Measure.Extents still returning different values

[llorden4]

This post appears to suggest this was resolved for measurement variations

It appears to be accurate for about 3 decimal places, but it's not matching model distance measurement results.  For example...

Using the Measure Tool, shows the Y distance value of 490.462375832 but the Extents is returning a value of 490.46245175 for the same Y distance value.  Would be great to get these values to align.

MsgBox(Round(Measure.PreciseExtentsWidth,8))

[J-Camper]

@llorden4,

What version of inventor are you using?  I found this in the 2024.2 release notes while researching something else:

"Resolved an issue where 'Measure' command is causing a slight discrepancy when it is set to 'All decimal'. INVGEN-71989"

not sure when it became a bug, but looks like they addressed an issue similar to what you are experiencing.

[llorden4]

We're on 2023.4.1

[J-Camper]

I only brought it up as I don't know when the issue "INVGEN-71989" was discovered.  It may exist in 2023 versions of Inventor right now too.  If you change your measure tool precision from "All Decimals" to 8 or fewer decimals, do the 2 appear more accurate?  Or is there no difference, aside from fewer decimals?

[llorden4]

Ran another test and still have accuracy only to 3 digits...

[J-Camper]

I can't reproduce this with a sample I made myself.  Can you share one of the problem parts?

Sample in 2023.4.1:

Logger.Trace(Round(Measure.PreciseExtentsWidth,8))

Manual measure tool:     23.680845864

code:                                  23.68084586

[llorden4]

Here's a sample file I've been using.  

[J-Camper]

@llorden4,

I still cannot reproduce the discrepancy you are seeing.  I even derived the assembly choosing the bounding boxes of components:
bounding boxes with dimensions displayed by 3D annotationslogger results B is trace 1 | A is trace 2

4001-B:

Manual:                             491.316954540

Code:                                 491.31695454
Derived bounding box:   491.31695454

4001-A:

Manual:                            493.471285685

Code:                                493.47128568

Derived bounding box:  493.47128569

I did make certain to zoom in on the body as close as possible to ensure the vertices i was selecting with the manual measure tool were furthest apart in the Y direction as that is what we have been focuses on during the tests. 

[llorden4]

So it's just my imagination captured with a screen shot?  I don't get why you get different results just as you don't understand why I am.  The methods of point selection you describe are identical to my methods.

[J-Camper]

I'm not sure what to say, but I attached the 2 parts with work points created @ the vertices that I used to get the matching results.  If you measure between "extent point 1" & "... 2" you should get the same result as the code.

[llorden4]

Now this is getting interesting.  The two points you selected match the findings you are posting.  I was using a bottom edge to a point, that yields a different result.

This tells me that the plane created to cut with isn't truly "level" or the loft creates a slight skew that is making one side of that edge longer than the other.  This was specifically designed for that edge to be "square" to the world planes, so something isn't behaving quite as expected here.  So this issue lies elsewhere of the boundary box.

[J-Camper · Accepted solution]

yeah I had to zoom in until I could not get any closer before I could identify which vertex to pull the measurement, so it would be hard to notice the edge was not perfectly square to origin.