[leveller] Re: 64-bit STL plugin?
- From: "Ray Gardener" <rayg@xxxxxxxxxxxxxxxxxx>
- To: <leveller@xxxxxxxxxxxxx>
- Date: Sun, 9 Oct 2011 20:49:04 -0700
Great. So we had a good Thanksgiving weekend in Canada after all. :)
-----Original Message-----
From: leveller-bounce@xxxxxxxxxxxxx [mailto:leveller-bounce@xxxxxxxxxxxxx]
On Behalf Of Chris Dordoni
Sent: Sunday, October 09, 2011 1:40 PM
To: leveller@xxxxxxxxxxxxx
Subject: [leveller] Re: 64-bit STL plugin?
Ray,
I tested that plugin fix with the 477 x 301 file, and then a larger version
1908 x 1201. Netfabb, Minimagics and Meshlab all report that it is a
manifold solid, at each size. I think you nailed it.
Thanks,
Chris
On 10/9/2011 2:57 PM, Ray Gardener wrote:
> Okay, I did find a possible off-by-one error in the fan generator; you
> can download
>
> www.daylongraphics.com/download/usr/stl64.ltf
>
> which has a fix.
>
> Ray
>
>
>
> -----Original Message-----
> From: leveller-bounce@xxxxxxxxxxxxx
> [mailto:leveller-bounce@xxxxxxxxxxxxx]
> On Behalf Of Ray Gardener
> Sent: Sunday, October 09, 2011 11:29 AM
> To: leveller@xxxxxxxxxxxxx
> Subject: [leveller] Re: 64-bit STL plugin?
>
> I did a test here; the binary STL output will write the number of
> triangles to a 32-bit value at offset 80 (50 hex). In this case, I got
290,266.
>
> Four extra triangles likely corresponds to an extra triangle in each
> bottom fan. My visual inspections of the models don't show anything,
> so the extra triangles are probably overlapping redundantly.
>
> The last step of your calculation was off; it should be 1558 = 2 x
> (477 +
> 301) + 2. There are as many triangles in the fan as there are
> gridposts, because after all the vertices are joined there is one last
> inner triangle that has to be made to fill in the connector edge.
>
> I'll debug it here and see if I can find the four excess triangles.
>
> Thanks,
> Ray
>
>
> -----Original Message-----
> From: leveller-bounce@xxxxxxxxxxxxx
> [mailto:leveller-bounce@xxxxxxxxxxxxx]
> On Behalf Of Chris Dordoni
> Sent: Saturday, October 08, 2011 8:56 PM
> To: leveller@xxxxxxxxxxxxx
> Subject: [leveller] Re: 64-bit STL plugin?
>
> Hi Ray
>
> I looked back at a few files I processed with the updated STL plugin
> in
> Win32 and discovered that I had processed them through Netfabb Studio
> Basic
> (free) to close them. My recall that the STL from the updated plugin
> was closed as it came out of Leveller in the Win32 version was wrong.
> I uninstalled the Win64 Leveller and reinstalled the Win32 Leveller
> with the STL plugin update and ran a test just to be sure. So the 32
> and 64 bit versions are doing the same thing.
>
> I think the problem is that there are just a few extra polygons that
> are preventing the object from being manifold, at least by the number
> of vertices and triangles that are reported compared to what should be
there.
>
> The resolution of my test object is 477 x 301 pixels, these are my
> calculations so you can check them:
>
> 285600 = 2(476 x 300) front face triangles
> 3104 = 4(476 + 300) edge triangles
> 1554 = 2(476 + 300) + 2 back face triangles
> -------------------------------------------------------------
> 290258 = total triangles, compare to Netfabb report 290266 triangles
> before and 290262 triangles after repair Somewhere I missed accounting
> for 4 triangles?
>
>
> 143577 = 477 x 301 front face vertices
> 1556 = 2(477 + 301 - 2) + 4 back face vertices
> -------------------------------------------------------------
> 145133 = total vertices , compare to Netfabb reports 145133 vertices
> before and 145133 after repair.
>
>
>
> It's an easy fix with Netfabb Studio Basic, and as long as the files
> don't get extremely large (Netfabb is not yet Win64)
>
> Chris
>
>
>
>
> On 10/7/2011 10:39 PM, Ray Gardener wrote:
>> There's usually some good viewer tools to check the resulting mesh
>> and its normals. My only misgiving is that the triangles that connect
>> the hi-res edge to the low-res bottom quad use a simple triangle fan
>> generator that makes lots of very long, narrow triangles as
>> heightfield resolution increases. Other than that, I can't think of
anything.
>>
>> Ray
>>
>>
>> -----Original Message-----
>> From:leveller-bounce@xxxxxxxxxxxxx
>> [mailto:leveller-bounce@xxxxxxxxxxxxx]
>> On Behalf Of Chris Dordoni
>> Sent: Friday, October 07, 2011 7:12 AM To:leveller@xxxxxxxxxxxxx
>> Subject: [leveller] Re: 64-bit STL plugin?
>>
>> OK, I will continue to examine what I did to to see if I can find
> something
>> wrong in my process.
>>
>>
>> On 10/7/2011 9:57 AM, Ray Gardener wrote:
>>> It all checks out here.
>>>
>>>
>>> -----Original Message-----
>>> From:leveller-bounce@xxxxxxxxxxxxx
> [mailto:leveller-bounce@xxxxxxxxxxxxx]
>>> On Behalf Of Chris Dordoni
>>> Sent: Friday, October 07, 2011 6:46 AM To:leveller@xxxxxxxxxxxxx
>>> Subject: [leveller] 64-bit STL plugin?
>>>
>>> Hi Ray,
>>>
>>> Is the 64-bit version of the STL plugin the same version as the 32-bit?
>>>
>>> I'm getting some results here that indicate it may be missing the
>> correction
>>> you recently made to joining the front polys to the back polys so
>>> the
>> object
>>> is manifold.
>>>
>>> Chris
>>>
>>>
>>>
>>>
>>>
>>>
>
--
tel: 508 341 5403
website: http://www.concept-4d.com/websites/Chris_Dordoni_website.html
gmail chat: chris.dordoni@xxxxxxxxx
skype name: chrisdordoni
Other related posts:
