Doing stuff again.

[Chris333]

OK so I printed a cube and did all 4 sides different.

Top board inset .0007"
second nothing
3rd inset .0014"
4th inset .0007"


.0010"
nothing
.0020"
.0010"


.0015"
nothing
.0030"
.0015"


No use posting the 4th side.  I think the middle one looks good?

Top one at .0007" is what 18 microns is and should be my pixel size.

[chessie system fan]

I like the top one best.

[garethashenden]

I like the top one best.

Me too. Its nice and subtle.

[robert3985]

I agree that I think the top one is the most accurate representation.  I'd randomize it a bit more however so no readily discernible pattern is evident.

Can't wait to see this behind the car's support framework!

Cheerio!
Bob Gilmore

[Chris333]

So what is random? I suck at random! I don't understand in my head what random is. I agree the boards have a pattern, but what should I use for a formula?

I think 0.00085" is a good split to vary the boards.

This is what I got so far:


Originally had .004" wide board gaps that would be .004" deep, but now that is plan B.

[Dave V]

So what is random? I suck at random! I don't understand in my head what random is. I agree the boards have a pattern, but what should I use for a formula?

An actual random generator would be relatively easy to code in Python...

[Chris333]

Google could not translate that into English 

There are 20 boards. How much should I move each board?

[Dave V]

Google could not translate that into English 

There are 20 boards. How much should I move each board?

Yeah, I deleted that code. That was too complicated. But the point (and challenge) is that what humans think is random, and what random actually is are entirely different.

Tell a model railroader to plant bushes randomly on a module. The harder he ties to be random, the more evenly distributed the bushes will be. Humans try to find patterns in everything, even when on some level they know they're trying to avoid a pattern altogether.

Random involves clustering as well. So in your example, there will almost certainly be two adjacent boards of the same thickness somewhere. And others very different.

What you have looks more like a pattern of three thicknesses repeating in the same order.

[Chris333]

I think this is why I hate weathering. Because the only way I can do random is if I follow a photo of random. If I do it myself I will even everything out till it looks normal.

It took me a few music players on my phone before I could find one that would actually play songs at random. Most kept playing the same songs over and over.

[chessie system fan]

Coding something is overthinking things. Just roll a die. Each number corresponds to a board depth. That said, I recommend fudging the randomness and not putting the highest board next to the lowest board. In your test piece, that made too large of a jump, IMO.

[Scottl]

More steps in offset, even if they are sub-voxel sized will help generate random appearance.  Adjust until they look random and there are not jarring adjacent boards.

Alternatively, only offset a few of the boards.  Less might be better.

[nickelplate759]

What if you use dice to select offsets for your boards?  Don't use just one die, unless maybe it's a gygax.

For example,

Make a table of (say 12) tiny offsets, then roll a pair of 6-sided dice to pick which offset to use on which board.

By the way, you shouldn't assume that the offset is the same on both ends of each board.  They don't appear to be uniform in the photo of the prototype car.

Hopefully you won't get double-sixes three times in a row...

[Chris333]

lol  I have one being test printed now and I just guessed and did different depths for each board.

It is like when you tile a floor you are supposed to just grab a tile. But when you place it down you start questioning how it looks next to the other tiles.

[Chris333]

Meanwhile the Doodlebug chassis is running.

The bottom of the motor mount


It screws in from the bottom


It ain't pretty, but it runs


Surprised how it just barely makes my curves. I mean they is Tomytec after all. 

And turns out the stock ratio is only 12:1, now it is ~41:1.

[Chris333]

Only printed one side and one end because you just copy the other side and delete the brakes.



That popped out middle board right in the center... I just pushed it in .0005". It was the only board causing problems with my brain