An ILM R2 Story

[Duncanator]

Sorry it took me a while to get the videos uploaded.

The first is my Blue/White front logic displays. I think I'm still going to play with the pattern break-up. I think there are too many large areas.
The colors are a little washed out in the video, but you get the idea.

 

[Duncanator]

...aaaaand the Green/Orange rear logic. I'm happier with the pattern break-up on this one.

 

[joberg]

Quite beautiful indeed

 

[3Dsf]

Patterns look great. Not all hyperactive or random like most replicas. But the original lights often had a lot of black areas as well didn’t they?

 

[Duncanator]

Patterns look great. Not all hyperactive or random like most replicas. But the original lights often had a lot of black areas as well didn’t they?

I think you're right. Maybe I should add some black areas to break it up too.

 

[3Dsf]

Of course the tricky thing is that the ANH ones didn’t work correctly for the whole film. Sometimes they were off at the start of a scene - presumably Baker forgot to turn them on immediately - and other times the motor didn’t seem to be turning but the lights were on. And so sometimes you got mostly lit fibres and other times you got moving black patches.

The ESB ones are much cleaner and actually work consistently through the film. Personally I think the ESB ones tend to move too fast, but I do really like the swooping writhing patterns you see, since they’re so much more interesting than just a bunch of randomly flashing points!

 

[3Dsf]

But your colour wheel makes it easy to experiment with options, which is pretty awesome. You could even do an ANH one and turn the speed down, then swap out the disc and crank up the motor speed.

 

[Vagabond Elf]

Wow, the difference between the wheel and even the higher-complexity LED approaches is pretty significant! My LEDs (that I got through Astromech, I'm not claiming any credit for the design or code here!) do pulse through different colours, but which "pixel" is changing is straight up random. The effect is a sort of twinkling. With the colour wheel, the lights are moving in waves. They're almost hypnotic!

I'm sure that one could program the LEDs to do that, but I also think often times going old-school works better, and this makes for a lovely, unique trait for your R2!

 

[3Dsf]

Vagabond Elf : totally agree! I don’t want to malign the efforts at people who have designed electronics for droids over the years, but the flickering flashing look has never looked at all right.

I’ve done some experimenting with trying to move bitmaps around to try and simulate the original look but I lack the programming chops to pull it off! It’s obviously definitely possible to do, given a fast enough microcontroller and a bunch of dot/colour addressable LEDs.

 

[TheJugg1er]

It's interesting. I recently rewrote all the code for the Fiber Logic Engines, and there's actually a couple of different "pattern" styles in the code. My initial code replicated this swooping sweeping pattern really nicely (IMO), but most people didn't like it. The code that ships has the more random patterning. Still there's lots of options in my code that will let you tweak the appearance to your liking.

I'm also with nkg that the ESB speed seems way too fast. I personally prefer the slower more controlled sweeps.

 

[swmodelfan1977]

It looks great!

 

[Duncanator]

It's interesting. I recently rewrote all the code for the Fiber Logic Engines, and there's actually a couple of different "pattern" styles in the code. My initial code replicated this swooping sweeping pattern really nicely (IMO), but most people didn't like it. The code that ships has the more random patterning. Still there's lots of options in my code that will let you tweak the appearance to your liking.

I'm also with nkg that the ESB speed seems way too fast. I personally prefer the slower more controlled sweeps.

That's unfortunate that folks didn't seem to like your sweeping pattern. I would have thought that would be the look to strive for.

Initially I was going to go with the LED puck system that Grant came up with because of how compact and solid it was, but I really liked the idea that I could make the whole thing myself if I went Old-school. There's a nostalgia to that, and I wanted to veer toward "A New Hope" as much as I could too.

 

[3Dsf]

1976 or bust!

Well. Except for the damaged middle foot catches and the weak shoulder joints and non-removable traction batteries.

 

[Vagabond Elf]

It's interesting. I recently rewrote all the code for the Fiber Logic Engines, and there's actually a couple of different "pattern" styles in the code. My initial code replicated this swooping sweeping pattern really nicely (IMO), but most people didn't like it. The code that ships has the more random patterning. Still there's lots of options in my code that will let you tweak the appearance to your liking.

I'm also with nkg that the ESB speed seems way too fast. I personally prefer the slower more controlled sweeps.

