MultiSurf & Rhino terminology, comparison & transitions??

[DCockey]

I'm interested in any information on the terminology MultiSurf (AeroHydro) and Rhino use to describe similar or the same features and geometry. I'm familar with Rhino but not MultiSurf.

I'm also interested in any comparisons between MultiSurf and Rhino capabilities.

Anyone with experience working in both MultiSurf and Rhino? How difficult was it to switch from one to the other? How much needed to be "unlearned".

[Ad Hoc]

I’ve not personally used Multisurf, but colleagues I know do. They say it is way way ahead of anything else, for its price and capabilities. They have used maxsurf/rhino and others, but still judge Multisurf to be superior. But it is ostensibly a geometry package and not trying to be anything else. It knows its limitations and sticks to it, hence being extremely good at one function, 3D geometry. I’m considering looking at it myself.

[yipster]

I looked at the demo a few times and thought its strenght was more the maritime applications as the 3d modeling
a good place to start may be glancing trouch the manuals of the modules
for boat design maxsurf is the better i think, its learning again tho specially when looking to master it

[DCockey]

A number of folks in the US are using MultiSurf in recording/documenting boats (creating lines, and construction drawings starting witht he vessel), and are familar with the terminology and capabilities of MultiSurf. There are also folks interested in recording/documenting boats who are familar with Rhino. Currently the two groups speak different languages. For example I've heard someone who uses MultiSurf talk about C-splines vs B-splines. As far as I can tell this is essentially equivalent to a "curve through points" vs "control point curve" in Rhino. That's just one example of differences in terminology. Hence the question about terminology.

I've also heard various claims about the capabilities of the two packages but haven't seen anything specific. It appears that for the 3-D portion of the purposes mentioned above both are capable but a specific comparison would be helpful.

There is also the inevitable question of what's needed for someone who uses one of the packages to become proficient in the other.

[CWTeebs]

They say it is way way ahead of anything else, for its price and capabilities. They have used maxsurf/rhino and others, but still judge Multisurf to be superior. But it is ostensibly a geometry package and not trying to be anything else. It knows its limitations and sticks to it, hence being extremely good at one function, 3D geometry. I’m considering looking at it myself.

I agree with this assessment. Multisurf is more accurate (in the numerical sense) but for many applications Rhino tends to be a bit easier to learn, and as mentioned the terminology in Rhino is a bit more standard/understandable. It's not wholly clear what the future holds for MSurf and AeroHydro.

[DCockey]

Originally Posted by CWTeebs

I agree with this assessment. Multisurf is more accurate (in the numerical sense) but for many applications Rhino tends to be a bit easier to learn, and as mentioned the terminology in Rhino is a bit more standard/understandable.

Off-topic from my original questions, but can you expand on "Multisurf is more accurate (in the numerical sense)" than Rhino? More accurate in what sense? Overall, or for particular types of geometry? Theoretical or do the differences show up in ordinary use?
With Rhino using the default tolerance settings, or irregardless of the tolerance settings in Rhino?

It's not wholly clear what the future holds for MSurf and AeroHydro.

MultiSurf pricing has been adjusted so that the minimum cost for MultiSurf is more competitive with Rhino.

[Ad Hoc]

Originally Posted by DCockey

but can you expand on "Multisurf is more accurate (in the numerical sense)" than Rhino?

From what my colleagues tell me, is that MS is 100% accurate. They trust the lines implicitly, having checked them over many boats over the years. Whereas from Rhino or Maxsurf the lines are sometimes not 100% accurate. Once the lines are drawn out large scale small errors are noticeable, but not with MS.

[DCockey]

What sort of errors? Can you provide an example?

Are these colleauges equally experienced with and knowledgable about MultiSurf and Rhino?

[Ad Hoc]

Originally Posted by DCockey

What sort of errors? Can you provide an example?

I don't have any examples, other than their feed back to my questions. In that the lines are always perfectly fair. The same cannot always be said for Rhino/Maxsurf. But this is also a measure of the tolerances of the end product you require too, and hence your level of "acceptability". A tolerance of 1mm on an ally boat of 50m is more than ok. But on a small GRP 5m boat, it is not.

Originally Posted by DCockey

Are these colleauges equally experienced with and knowledgable about MultiSurf and Rhino?

Their combined experience of doing this manually and on a computer via maxsurf, Rhino et al, is considerable.

One of them, for example, has been using maxsurf for almost 20 years. They even call him up for advice. He agrees MS does produce better more accurate lines and is more versatile, in the sense you can create what you like how you like, and developable. Masxsurf for example cannot produce perfectly straight lines, despite what the manual says!

[CWTeebs]

Originally Posted by DCockey

What sort of errors? Can you provide an example?

Are these colleauges equally experienced with and knowledgable about MultiSurf and Rhino?

