View Full Version : "Merge Average" - exists?? / enhanced "Merge Points"

08-15-2014, 09:14 AM
Is there a version of "MERGE POINTS" that allows the user to?:

Group select a bunch of points
Perform a FIXED merge (distance based) on those points, but
Place the merged points at the average location of those points which DO get merged?

AFAICT, Merge Points moves one point to the location of another point. I'd like a version that moves the merged point to the AVERAGE of the points selected to merge.

Here's a test object that I hope makes this clear. Use the Selection Set "Interior Points" to select the ...and merge at 60mm, then 120mm. You'll see what I mean.


08-15-2014, 09:20 AM
weld average

08-15-2014, 09:38 AM
weld average
That leaves me with ONE point, in the middle. I'm looking for something different.

Let me restate:

If one performed a MERGE POINT with a fixed distance, points would be merged in smallish groups. Let's call these "candidate points".

Right now it seems that MERGE POINT follows its documented behaviour in that the... heck here's a C&P from the dox:

Fixed merges any points within the specified Distance of one
another. When merged, one point will move to the location of
another — Merge will not average the distance between points
to create a new point midway between them. As such, the related
polygon may be stretched as necessary.

That's great and all, but I'm looking for the same grouping of points by distance, but then an AVERAGING of the locations. As it currently stands, MERGE seems to be acting more like WELD.

So, a new behaviour, as indicated by the dox, that DOES create a new point midway between all the candidate points.

(For that matter, is there a Selection tool that indicates which points would be candidate points for Merging???)

08-15-2014, 09:59 AM
Well, there is Weld GROUP Average (https://www.lightwave3d.com/assets/plugins/entry/cp_weldgroupaverage/) which is awesome!!

....but, it doesn't have a fixed distance option. (so you're going to have to get a little more physical with your selection set)

....And, it relies on point selection order. Which could, depending on how a particular model was made, require hand selecting (as opposed to a mass selection).

However, on your model, since it was made as one sheet, I used statistics to grab all points associated with three polys and deselected the outer loop of that selection, which left me with the pairs... and the plug worked like a charm. :thumbsup:

08-15-2014, 08:49 PM
Weld Average
That leaves me with ONE point, in the middle. I'm looking for something different. How so? Weld average does almost exactly what you want. Select a group of points, weld average, and bam, the points are merged into one point located at the computed center of the group. The only thing it doesn't do is ignore selected points outside of a desired distance range (the tool doesn't have a fixed range option because how would you tell it where the center of such a range is located?) Can't you get the same result by just selecting a cloud of points within your desired range (albeit not a perfect sphere, but does it need to be)?

08-15-2014, 10:43 PM
Because he wants different sets of points all merged to their own group averages, not a whole bunch of points merged into one.

08-16-2014, 12:43 AM
maybe >


or like Joe suggests, Weld Group Average...