New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Adopting SI units for all ThreeJS physical quantities? #6259

bhouston opened this Issue Mar 19, 2015 · 6 comments


None yet
6 participants
Copy link

bhouston commented Mar 19, 2015

I would suggest that ThreeJS officially adopt SI units for its physical variables including distance, time, mass, light intensity, etc.

Right now ThreeJS is trying to be unit agnostic. While I can understand the motivation, it makes it harder to combine assets together in a ThreeJS scene -- artists need to come up with their own standards. Also many materials only look proper if you adjust their properties correctly relative to the scale of the object -- normal and bump maps in particular.

Also if you pick a non-standard unit, it can be hard to integrate physics properly. If you make 1 ThreeJS distance unit = 1km. You can not use SI units for mass specification in you are using ammo.js -- you need to also adjust those to match your unit scaling.

Unreal Engine 4: 1 Unreal Unit = 1 CM.
Unity 3D: No official unit system, but it is recommended to use 1 Unity Unit = 1 Meter if you want to use physics.
Softimage: 1 Softimage Unit = 10 CM. 1 Clara Unit = 1 Meter.
3DS Max: Various units, which causes a lot of conversion work.

My recommendation is to recommend standard SI units for all quantities:

  • 1 Meter for distances.
  • 1 Meter/Second for velocities.
  • 1 Meter/Second^2 for accelerations.
  • 1 Kilogram for mass.
  • 1 Lumen for light intensity when using physically based lights.

I think using standard SI units makes everything work together nicely. And it will lead to further maturation of the ThreeJS content pipeline, and increased interchangability.


This comment has been minimized.

Copy link

Mardonis commented Mar 19, 2015

Thank you for suggesting this idea. Standards in anything help work together. Was meaning to suggest this but you beat me to it.


This comment has been minimized.

Copy link

ghost commented Mar 19, 2015

👍 I am pro SI


This comment has been minimized.

Copy link

jeromeetienne commented Mar 20, 2015



This comment has been minimized.

Copy link

makc commented Mar 21, 2015

what code changes would be required, in order for 1 to become 1 meter? what would prevent anyone from creating a lizard model with -100 ... +100 (now meters) bounding box? so far, the only useful SI unit in programming is a second. mostly in the form of ms, in setTimeout-s,, etc.


This comment has been minimized.

Copy link

crobi commented Mar 24, 2015

I don't think this requires any code changes. You just put "We recommend you use models with a scale of 1 unit = 1 meter" in the documentation, and set all default values (like the camera clipping planes) assuming the scene is a couple of meters large.

As long as you can freely change camera parameters, don't have absolute thresholds inside the engine (like not rendering meshes with a radius of less than 1e-4 units), and have control over all assets in your application, there's nothing that prevents you from using feet instead of meters. Having a recommended unit system just makes it easier to exchange assets between three.js applications, or convert assets from another engine to three.js.

If you want to integrate a physics library, the unit becomes more important. The stability of the simulation might depend on the scale of numbers (due to the numerics), and there's usually many thresholds (for example, to detect stationary objects).

Example formulation from the Bullet engine wiki: By default, Bullet assumes units to be in meters and time in seconds. Moving objects are assumed to be in the range of 0.05 units, about the size of a pebble, to 10, the size of a truck. The simulation steps in fraction of seconds (1/60 sec or 60 hertz), and gravity in meters per square second (9.8 m/s^2). The page also goes on explaining that for some scenarios, centimeters, kilometers or even light years might be the more appropriate unit.

@mrdoob mrdoob added the Suggestion label Mar 24, 2015


This comment has been minimized.

Copy link

mrdoob commented Mar 24, 2015

Sounds good to me 👍

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment