Well, I always wondered what that darn “Curvature” feature was good for…
Turns out it’s a kind of a way to fair your curves and the results are absolutely inspirational.
I have been going back through some favorite and not-so-favorite shapes and you can get things refined to an amazing extent. (I would not gush like this but it’s the most fun I’ve had in a while, and the level of confidence in your design goes way way up.)
So the game is this: especially with rocker, but also with the outline
when you make the little gaps disappear
and smooth out the curves in the Curvature display line–
then unView those lines and the control points and everything else but your shape and maybe the base line …
holy shihtzu
My fish, fun gun, flying saucer, stubbies, etc just went from neat to beautiful.
I’m not a board cad user, so this might be way off.
The curvature a rockers is a series of arcs, with different radii ( the plural of radius ). A smooth rocker will be a graduation of these - 5 foot radius next to a 6 foot radius next to a 7 foot radius etc.
In this rocker, see how the progression of arcs runs from 5 foot radius to 7 foot radius, to 14 foot etc.
This will give a smooth transition with no flat spots or bumps.
The same concept works for templates.
Sorry if I over analyzed it again…just couldn’t help myself.
Is that your guess to how that curve is generated?
or a reality?
I always figured it was most likely an interpretation/chart of the slope m of the tangent lines. As that can be easily derived for any point. This imaginary circle stuff sounds a little mathematically obtuse.
More specifically,
Are these smooth analysis curves a byproduct of a great hand
shaper and great boards.
Or an ideal imposed by CAD standards rather than from hand shapers?
How do these analysis curves for outline and rocker relate
to actual great boards designed by top shapers. As in what
do their hand shaped scanned board analysis curves look like?
Just trying to figure how much time to spend smoothing out the curves(and that can be extensive, outline) and its value.
I think the really well-faired curve is bound to be more hydrodynamically sound in terms of even lift and release and separation, and that’s what came back to me when I started to play with this thing–the more I faired them into correct curves according to the program, the more beautiful and graceful and the more like the work of the masters they looked. I figure their eyes just see that way, but when you’re working with limited resources and not doing it for a living, you can be forgiven for using the crutch of a program that does a good job of it.
I see the value of wings and bumps, but I like what I’m getting here on the curves.
Those curves were drawn on Autocad and are reality. When you see and feel the outline and the board with the graduated curves you’ll tell that it is perfectly transitioned. I think that the real masters templates, if measured would flow like this. It’s just that most templates are just sort of “fudged” untill they feel right.
I like the idea of quantifying the measurements, to be reproduced or modified for the next board. The closer the measurements taken from that “magic board” the better able the shaper is to duplicate. Since you can’t really measure the curve of an existing board like this, the idea is to draw out the perfect rocker template and make that board. Then you know what you have.
Make your templates with a tape measure and a nail as a compass. “Connect the dots” like a normal template.
All this analyzing lets a dope like me understand what the master is doing, and copy from his experience. I know other guys talk about running their hands along a board and feeling how good it will ride. I’ve just never been that clever; I need to measure.
I only use 2 points to define my curves in S3D and BoardCad (no mid rail point or mid rocker point). This means that I work with “pure” outlines on my pintails and my rockers are continuous curves.
In comparing to a friend’s 10 year old “perfected” hand faired rocker (over 100’s of boards) we’ve noticed that they are really close, but the computer curves are more subtle through the centre 1/3 (i.e. flatter/bigger arc) and carry more belly in the mid to tail 1/3 (we’re talking 1/8" max light showing between the curves here). He used chains to generate his original curves, but I guess in pulling the chain away from a 50/50 surve he created a “midpoint” between 2 curves.
Wow. So I tried that and it works. With Show Original Board checked, its outline visible, I deleted the mid-points, and I was able to get the lines back to where I wanted them, and the Curvature shows perfectly-faired curves–have to play with nose and tail, but can get it into a workable range for the rocker temp
This is what I do when I draw out my templates full-size with Adobe Illustrator. I have lately been taking some of my template collection and converting them to templates drawn in Illustrator. It is always amazing to me how much cleaner the outlines turn out.
