Calypso GD&T library vs. SmartProfile

[[Aa...]]

Background info:
When we first set up the measurement procedures for our metrology lab, Calypso's handling of geometric specifications was lousy at best.  (This would have been around 2016-2017.) So, at that time, the decision was made to ditch Calypso's reporting and only use the measured points from Calypso and run them through SmartProfile (a third-party software for evaluating measurement results from measured points).
 

Question:

Is there anyone here who has used the new GD&T library and SmartProfilie who can give me an idea of how comparable they are, now?

[[Aa...]]

Shameless bump...

[[Ke...]]

It might be better to explain exactly which specifications were "lousy at best", and if there has been any improvement in them. 

Does anyone else in the group even use SmartProfile? ...IDK.

[[Je...]]

I’m not sure what criteria were used to evaluate Calypso against SmartProfile, but I’m confident the evaluation was flawed. Your company doesn’t know what it’s been missing.

Yes, pre-2018, Profile of a Surface as a singular characteristic in Calypso needed improvement. No argument there. But everything else about the software blew SmartProfile out of the water. And Zeiss didn’t sit still. By 2019, they brought Profile of a Surface up to speed, with further refinements in 2020 and beyond. The new GD&T engine shows promise, but the software is superior even without it.

No disrespect to the great Dr. Greg Hetland, but CMMs are much more than point collectors. Calypso provides feature-based dimensional management and a visual representation of actual-to-nominal geometry. The power is in the pairing: sensor navigation, data acquisition, and evaluation working together seamlessly. Calypso’s reputation is so admired that competitors—looking at you, OGP Zone3—are out here trying to reverse-engineer it.

What your company is doing with Calypso is like hitching a Ferrari to a team of horses. You’ve got the keys to a metrology masterpiece, and you’re letting it idle in the driveway. Fortunately, it’s not too late to put the keys in the ignition and let it prove its superiority.

[[Ri...]]

The biggest thing that Calypso lacked was proper utilization of LMB/MMB and Simultaneous Evaluation. This has largely been addressed with the new GD&T engine. 

The strongest aspect of Calypso has always been its flexibility though. I'm not familiar with SmartProfile, but I believe it was directly integrated into Zone3, which I am familiar with. The biggest complaint I would have about their GD&T engine is that you have to follow a standard, and trying to wiggle out of that is either extremely difficult or impossible. I will say though that I always did like their option to select which version of the ASME standard you wanted to use (1994, 2009, etc.). 

 

[[Aa...]]

Thanks to all of you for your responses. 

Please sign in to view this username.

, I have to respectfully but strongly disagree. 

Perhaps the following rant will show you why and answer

Please sign in to view this username.

's question of what I mean by "lousy at best".

Our biggest issue with Calypso's GD&T evaluation had nothing to do with profile of a surface and everything to do with its handling of "Datum Reference Frames".  When we ditched Calypso's evaluations, their "Datum Reference Frames" were basically just cloaked Alignments that would not adjust to optimize the results in each degree of freedom to the extent it was unconstrained.  E.g., If one had a plane (defined as datum feature A) and two cylinders nominally perpendicular to it, with one of them defined as datum feature B and the other controlled with a Position tolerance to the DRF A|B, the clocking about the axis of the datum feature cylinder is fully unconstrained, so provided the second cylinder is oriented properly and the distance between them is good enough, that clocking could and  be shifted to make the reported position of the second cylinder fall in.  Yes, one could probably use PCM to iteratively solve for the clocking to optimize the results, but why?  To extend that example, if the second cylinder were defined as datum feature C, the DRF A|B|C would be interpreted completely wrongly by Calypso, and the clocking would be set by the direction of the axis of datum feature C, not by its position relative to the preceding features.  At the end of the day, the fact that Calypso had no way of selecting to which standard it would adhere was the strongest evidence that it wasn't designed to follow the standards.

I don't mean to start a heated debate about standards, here, but perhaps a respectful discussion could help us all understand each other.  The company where I work deals in a world where every micron counts (why else pay for a high-end Zeiss machine), so even though the complaints I stated above might seem small, they're not.

[[Je...]]

Please sign in to view this username.

 

I truly appreciate your willingness to stick to your convictions and follow up with clear technical reasoning.

Your example regarding DRF handling—especially with clocking degrees of freedom—is helpful, and I’m going to spend some time reflecting on your explanation. I agree with you that the way Calypso handles datum assignment in characteristic windows (if I’m understanding you correctly) acts more like a simplified secondary alignment with limited user control over specific degrees of freedom.

I’m not sure Zeiss intended this behavior to be “cloaked” so much as they were aiming for a sort of “GD&T easy mode.” From their perspective—especially around 2016–2017 as you mentioned—this design philosophy was likely biased toward ISO standards, where rigid interpretation often overrides user autonomy in defining constraints. Thankfully, Zeiss has steadily expanded its accommodation of standard-specific GD&T, and the new GD&T engine is another step forward.

Still, I have to chuckle at the fact that in order to enable ASME-specific functionality, Calypso now makes the user select a toggle labeled “loose” 🤦‍♂️. A cheeky choice of terminology, for sure.

Like you, I work in sectors where every micron matters. I’ve personally never found enough value in exporting evaluation to a third-party software like SmartProfile. But I have found it critical at times to consult the expert Application Specialists at Zeiss to help me understand what’s happening behind the curtain in Calypso’s calculations—especially when trying to achieve strict adherence to a specific GD&T standard.

