Thanks Steven,
Well I'm going to try getting some 3d prints made and will try both
methods. Your way is quickest, but the files are big if I try and keep
sharper detail. The alternative which I now have working is to break the
3000 cube model into its individual pieces using Joeys plugin, then Boolean
them all back together using a script supplied by Peter B. This is a much
slower method when dealing with so many cubes, but it works and generates
files that are much smaller.
Thanks again for your input.




On 28 November 2013 19:20, Steven Caron <[email protected]> wrote:

> if you need those perfect edges from the original cube then booleans will
> probably be your best bet. otherwise using my custom ice nodes you can
> decrease the voxel size a lot (increasing total voxels) and use the
> 'adaptivity' parameter on the volume to mesh node to reduce the density to
> something manageable.
>
>
> On Thu, Nov 28, 2013 at 5:29 AM, Chris Marshall <[email protected]
> > wrote:
>
>> Thanks Peter, much appreciated. I shall give it a go and report back.
>> Cheers
>>
>>
>>
>> On 28 November 2013 12:49, <[email protected]> wrote:
>>
>>>   as matt said, first separate each polygon island.
>>> there are some scripts out there that do this, but worst case: select
>>> with polygon island filter, then extract and so on.
>>>
>>> then boolean all of them one at a time - in the top right of the image I
>>> attached, was the line of vbscript I used for this:
>>>
>>> for i = 1 to 71
>>> ApplyGenOp “BooleanGenUnion”, , “polymsh” & i – 1 & “;cube” & i, 3,
>>> siPersistentOperation, siKeepGenOpInputs
>>> next
>>>
>>> It expects there to be a series of objects named cube1, cube2,...
>>> (select all your extracted meshes, alt+enter for a multippg and rename them
>>> – they should now have sequential names) and a series of objects polymsh0,
>>> polymsh1,... which are the results of each boolean.
>>> it’s not very smart for naming, so create the first boolean by hand,
>>> duplicate it and name it polymsh0.
>>> Now it should find all subsequent booleans which will be named
>>> polymsh1,polymsh2 and so on - that get created by the script.
>>>
>>> it runs a few seconds and the very last polymsh should be the result of
>>> all booleans, *if nothing went wrong*.
>>>
>>> If it has only part of the objects, that’s because somewhere along the
>>> line a boolean didn’t work – and resulted in an empty mesh. So find the
>>> last one before the empty one, and boolean them by hand.
>>>
>>>
>>>
>>>
>>>  *From:* Chris Marshall <[email protected]>
>>> *Sent:* Thursday, November 28, 2013 12:30 PM
>>> *To:* [email protected]
>>> *Subject:* Re: skin lots of cubes
>>>
>>>  OK I accept that, it's not ideal but it's a solution. I'd rather get
>>> the result you have with the multiple booleans / hard edges, which is what
>>> I was after, but how are you doing it?
>>>
>>>
>>>
>>> On 28 November 2013 11:26, <[email protected]> wrote:
>>>
>>>>   I’m kind of with Matt on this.
>>>> If it’s simple cubes and no rounded corners you’re after, booleans
>>>> would do.
>>>>
>>>> see the attached, the middle one is a scripted boolean between 70 cubes.
>>>> Nothing so fancy as what Matt suggests – just a boolean between two
>>>> cubes, then a boolean between the result and a third cube and so on.
>>>>
>>>> I was surprised to see that you can get something acceptable with
>>>> rounded corners even, depending on how demanding you are.
>>>> the one on top has the rounded corners shader - which unfortunately
>>>> only considers the convex angles.
>>>> the one below has a rounded bevel on all edges – so concavities are
>>>> treated as well – but there are some nasty spikes.
>>>>
>>>> Booleans generate less than ideal topologies for working with – doing
>>>> smooth deformations on top of all of this will be troublesome.
>>>>
>>>>  *From:* Matt Lind <[email protected]>
>>>> *Sent:* Wednesday, November 27, 2013 7:25 PM
>>>> *To:* [email protected]
>>>> *Subject:* RE: skin lots of cubes
>>>>
>>>>
>>>> Couldn’t you separate by polygon island, Boolean, then re-merge?
>>>>
>>>>
>>>>
>>>> That much could be scripted without too much trouble as you could do a
>>>> distance search between cubes to know which cubes should be Booleaned with
>>>> each other.  You could script it to Boolean while merged, but that would
>>>> require a little more work in the form of the algorithm.
>>>>
>>>>
>>>>
>>>> Matt
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> *From:* [email protected] [mailto:
>>>> [email protected]] *On Behalf Of *Chris Marshall
>>>> *Sent:* Wednesday, November 27, 2013 7:01 AM
>>>> *To:* [email protected]
>>>> *Subject:* skin lots of cubes
>>>>
>>>>
>>>>
>>>> Hi All,
>>>>
>>>> After managing to get a frozen poly mesh from hundreds of ICE cubes,
>>>> what I need to do now is perform something like a boolean union on them all
>>>> (though they are a single merged polygon), a bit like polygonizer but
>>>> without the round corners. Any tools out there that could do that?
>>>>
>>>>
>>>>
>>>> Thanks
>>>>
>>>> Chris
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>> --
>>>
>>> Chris Marshall
>>> Mint Motion Limited
>>> 029 20 37 27 57
>>> 07730 533 115
>>> www.mintmotion.co.uk
>>>
>>>
>>
>>
>>
>> --
>>
>> Chris Marshall
>> Mint Motion Limited
>> 029 20 37 27 57
>> 07730 533 115
>> www.mintmotion.co.uk
>>
>>
>


-- 

Chris Marshall
Mint Motion Limited
029 20 37 27 57
07730 533 115
www.mintmotion.co.uk

Reply via email to