You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am uncertain if this is known issue, but the cellular distance algorithm will output values outside the range of of -1.0 and 1.0 under normal conditions, specifically with the "hybrid" distance function with my tests.
Other language ports output the same out-of-range values with equivalent settings. The web preview is coded defensively and clamps the values into range before conversion to an integer, but it is visually noticeable with the abundance of white in the output compared to similar configurations.
For the C port, this is the offending part of the algorithm on line 1647:
Cellular output is actually never bounded, it's just normally falls within bounds unless using hybrid distance.
I have done some work to add correct bounding to it, but since it is a output changing fix I'd want to bump the minor version at least
I am uncertain if this is known issue, but the cellular distance algorithm will output values outside the range of of
-1.0
and1.0
under normal conditions, specifically with the "hybrid" distance function with my tests.Minimal example using the C port:
Output:
Other language ports output the same out-of-range values with equivalent settings. The web preview is coded defensively and clamps the values into range before conversion to an integer, but it is visually noticeable with the abundance of white in the output compared to similar configurations.
For the C port, this is the offending part of the algorithm on line 1647:
Assuming this is correct, would it be better to clamp it before returning the result from this function?
The text was updated successfully, but these errors were encountered: