When your dimensions are using a decimal format, the precision value directly specifies the number of digits after the decimal point. For example, precision 5 gives you xx.yyyyy, and precision 2 would give you xx.yy

Negative values round the dimension data to the nearest power of ten, so -3 would round the value to the nearest 1,000 units, and -1 would round to the nearest 10. This may be useful if you are measuring very large distances (or large quantities of very small units), though I've rarely seen it used in practice.

When the dimensions are in a fractional format, then the precision value specifies the power of 2 in the fraction's denominator, so

- 1 -> 1/2
- 2 -> 1/4
- 3 -> 1/8
- 4 -> 1/16
- 5 -> 1/32

etc.Hope that helps.

DT