O-Inspect programs stopped working

[[Am...]]

So I have a couple programs that I have run hundreds of times and now all of a sudden they would not work and I had to adjust them to work but I don't want to leave them this way.

The issue I'm having is every single feature I had that was using the edge type deacreasing (or increasing) now all of a sudden will not get a single point. The only thing that works is if I change them to increasing and decreasing, which I do not want, I set them to decreasing for a reason and that is because increasing and decreasing often will plot points where I don't want them to be because of surface finish. When I had the edge type set to decreasing the programs worked very well for hundreds of runs and now i cannot get it to measure a single feature (even newly created features) to plot a single point.

Anyone seen anything like this? This is very frustrating and I really hope someone out there has some advice.

[[La...]]

Have you recently upgraded to a service pack or from 2017 to 2018? There were some image algorithms that changed, I remember reading that in the release notes. Just throwing that out there.

[[Am...]]

We had the calibration tech out here do the annual service so he might have updated that, and we just recieved a new o-inspect this week which is running calypso 2019 which is also giving me the same issue. That probably sounds like whats going on, however what am I supposed to do in this situation I've played around with it on a new program and no matter what Decreasing edge type doesn't see to be picking up anything anymore.

[[Ri...]]

Calypso 6.6.12 introduced a completely new image processing system.

"New image processing method
CALYPSO Service Pack 6.6.12 features a new, improved image processing method. This new image processing method provides the foundation for a considerable advancement in stability, performance and range
of functions of image-based functionality in CALYPSO.
With this service pack, we provide maximum compatibility to existing
versions of CALYPSO. Extensive testing ensures that measurement plans
from earlier versions of CALYPSO still run and deliver the expected results.
However, there are instances in which full compatibility is not available.
This applies when there is a complicated edge or incident light situation
and a fully automatic conversion to the new image processing method is
not possible.
If you encounter the following problems after installing CALYPSO
6.6.12, the measurement plans must be reworked:
– Features cannot be calculated
– Results differ compared to previous versions
In such cases, it is usually enough to adjust the light settings. In some instances, additional modifications to the edge detection may be necessary."

It's interesting that you are messing around with the increasing/decreasing edge. I don't think I've ever had to change that.

My initial suggestion is to phone customer support, and see if this is intended or not.

If it is intended, you will need to eventually rework your programs, but in the mean-time you can install Calypso 2017 (your license always allows for backwards compatibility) and use that until the issue is resolved.

[[Am...]]

The problem is that its doing the same thing on my 2017 version too which has had the service pack updated. Even if I create a new feature from scratch. The reason I had it set to decreasing is with the wide window for the feature increasing and decreasing will pick up too many random points on the surface, my options were have a very tight window or use a different edge type which prevented the points being placed in the undesired area. Ive had to change my program to use increasing and decreasing and make my range much smaller which forces me to adjust the location everyone in a while because the parts will vary. Its very frustrating because I had the program working without issues or needing to have an operator who can tweak the program.

[[Ri...]]

Nothing was changed in 2017.

[[No...]]

Did you eventually change the detection threshold to manual and accidentally set an unrealistic value? This happened to me a few times in the past and made the programs behave differently (although it shouldn't) until I set it back to automatic.

[[Am...]]

Threshold still set to automatic, I'm just gonna go ahead and re-write the program.

[[No...]]

Wait a minute...are you saying that existing programs ceased to work, but new ones do?

[[Am...]]

No, trying to use a decreasing type edge now doesn't give me any points on new programs either. I really needed to just get these parts measured so I've adjusted my programs to use increasing and decreasing and tweaked it to work that way a little better.

Still pretty frustrating though that it just starts working differently one day.

[[No...]]

This is not right. Especially the fact that one machine changed behavior after a tech guy was there makes me suspicious (did he also set up the other machine?) I would call support and ask if they can tell you what may cause this.
I know from a problem we once had with autofocus, that there is an undocumented parameter file controlling AF behavior. Who knows if there isn't something similar for edge detection?