blog/content/sundries/a-thoroughly-digital-artifact/notes.txt
2023-01-19 16:31:29 -08:00

143 lines
6.9 KiB
Text
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

inital comms with CNC shop:
------------------------------
Me: "project description": A relief map of California, carved from wood. Height exaggerated enough
to visibly discern the Santa Monica mountains. I can provide an STL file if needed.
Shop: STL can work but I cant manipulate it, which could save some money. If possible can it be
exported to an .igs or .iges or .stp format?
Me: Yeah, STP should be no problem. Can you give a rough estimate of the cost for 1x2-foot relief carving?
Shop: Without seeing the drawings, I cant give even a close price but in the past they range from a
few hundred dollars to several thousand dollars.
Me: That's totally fair! I'll get you some files in a few days.
------------------------------
next comms with shop three weeks later:
------------------------------
Hi Steve, I'm sorry for taking so long to get back to you! I had a harder time producing the IGES
file than I thought I would, but I think this should be OK:
(snip url to file)
It's 51 megabytes, otherwise I'd attach here.
As modeled, there's probably more high-frequency detail in the mountains than is necessary, as I'm
going for something that feels nice to the touch so smoother is better. It's also modeled at a
slightly larger scale than necessary, though not too far off (it's 500x577mm, and I'm interested in
the 400-500mm range for width; the relief height is in the 20-30mm range depending on scale). I was
imagining it would be carved with contour cuts in some thick nice ply, though I'm happy to hear
better ideas; I have literally no experience with making something like this.
(NOTE: go back to email thread and summarize the back-and-forth over tooling)
---------------------------------
Note that the shop did the extra work anyway just because they were nice, and that he was glad when
I told him it was a gift for my wife.
Zulip dump from 10 days after initial contact:
-----------------------------------
It IS Mt. Shasta!
After I made the mosaic out of the tiles I downloaded to cover the area, I masked it with an outline
of the state that I downloaded from a California gov geo site, then used a program called
gdal_translate to turn the image, a "geotiff" file with height data encoded, into that heightmap png
with the lowest value mapped to 0 and the highest to maxint.
I also had to reproject the geotiff with the height data into the same coordinate system as the
state outline was in. The height data was in a system using lat/long called "EPSG:4326", while the
state outline was made from line segments with 2d vertices in a projected coordinate system called
"EPSG:3857" with units of "meters". 3857 is "web Mercator", and is the coordinate system used by
Google and Open Street Map for their map tiles and other shapes.
It may or may not be surprising that cartography is very complicated!
My next step is to turn this heightmap into solid geometry that I can 3d print and/or send to a
local CNC shop to have them carve a relief of California out of wood, which is close to the final
step of producing an artifact as a present for my partner.
There are a bunch of python packages for working in this domain, but they're all just wrappers
around various GDAL libraries or tools.
The raw topo data I got from
https://portal.opentopography.org/raster?opentopoID=OTSRTM.082015.4326.1 (that was the epsg 4326
geocoded tiff; actually, several of them because you can only download up to 450km^2 at a time,
hence having to mosaic them with the gdal_merge.py command (the '.py' is in the name of the command
that gets installed when you do apt install gdal-bin)), then use gdalwarp to re-project to 3857,
then I had to write a python program to mask it for some reason, then gdal_translate (no .py on that
one, but they're all just python scripts) to convert to the png heightmap. I'm leaving out a couple
details in the workflow, but that's the main shape of it.
OK, actually, now that all that context is established, here's the actual command that produced that
file from the geocoded tiff:
gdal_translate -of PNG -ot UInt16 -scale -130 4412 0 65535 cropped_ca_topo.tif heightmap_cropped.png
and then I used convert (the imagemagick program) to scale the png from 33,311x38,434 to
2,000x2,2308 pixels.
the -scale -130 4412 0 65535 is mapping the height data min/max to the png whiteness in the output
file.
---------------------------------------
Zulip musings from a few days after that, still working on the heightmap:
---------------------------------------
(re: non-linear scaling of height to reduce pointiness)
ok, it was easier than I thought it would be. gdal_translate has a -exponent flag you can use with
-scale, so I remade the heightmap with an exponential scaling, using 0.41 as the exponent.
funny enough, I'm still working on this, since even when I drastically scale the size of the mesh
down in Blender (which I export to OBJ for import by FreeCAD), doing anything like modelling (eg,
extruding downward or joining with a solid base, or cutting the shape so it's CA-shaped and not a
rectangle) requires tens of gigabytes of resident memory and I keep having to kill the program and
start over.
a 60-megabyte OBJ file turns into 21 GB of resident data in the modelling software.
I have 32GB of RAM installed
that 21GB expands to 30 when I try manipulating it
------------------------------------------
Zulip from two weeks later (July 7):
--------------------------------------
Two weeks later I'm finally printing something out. I've given up on converting it into a parametric
CAD-like object; it seems this is a Hard Problem, but I'm not sure why. My goal with doing that was
to give a parametric CAD file to a local CNC milling shop, per their request, since when I suggested
a mesh-based file (STL), the guy was like "I can't do much manipulation with that to make it more
manufacturable, so a real CAD file would be best".
But at least with an STL file, I can print it myself. So that's going now, we'll see how it turns
out in no less than eight hours.
I haven't really done anything else with my computer besides this for a while.
(next day)
ok, I got something printed out, but I'm not super stoked on it. Also, I'm still chasing the elusive
dream of turning this into a parametric solid for easier CNCing. Vape pen for scale:
(insert shitty print photo)
(next day after that, the 9th)
I've finally "finished": I've created a mesh that has no missing faces, is not too crazy, and can be
converted into a parametric solid, and sent that off to a local CNC shop for a quote on having it
routed out of wood. I'll also do another 3D print, since the base is now a larger version of the
coastline instead of a rectangle, and the high frequency detail is a little diminished.
----------------------------------------
Links:
https://data.ca.gov/
https://portal.opentopography.org/raster?opentopoID=OTSRTM.082015.4326.1
https://www.printables.com/model/240867-topographic-california
https://touchterrain.geol.iastate.edu/
https://en.wikipedia.org/wiki/Shuttle_Radar_Topography_Mission