I'd genuinely be interested in hearing more about your workflow—how you’ve achieved a handshake between Calypso and SmartProfile, and where you think the strengths and blind spots of each tool lie.
 

[[Br...]]

Please sign in to view this quote.

I'm surprised to hear this. There are fairly easy work-arounds for this like utilizing the secondary alignments. You can even construct a perpendicular line between datum feature B and C to constrain your tertiary rotational degree of freedom. Another thing you could do is create an intersection between datum feature C and A, and make the axis vectors constrained to datum feature A. This will create an intersected circle feature that is essentially the RAME. Then you can just pop that right into the tertiary box of your position characteristic and will solve correctly. Like Richard mentioned though, where it really was lacking was in utilization of MMB/LMB modifiers, especially with patterns of features.

Edited May 24

[[Aa...]]

Please sign in to view this quote.

Yes, you can do all these things, and I had started utilizing the intersection point trick somewhat regularly (The plane and two-cylinders scenario is all too common with some of our customers.), and while it would get close to the optimal "Candidate Datum Reference Frame", as ASME put it, it would not quite get there.  (The only real way would be to purchase the PCM add-on and program the measurement plan to iteratively solve for the Candidate DRF within the constraints imposed by the FCF, effectively mimicing what SmartProfile would do.)  Even if the workarounds work, why should we have to deal with workarounds?  Workarounds are basically an admission of "No, the software doesn't evaluate the results according to the standard, but if you twist it's arm just right, you can get it close, and then tell your customer that's what you've done and hope they consider that good enough to deem your reports credible because your reports have the Zeiss logo on them."  That dance just wasn't good enough for us when we had one customer say they would not send us any more work if we didn't start using Smart Profile.  Zeiss really missed the boat when they let QVI purchase Kotem.  (I guess those Germans were too stubborn to admit the Hungarians were doing it correctly, and thought they could do it better.)  It doesn't seem like they had any interest in catering to their U.S. customers' needs.  I'll stop here because I'm sounding old and stubborn, myself, talking about such ancient history.  (We bought our first license to SmartProfile in 2016, and at that time, Kos, the president and founder of Kotem, was still dealing directly with customers who were purchasing the software.)  

Edited May 27

[[Br...]]

Candidate datum reference frame? That's Y14.5.1 lingo.  So are saying SmartProfile just does a better job at optimizing the datum features that "rock" and selecting the best candidate datum? Calypso just kinda lets you pick the fitting algorithm and spits you out a datum. Sometimes the outer tangential evaluations struggle with rocking planes for sure. If you're working to Y14.5-2018 btw they moved to a single solution that minimizes the separation between the feature and the true geometric counterpart. So you don't really have candidate datums anymore I don't think. 

[[Aa...]]

Brett,

You're absolutely correct that Candidate DRF is Y14.5.1 lingo.  But no, the concept of candidate DRF's is not just about choosing the optimal combination of candidate datums (as implied by your reference to datum features that "rock").  It's about selecting the best Candidate DRF (Read: Candidate Cartesian Coordinate System, or Candidate Set of Three Mutually Perpendicular Planes).  "Candidate" within this context is determined by whether it meets the requirements stated by all datum references in the FCF.  My first example (plane as primary datum feature, perpendicular cylinder as secondary) sets no constraints on the clocking about the RAME axis of the secondary.  That's where Zeiss falls flat (or at least, used to).  Instead of finding the clocking that optimizes the result, it merely inherits that DoF from whatever Alignment serves as the parent alignment of the "DRF" Alignment (likely the Base Alignment).

I'm interested in what you have to say about the change in Y14.5-2018.  Can you expand on that a little more?

[[Je...]]

Please sign in to view this username.

I agree with your choice of the word "inherit."  Calypso has a tendency to act like a rube goldberg machine set in motion by the first assignment i.e. the base alignment.

I love the feature-based architecture, but I have to reserve a chunk of my mental DRAM to keeping track of what inherits what (or who's on first 😂).  They should add optional visual cues similar to Solidworks to display heredity.

I also agree that Zeiss missed a strategic acquisition with SmartProfile.  However, in the past 2-3 years they have shown remarkable improvement with standard-aherent GD&T, and the new engine is a huge step forward.

[[Ri...]]

I think they made the correct choice is passing on SmartProfile - I wouldn't be a huge fan of them trying to shoehorn in a 3rd party software just to handle GD&T. And based on my experience with the Inspect software I'm happy that Calypso allows you to measure per the standard, and per your own requirements. The standard is not absolute rule of law, and given the fact that they have walked back the use of OTE just goes to show that it isn't always correct or the best approach. 

[[Ch...]]

Please sign in to view this quote.

Great posts gentlemen ,   

Please sign in to view this username.

  I agree, I have to do extra tasks and use extra mental resources to ensure Calypso is doing what I would like it to do....

 

[[Ke...]]

Please sign in to view this quote.

I think I missed the memo on that one... details?

[[Th...]]

I don't know about walking back OTE, but there was some interest in a constrained LSQ definition for datum features several years ago. I have an academic paper I downloaded from ResearchGate, but the DOI resolves to this page, so if you want to give ASME $25 that's the official source.

[[Ri...]]

Please sign in to view this quote.

Per ASME Y14.5.1-2019

Please sign in to view this quote.

It all revolves around datum candidates, and if one isn't suitable to be used that it is appropriate to use the constrained L2 which is just LSQ shifted to the outside. So who determines what is suitable? Big grey area. 

This would really only applies to the measurement world, the functional world shouldn't really worry about this. 

I'm also attaching a research paper about it. 

ASME2016IMECE2016-67753.pdf