Mr. Cockey,

Ad Hoc has answered this question, at least conceptually. I normally use Rhino because it's easy to use, but I recently ran into a case where I needed the Accurate surface evaluation methodology in MSurf. The WAMIT documentation describes "Fast" vs "Accurate" modes of the Relational Geometry Kernel (the heart of MSurf originally written by Dr. John Letcher):

Originally Posted by Wamit User Manual v6.7 Appendix C

RGKernel has two evaluation modes, denoted “Fast” and “Accurate”. In “Accurate”
evaluation, all curves and surfaces are evaluated recursively, using the actual math
functions that provide their definitions. In “Fast” evaluation, curves and surfaces are
evaluated approximately by interpolation in a stored tabulation of the curve or surface;
the fineness of this tabulation is controlled by the divisions x subdivisions products.

I don't know that I would say MSurf's "accurate" mode results in 100% accuracy, but, it can sure beat the pants off Rhino's NURBS representation of the same geometry.

[DCockey]

Originally Posted by CWTeebs

Mr. Cockey,

Ad Hoc has answered this question, at least conceptually. I normally use Rhino because it's easy to use, but I recently ran into a case where I needed the Accurate surface evaluation methodology in MSurf.

I appreciated Ad Hoc's "conceptual" response but was looking for more specific information rather than vague generatilities. Can you provide some specifics of what lead to the need for the "Accurate" mode in MultiSurf?

The WAMIT documentation describes "Fast" vs "Accurate" modes of the Relational Geometry Kernel (the heart of MSurf originally written by Dr. John Letcher):

I don't know that I would say MSurf's "accurate" mode results in 100% accuracy, but, it can sure beat the pants off Rhino's NURBS representation of the same geometry.

The description of the "Fast" versus "Accurate" modes in RGKernal/MultiSurf you posted describes different modes RGKernal/MultiSurf users for computations and is not a MultiSurf - Rhino comparison. I found another discussion of accuracy and the "Accurate option" in the MultiSurf V8 Manual, p207. In addition to a very similar description of the two modes, it also mentions that accuracy in the normal/"Fast" mode can be increasing the number of "divisions". I wasn't able to find any mention of being able to directly set accuracy or tolerances in MultiSurf.

In contrast Rhino allows the user to directly set tolerances and therefore accuracy. http://wiki.mcneel.com/rhino/faqtolerances The equivalent to "Accurate" mode in RGK/MultiSurf in Rhino is to tighten up the tolerances appropriately. To the best of my knowledge Rhino calculates geometry by evaluating the equations to the specified tolerances, not by using a look-up table. Has anyone making claims of superior accuracy for MultiSurf compared to Rhino tried tightening the tolerances in Rhino to achieve the same accuracy?

When looking at Rhino with the display in the various shaded modes or Rhino Render mode I've been fooled into thinking there were gaps and other problems with the geometry when what I was seeing was actually artifacts of the mesh used for the displays (not the underlying geometry). Refining the display mesh by the provided controls eliminated the gaps. I have to wonder if others have also seen gaps, etc on the screen, assumed they were due to fundamental problems with the geometry and didn't realize they were in reality due to the coarseness of the display mesh used.

[DCockey]

MaxSurf has been introduced into this discussion, seemingly as if comments about MaxSurf are relevant to Rhino.

MaxSurf and Rhino are different software, developed by different teams on different continents, and I'm not aware of any relationship between them. Based on a brief review of the MaxSurf users manual and my knowledge of Rhino there appear to be numerous, significant differences between the two programs relative to geometry manipulation, calculation and display.

MaxSurf is not relevant to a comparison of Rhino and MultiSurf.

[Tad]

I started using Multisurf in the late 1980's when it was called Fairlines. We thought it fantastically powerful at the time. Once it became Mutisurf we actually modeled decks and houses with it but it became unbelievably (now) time consuming. Multisurf still only allowed typing in coordinates (no dragging with mouse) when Rhino came along in the late 1990's. I continued to use Multisurf for a while to fair hulls but never any further modeling after I tried Rhino. Switched to Freeship for hull modeling as soon as it came along and never went back to Multisurf.

Pretty sure I've forgotten almost all the terminology but I remember "snakes"........maybe that's all?

[CWTeebs]

Can you provide some specifics of what lead to the need for the "Accurate" mode in MultiSurf?

I can't provide the side-by-side comparison of my geometry in Multisurf vs Rhino because it contains proprietary information related to an offshore drilling project and sanitizing it would not be a trivial task.

I believe the problem of "tricking" yourself into seeing "artificial" gaps may actually be related specifically to trimmed surfaces and the intersections between two mated surfaces, that the tolerance set in Rhino can adjust the level of continuity between the surfaces but doesn't actually iterate to ensure the residual error between the NURBs description and the "actual" geometry is below a certain threshold at every single field point on the surface.

I fully admit I may have mis-understood your description of the problem.

[CWTeebs]

I've sent an e-mail to two of my colleagues, one of whom is a developer at AeroHydro, to give a more concise description between accuracy modes in Rhino and Multisurf.

In the meantime, I still think Ad Hoc's initial feedback was accurate.