Comparing Freestyle to TSplines-adding new controls

[KQD]

We use Freestyle in Creo alongside TSplines in Rhino5 and Fusion360. In general, Freestyle is better in many areas, but there are a couple where TSplines wins.

1. When you split a control mesh face to add more capability for local editing the behaviour in Creo and TSplines is very different. In Creo, it immediately changes the shape. In TSplines it retains the existing shape. This is a crucial difference. Imagine you are carefully sculpting your form, are happy with it but decide you need a bit more localised control. In Creo, adding this changes the shape, so requiring you to try to backtrack to remake the form. In Tsplines the new control gets added so you can continue to tweak safe in the knowledge that your work up to that point is unaffected. This is a serious issue for us and probably our number1 annoyance with Freestyle.

2. Symmetry..TSpline symmetry can be applied at any stage and allows many planes of symmetry. Freestyle symmetry, we find, has to be done very carefully and is only in one plane. 

Given that Creo6 had little in the way of new features for core shape modelling I'd like to see enhancements in these areas....and of course multi body part modelling, which missed the cut in 6.

[Pettersson]

I'm nowhere near a Freestyle expert, but for your point 1, isn't that what changing the Level is for? If you change the Level from Base Level to Level 1, you will have more control points without affecting the geometry.

[KQD]

No. Changing the levels just changes the subdivision level. What T splines does is let you add loops without changing the form. When you add loops in Freestyle the shape changes

##- Please

[bbrejcha]

Ive done an extensive study comparing Freestyle Sub'D to Catia and T-spines. I prefer Freestyle as well and if your interested I could make a comprehensive list 

Not that we don't have  a few problems with the Creo version of Sub'd but ....

Bart Brejcah

Design-engine.com

[KQD]

Nice bike! When are you guys going to offer video short corses for sale - as much as I'd love to come to Chicago and all 😀

We absolutely prefer Freestyle to TSplines (it was Freestyle that made the Creo move a no brainer - just the rest of the software is not quite as no brainer!!)

What Freestyle does best (IMO) -

1. Variable Edge creasing (Tspline is on or off) - but the problem with this is that STEP exports sometimes don't work that well when you use this extensively. I'd like to see an option for exporting that inserts extra loops at the edges to mimic the crease - I think this might help the export issues.

2. Patch layout conversion - the top level subd mesh = the nurbs surface patches (Tsplines or PowerSurfacing don't).

3. Sub D levels (much like Modo in its ability to go up and down subD levels)

4. Curvature matching to existing geometry - pretty robust!

5. Speed. Freestyle is the Ducati. T Splines is the Vespa.

What TSplines does best:

1. Symmetry - Creo's symmetry in Freestyle is pretty poor by comparison.

2. Adding loops has no effect on the shape (see above thread) - this is actually a total ball ache for editing - would like to see this sorted!

3. Evaluation of the surface during edits - could be our lack of knowledge but Tsplines inside Rhino (not sure about Fusion or Alias) has active surface evaluation during edits (zebra/reflection etc).

4. Whole shape warps

The problem with all SubD implementations are what happens at export time. We do a lot of this work in the toy sector and all that goes to China or India exported as STEP primarily or SolidWorks (imported to SolidWorks from Creo, then various details added that are "hard to do" in Creo like surface splitting for face painting up). This is where (in my experience) it all falls apart in ALL systems. TSplines exports can be "troublesome". Power Surfacing (Add on for SolidWorks) ditto. As is Creo. I know the vendors want us all to use 1 system but 2020 team reality is designers use Creo/SolidWorks/Rhino/Modo/Fusion (us), dealing with others doing animations in 3D Studio or Maya or production tooling (often the same or things like PowerShape, NX/CATIA etc). No point in creating a wonderful form in an hour if it takes you 5 days to get it across intact to the next phase! Hence more "helping" export options would be useful.:

Convert Creo edge creasing to sub d loops at the edge zone and additional nurbs patches in these areas

Export the control cages for the freestyle features as a mesh object (selected feature or all)

Export the control cages as NURBS bodies (last resort option).

[Patriot_1776]

Crikey that's a complicated driveline!  Why the idler?

[bbrejcha]

no need to lock out the bike for accents.  

Bart Brejcha

Design-engine.com

[Patriot_1776]

Huh? Did you mean ASCENTS?  I know bicycles are absurdly complicated these days, but I'd rather lock it on those occasions than have to deal with the extra drag and complexity ALL the time as a result of it.  Also, I'd say that it's far easier to get things snagged in that idler, plus just the extra wear and tear and ginormous chain length...  Good luck with it!

Did someone actually build a prototype, ride it, and confirm this?  I'm not seeing how it would work that way.  Your weight moving up and down would still compress the suspension.  And the torque reaction on the rear sprocket from the chain remains the same and would RAISE the suspension on the powerstroke.  The only thing you can do to limit that is lengthen the distance from the swingarm pivot to the wheel axis.  Sportbikes (motorcycles) have done this as much as possible for this reason to reduce extension/compression of the rear suspension under driving or engine braking conditions.

Me, I have a bike with lock-out.....and never use them.  I just pedal and deal with it.