How to keep curve length?

[pyrochlore]

Hi all,

I have curves modified by edit SOP and want that curve

keep the same length with the unmodified one.

I also found an old post said "Resample SOP" could achieve this.

The attached file is the one I tried but I still didn't get what I want.

(I want "perimeter_mod_new == perimeter_ori")

Please give me some suggestion to get it done.

Thanks.

keep_length.hip

Edited January 28, 2013 by pyrochlore

[ben]

You could resample by lenght, so you'd know how much segments are in your curve (create an attrib to sore NPT)

after your edit sop resample again with same length

delete all PT with number > NPT

This will not be exactly accurate, but depending on what you what to achieve it could be ok.

(see attach file)

Or

You could store your perimeter value

edit your curve

store new perimeter

calculate the difference (perimeter1/perimeter2) and use a carve sop

(this would be accurate)

keep_length.hip

Edited January 28, 2013 by ben

[pyrochlore]

Hi,

Thanks for the suggestion from ben.

I made a comparison to these methods (see attach file)

and found the method using Carve SOP get least length deviation.

Even for this method, the deviation is up to 6% { (perimeter_mod_new - perimeter_ori)/perimeter_ori }

and I want the deviation down to 1%.

I don't know where the deviation comes?

Is there any better way to solve this?

keep_length.hip

Edited January 29, 2013 by pyrochlore

[michael]

here is a file I got from SESI way back in the day...it'll get you within >1%

this is something that REALLY needs to be built in...

maintain_length_vop_example.hip

[pyrochlore]

Thanks, Michael.

That's a great solution I never thought.

In short, the method you provide is to apply the orient of segment in the modified curve to a copy of unmodified one.

So in the final curve we will have segment length equal to the unmodified curve and orient equal to the modified curve.

Hence the length of curve is kept the same.

This method is restricted to the case that the variation of each single segment is uniform.

The method will shrink\expand those segments to the original length.

It is not the same as we using Carve SOP, cutting the tail of the curve.

In my case, ben's solution fit my request more. (though the accuracy is not good)