Hunh. Somehow I was unware of that option! (I'm assuming you're referring to JoyMonkey's logic engines and not the Teecee's versions.)

When did the "swooping" pattern get added? Perhaps I'm merely out-of-date.

 

[TheJugg1er]

Hunh. Somehow I was unware of that option! (I'm assuming you're referring to JoyMonkey's logic engines and not the Teecee's versions.)

When did the "swooping" pattern get added? Perhaps I'm merely out-of-date.

I rewrote the code starting back in July 2020, because Paul (JoyMonkey) was looking for some help with the code. Having rewritten the code for the MagicPanel, and the code for the PSI Pro's from scratch, I seemed like I could probably provide some updates. I'm a FW engineer for the day job, so the code stuff is "easy".

Anyway, the code is designed for both the Teensy and the Zero versions of Paul's boards, and will provide the same functions on both versions of the hardware. The LED->fiber optic can look a lot like the original colour wheels that were used, and with the work I had, we get the "black" 'pixels' and I have matrix control so that if you want to have "dead: pixels to simulate the gaps in the matrix on the ANH R2, you can do that.

For what it's worth I have a laser disk version of the films, and spent over 4 hours re-watching the film and recording the PSI movements and Logic movements to replicate the motions as closely as possible. For example, there's only one on screen sweep of the PSI's in ANH - 1.06.50 left half red, right half blue -first on screen swipe!!!!!! Red increases to 3/4

Ok, enough background, and back to the question that was asked. Within my updated code, which you can get on my github (nhutchison/LogicEngine) I have a small piece of commented out code that uses a sin8(count*32) modifier. That tiny change will create the sweeping patterns, and since my own R2 is trying to be as close to ANH as possible is what I use.

 

[TheJugg1er]

You can see the "final" version that is the shipping code, where the "blinkies" are much more random. They're also a fair bit faster. Of course, you can change the speeds, as well as most anything else in the code if you care.

Sorry ... not trying to hijack the thread Duncanator

 

[Vagabond Elf]

You can see the "final" version that is the shipping code, where the "blinkies" are much more random. They're also a fair bit faster. Of course, you can change the speeds, as well as most anything else in the code if you care.

Sorry ... not trying to hijack the thread Duncanator

Ah, I'd misread your post. Or rather, interpreted "recently" as much more recently than you meant.

At some point I may reach out for that "sweeping" code... but that won't be until I'm working on an R2 dome, which means probably at least a year from now, and in any case as you've noted this isn't the place for that conversation.

 

[Duncanator]

You can see the "final" version that is the shipping code, where the "blinkies" are much more random. They're also a fair bit faster. Of course, you can change the speeds, as well as most anything else in the code if you care.

Sorry ... not trying to hijack the thread Duncanator

That's really cool! I like that you can switch up the colors, as well as run text.
Neat stuff!

 

[Pwrsrce]

Hey John… Quick question… What was your process for the domes finish?

I know the original dull spun look isn’t as accurate as we would have liked, and most people tend to make it too “mirror like” when they polish. I love the look you ended up with, and am curious if you documented the process.

Jason

 

[Duncanator]

Hey John… Quick question… What was your process for the domes finish?

I know the original dull spun look isn’t as accurate as we would have liked, and most people tend to make it too “mirror like” when they polish. I love the look you ended up with, and am curious if you documented the process.

Jason

Wow! sorry for the late reply Jason.

Our spun domes came with coarse spin marks form the forming process; so to get mine to a similar finish as the real ones had, I sanded it with a palm sander as well as by hand. I think I started with 150 or 220 grit, then brought it to 360 grit paper. Once I reached that point, I hand sanded it with red scotchbrite pads. That gave it the the smooth-but-not-polished look that the real domes had.

Now I'm only basing my dome's look on the condition of the original domes as they existed when they were all renovated in the ILM modelshop for the prequel films. I couldn't say how close they may or may not have been to their original finish back in the 70's for New Hope. There have been decades of who-knows-what had been done to them. I tried to match mine to the dented New Hope dome that I had to hammer back in to shape in 1999. To smooth out the hammer and dolly marks from my repair, I used the same method as I mentioned above to match the repaired area to the surrounding dome. I think all-in-all it blended in pretty well with all the other dings and scuffs on Artoo's noggin.