The really interesting part is when I put my original templates on the computer generated one, they always end up being very very close but for some reason the subtle differences with the computer curves just seem to make them look so much better to the eye.
Some of them surprised me though as the original was almost identical to the computer generated one, I guess I just got lucky!
I now do all of my master templates this way, then I print them out on a large format printer on roll paper and transfer them to masonite. I have done about 20 templates so far, out of my 120 template collection, and every one of them has been a winner when I shaped a board from the template.
I have done a few tests where I made two identical boards shaped from my original template and shaped from the computer template. I then have a surfer/customer look at the two and ask them which one they prefer and they always pick the computer template.
The vast majority of the templates I have done only have one control point at the wide point and then two end points at the nose and tail, although on rounded pins the tail end point is a control point to control the tail curvature. I think there have only been two templates out of the last 20 that I had to add an additional control point at the nose 12" back.
I actually though this was pretty old news now, but I’m glad it’s catching on. If anyone want’s to know more about what this curvature really is all about check wikipedia, it has a pretty good article.
As Red_Boards suggest, if you use only two control points you can guarantee that the curvature is smooth and continuous, it’s on of the criteria for curve algorithms like the bezier. However, if you use more control points I would say it’s good practice to check the curvature to ensure that it’s smooth at the control point, otherwise the curve might look off. This is really hard to tell on the computer screen, but if you use CNC or print it an dicontinousy in curvature may well be visible. Indeed you can make exactly the same curve with two curves or more (you can test this by making one long curve and splitting it by adding control points and checking the curvature.). More control points offer a greater amount of control, but a single curve is quite flexible too.
I’ve read somewhere that you can visibly see if the first derivate of a curve is not smooth, but not an unsmooth second derivate. The curvature rely on the first and second derivate and should thus be smoother than anything you could do by eye.
One other thing I think the curvature is good for in particular is to analyze rocker. We get discussion where someone says that the rocker is increasing through the tail, some others like to decrease the rocker through the last few inches(supposedly good if you have some flex in the tail), you have the three stage rocker, continous rocker, constant radius rocker, etc. With curvature you can actually check this and quantify it, make a board with a flatter center section and a little more flip in the last foot, or flip starting exactly at the fins. Given that you can analyse this, you can compare boards, see which curvature works best and have more information than from straight rocker measurements. I might add that you can check the radius of the rocker at any point in the slidning info for the bottom.
I actually though this was pretty old news now, but I’m glad it’s catching on. If anyone want’s to know more about what this curvature really is all about check wikipedia, it has a pretty good article.
Thanks for your comments Haavard–I forgot to mention that part about accelerating rocker and flattened rocker–you can do a stagey rocker with it–it just is so much easier to blend it. Seriously great.
One of the cool things you can do with Illustrator is use the blend tool to morph steps between 2 outlines. For example if you had a funboard outline and a shortboard outline and you wanted to see an outline that would be somewhere in between the two.
Now if I could only figure out a way to import Illustrator splines into BoardCAD…
I agree with you I would be really stoked if I could import an outline from Illustrator, especially now that I finally have BoardCAD up and running and having fun with it!
In theory you could import the outlines by hand as I suspect that illustrator used beziers just like boardcad. If you have tool in illustrator where you can get the exact coordinate of the tangent and endpoint, just plug that into the appropriate edit boxes in the lower right corner and press set. You might have to swap x and y coordinates if you have the outline pointing upwards in illustrator.
Importing an illustrator file could be added to boardcad, but it’s not trivial. However we could add morphing from one outline to another quite easily. Added to the todo list.
Speaking of the to-do list, Haavard, I keep thinking it would be cool if you could jump to the next control point (clockwise) by using the TAB key. Been meaning to mention it for over a year.
