of Chinese tubes arrived and they were the perfect size. They fit in the bushings and had walls just the right thickness to handle the 3mm taps that arrived from China in the same shipment. So at the last moment I was able to put together 3 fully functional and perfectly straight shaft assemblies.

the culprit

But that was not the biggest issue - the real problem was that I had somehow messed up my measurements when making a drilling template for transferring the hole positions from the motor bracket to the wall of the focus block - the holes were all about 2mm away from where they needed to be for the motor bracket to clear the course focus knobs of the focus block. This was a big deal because the only solution that I could see was to reprint all four motor housings - a huge task that took 23 hours and a spool of filament. And I wasn’t even sure I could redesign the mount with sufficient accuracy to get the mounts aligned perfectly with the holes. Then this morning I had an idea. I don’t know what this is called, if it even has a name, but this printed piece was what I had come up with.

The part on the left saved my bacon!

It fits perfectly over the base of the unmodified cage and holds it in exactly the right position to clear the focus knobs while allowing me to reposition the mounting holes to a new part of the microscope body. And they work perfectly! Of all the cool gadgets I have made with this printer, this mount adapter is by far the most satisfying.

The adapter plate in position

So, with the microscopes dismantled, deep cleaned and fully reconditioned, I can finish the final assembly and get them all tested and ready for shipping in the next few days - I hope. I have learned more from these two projects than I could ever have thought possible. This kind of problem solving is why I love this part of my work so much. You should give it a try! Time to change gears and tell you about this week’s programming…

The coming week is going to be very interesting. I have been thinking a lot about how the exploding field of artificial intelligence is starting to change everything, and I mean everything. Of course, I am most interested in getting myself up to speed on how it might impact me as a photographer. I have found some new applications for this technology that have shocked even me. So I decided it was time to do an AI update for macro photographers - and before you remind me that I did this only a few months ago, I think an update is long overdue. So on Tuesday’s Macro Talk, at 8PM, I am going to let you know about a few developments that could have a hugely positive impact on us as both amateur and professional photographers. I will be introducing you to five new applications for large Language Model-based AIs and showing you how close they are to upsetting “business as usual” in the world of macro. I don’t want to spoil the fun by telling you about these things ahead of time, but make it on Tuesday, if you can. Here is a link to the stream. https://youtube.com/live/i9fl4bbIMxI?feature=share

Following on from that we are going to look a little closer into how a relatively new kind of application that uses existing technologies to create AI Assistants, could make us better photographers. It is going to be a contentious discussion, I suspect, but there is a lot to talk about and getting the questions asked is a good start. Here is your link to Macro Talk Too, Thursday at 2PM. https://youtube.com/live/o6Vur5r1zZM?feature=share

Holding to our theme for the week, I will be bringing another controversial topic to the AfterStack conversation on Saturday morning at 10AM. I will propose an entirely new approach to photographic post-production and, by extension, photographic training, This is going to be charged discussion but one that we are going to have to have sooner or later - what comes next. We already rely quite heavily on AI to help us edit our images. So what else is right around the corner, or in some cases, right here, right now? Sure to ruffle a few feathers, this discussion needs to be had and on Saturday we do just that! Here is your invitation…

Allan Walls is inviting you to a scheduled Zoom meeting.

Topic: AfterStack with Bud Perrott and Allan Walls - Episode 23

Time: May 17, 2025 10:00 AM Central Time (US and Canada)

Join Zoom Meeting

https://us02web.zoom.us/j/6916802815?pwd=TS9tZi9ZL1NXeVUvOUF4eTg5YjdlZz09&omn=85292399173

Meeting ID: 691 680 2815

Passcode: 678122


So with a lot of controversy on the horizon, let’s get into it tonight and see what has changed with machine intelligence since we last talked.

I hope you enjoyed last week’s Stereo conversation and learned enough to try your hand, wowing our celebrity judge with your stereo images! I can’t wait to see what you all come up with!

Please try to make it tonight and let’s get this week of AI off to a good start!

And don’t forget to check out the instruction piece for the new crystal viewer - it is the next post after this.

One of my favorite crystals - top secret recipe

All the best,

Allan