Rounding units to metres and centimetres


I’m working on project where I use only two kinds of units - metres and centimetres. Is it possible to round Blender units to use only two places after comma? In other words - I’d like to turn off mm and μm values.

X m XX cm - that’s all I need.

there are addons to show dim in whathever units you want
but not directly in blender !

happy bl

I’d like to work only with meters and centimeters.

I have walls, for example 30m x 3m with many windows. I use numpad to keep good values and sometimes snipe vertex to vertex. Finally I have found, that some snipped vertices are shifted a few μm and walls doesn’t match at 100%. I’d like to round values to centimeters, to avoid such problems…

Here You can see what happens:

There is no way to avoid that… What you are seeing there are precision issues due to the way numbers are processed within Blender/computers in general (see here for a deeper insight).

Usually it’s best not to obsess about it, as those imprecisions are inevitable anyway. Food for thought: Even with a deviation of a few μm, Blender still is much more precise than that wall could ever be built in the real world…:wink:

I think it is way more then the floating math errors

Bl inside is limited to 7 digits
so with a scale of one you get 9999.999 Meters

so this gives you a precision down to one millimiters
but you can always change the scale to change numbers of digits on the right but then you got less on the left
always limited to 7 digits

also what you see in N panel can be modified cause this is using Microsoft Format
which can show more digits then what you have inside Blender !

happy bl

It’s due to the floats. Blender only displays 7 digits, but would do internal calculations with floats before rounding the result off to 7 digits. Floats don’t always translate to rational (in the mathematical sense) decimal values. In fact they usually don’t, so there is usually some small amount of error involved when rounding off.

For example, if you enter 3.1 into a standard decimal to floats converter, it will spit out 66 66 66 40 as the answer (Little Endian). However, if you then clear the values and input 66 66 66 40 as your starting point, the decimal value returned will be 3.09999990 which is obviously not quite the same as 3.1. In this case it will round off to 3.1 at seven digits, but this is a very simple example. Once you start using transformations on an object it would be perfectly possible to get larger errors that would affect a seven digit display.