"32mm" means nothing
If you have been in this hobby for more than a week, you have seen it. You download a "32mm" miniature, print it, and it towers over your "28mm" army. Or it looks like a child next to a tank that was supposedly "1:56."
We call this scale creep. Every sculptor measures differently. Some go to the top of the head, some to the eyes. Some account for base thickness, some do not. For anyone trying to build a coherent collection, it is a mess of failed prints and wasted filament.
I have thrown away enough mis-scaled prints to know. That is why I started measuring everything with digital calipers.
The 175cm baseline
I built STLACCESS in a bedroom workshop. I realized that while "28mm" is a marketing term, "1:56" is a mathematical ratio. To bridge the gap, I needed a fixed point.
I settled on the 175cm human baseline as a consistent reference point. There is no single industry standard — different manufacturers, sculptors, and game systems all use different baselines. The 175cm figure is a practical compromise I chose so that every conversion on this site uses the same starting point. If you are used to a different convention, the ratios between scales stay the same regardless of the baseline.
Eye level (160cm)
The standard for fantasy and sci-fi miniatures. Measures to the eyes, not the crown, because helmets and headgear vary. A 175cm human has eyes at 160cm.
Top of head (175cm)
The standard for historical military kits and display busts where realism matters. Total height from feet to crown.
Why this works
The 175cm baseline is arbitrary but consistent. It does not matter if the "average" human is 170cm or 180cm in real life. What matters is that every conversion on this site uses the same starting point, so the ratios between scales are mathematically sound. A 28mm to 32mm conversion is 114.3% whether your reference human is 175cm or 185cm. The baseline just gives us a common language.
How I test the numbers
Math on a spreadsheet is not enough. I print test pieces and measure them with digital calipers (0.01mm resolution). Here is what I have learned from my own prints:
FDM accuracy
Over-extrusion can add 0.3% to 0.5% to dimensions on FDM printers. I calibrate e-steps and flow rate before each test.
Layer height bias
At 0.08mm layer height, a 28mm model has 350 layers. At 0.2mm, it has 140. The coarser layer height can create a 0.1-0.2mm stair-step error on curved surfaces.
Thermal expansion
PLA and PETG can shrink 0.2% to 0.5% as they cool from printing temperature. This varies by brand and color. I note material-specific offsets where they matter.
I test every percentage on my Bambu Lab A1 Mini before publishing. If a number fails on my machine, I adjust it.
From a 40 euro printer to calipers
I started 3D printing back in 2017 with an entry-level FDM printer. Prints warped, dimensions were wrong, and I had no idea why. I spent more time troubleshooting than actually printing. That first year was a crash course in extrusion rates, bed leveling, and the painful reality that not all STL files are scaled to anything meaningful.
In 2019 I tried turning the hobby into something bigger. I started a local print service in Enschede, Netherlands with a Bambu Lab X1 Carbon. The goal was simple: help other hobbyists get the parts they needed without owning a printer themselves. It did not work out. The margins are brutal, the demand is inconsistent, and running a service business left no time for actually enjoying the hobby. I had to sell the machine and walk away.
It took years to get the finances back in order. Now I print on a Bambu Lab A1 Mini — a fraction of the cost, none of the business pressure, and I am happier for it. I print Warhammer models, board game accessories, figures for friends, and the occasional household hack that makes life easier. All FDM, all PLA and PETG, all calibrated the same way.
That failed business taught me something important: most people do not have a room full of printers. They have one machine, a spool of filament, and a desire to print something that looks right. Every percentage on this site is tested on the same kind of setup most hobbyists actually use.
What STLACCESS is
It is a tool I built for myself. I do not sell STL files, filament, or printers. There are no affiliate links. It is free because I needed it and figured others did too.
The numbers come from my prints. I calculate them, I print them, I measure them. If you submit a correction through the reference pages, I test it on my machines before updating. If enough people tell me the same thing, I change it. That is the deal.
It is never finished. New scales appear every year. New games use different baselines. I update the library when I find discrepancies or when manufacturers change their sculpting standards. If something looks wrong on your table, send it in.
By the numbers
44+
Scale conversions
5
Free browser tools
3
FDM profiles tested