emPolygonizer - metaballs for SOFTIMAGE

Discussions concerning plugins for SOFTIMAGE©
Letterbox
Posts: 391
Joined: 17 Jun 2009, 14:49

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Letterbox » 13 Jul 2009, 14:06

Mootzoid wrote:Thanks for all the advice, guys :ymhug:
I looked into all the formats and here is what will happen: PC2 and MDD are nice formats, but only for points => those formats aren't suited to cache the geometry (points and polygons).
So for a start I will support an own format (that can also cache the vertex colors) and the well-known and widely supported Wavefront OBJ format. That should be a good start, I think :)
What about Collada?

User avatar
Mootzoid
Administrator
Posts: 243
Joined: 03 Jun 2009, 18:48
Location: Germany/France
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Mootzoid » 13 Jul 2009, 16:21

Ah, yes, I hadn't thought about Collada. Definitively a format I should also implement.
Just downloaded the PDF of the Collada file format specifications... it is very detailed:

it has 298 pages! :((

Good thing I only have points, polygons and vertex colors
=> I will be able to skip 98% of the chapters ;)

Letterbox
Posts: 391
Joined: 17 Jun 2009, 14:49

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Letterbox » 13 Jul 2009, 16:30

Mootzoid wrote:Ah, yes, I hadn't thought about Collada. Definitively a format I should also implement.
Just downloaded the PDF of the Collada file format specifications... it is very detailed:

it has 298 pages! :((

Good thing I only have points, polygons and vertex colors
=> I will be able to skip 98% of the chapters ;)
298 pages, you cant say the details aren't there. ;)

But I'm not sure if it supports "difference" so it might end up as a mesh for each frame. Which is not the end of the world, you might need to look at shaders, eg you dont want n shader's per frame, and every frame, I dont know if that's possible though. Thats possibly the only downfall (good movie) that i can see.

Cheers
Terry

janimatic
Posts: 6
Joined: 02 Jul 2009, 18:36

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by janimatic » 13 Jul 2009, 19:35

obj is fine,
except it's an ascii file format so it's not very fast to read...

But if you provide a separate operator to read your own binary mesh format we could recompil for 6.5 and use the whole render farm...

Letterbox
Posts: 391
Joined: 17 Jun 2009, 14:49

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Letterbox » 13 Jul 2009, 19:41

janimatic wrote:obj is fine,
except it's an ascii file format so it's not very fast to read...

But if you provide a separate operator to read your own binary mesh format we could recompil for 6.5 and use the whole render farm...

But ascii IS easy to edit, just in case :-)

Cheers
Terry

janimatic
Posts: 6
Joined: 02 Jul 2009, 18:36

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by janimatic » 13 Jul 2009, 19:54

not so easy to edit at 24 files per second !
seriously ascii is really slower than raw binary data, specially if you need topology data

User avatar
CiaranM
Posts: 87
Joined: 09 Jun 2009, 01:37
Location: London
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by CiaranM » 13 Jul 2009, 20:09

janimatic wrote:seriously ascii is really slower than raw binary data, specially if you need topology data
Yeah, our networks would not appreciate it, especially when we start to deal with heavy meshes...So, it'd be great if your binary format could support compression, like the ICE cache format does (I think it uses Zip).
Oh and did I mention motion data... :D
Speaking of which, I really look forward (in hope) to the day when the ICE cache format can support topology descriptions. Would be great to be able to attach any custom attributes to a mesh cache.

User avatar
Mootzoid
Administrator
Posts: 243
Joined: 03 Jun 2009, 18:48
Location: Germany/France
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Mootzoid » 13 Jul 2009, 21:41

Hey guys, I somehow get the feeling that caching geometry is more than just a "nice to have" :)

PS: compression for the own format is a good idea.
PPS: the obj file format exists in ascii... and binary (.mod)!
PPPS: I will definitely not write a cache reader for XSI 6.5, sorry!
PPPPS: but I might try to compile the plug-in with the XSI 6.5 SDK.
PPPPPS: end of post scriptums :p

Letterbox
Posts: 391
Joined: 17 Jun 2009, 14:49

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Letterbox » 13 Jul 2009, 23:19

Ok what if you used Collada, which is ascii, but output it thru a LZW open source format, that would get you binary which would be faster. Which I believe some web servers and networks can do to minimize bandwidth use.

And if there were any problems you could unzip, and yes run edit's across all the files. With something like Ultraedit or grep if your a linux user. Worse case at least you'd have method for a fix, rather than a proprietary format which you couldn't edit.

Just a thought.

Cheers
Terry

User avatar
Mootzoid
Administrator
Posts: 243
Joined: 03 Jun 2009, 18:48
Location: Germany/France
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Mootzoid » 14 Jul 2009, 00:09

@ Terry :
Yup, I like it! That's just the kind of little feature that is highly appreciated when you (must) work with lots of files.

By the way: I added some example animations to the emPolygonizer web page, and tomorrow there will be a further one (Some tests using emPolygonizer with Costa's SPH stuff).

Letterbox
Posts: 391
Joined: 17 Jun 2009, 14:49

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Letterbox » 14 Jul 2009, 02:13

Mootzoid wrote:@ Terry :
Yup, I like it! That's just the kind of little feature that is highly appreciated when you (must) work with lots of files.

By the way: I added some example animations to the emPolygonizer web page, and tomorrow there will be a further one (Some tests using emPolygonizer with Costa's SPH stuff).

Thanks, and I want to look at the pages trust me, i do, but there are OTHER web pages i have to finish :)

Cheers
Terry

User avatar
Mootzoid
Administrator
Posts: 243
Joined: 03 Jun 2009, 18:48
Location: Germany/France
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Mootzoid » 14 Jul 2009, 12:05

Just added another little example animation,
here's a screenshot, below it is the link:

Image
"Care for a drink?"

cheers,
Eric

@Terry: No hurry, man ;)

User avatar
Hirazi Blue
Administrator
Posts: 5107
Joined: 04 Jun 2009, 12:15

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Hirazi Blue » 14 Jul 2009, 12:30

Looks great!!! :-bd
But to answer the question "care for a drink?": "Not that one, thank you!!!" :ymsick: ;)
Stay safe, sane & healthy!

Letterbox
Posts: 391
Joined: 17 Jun 2009, 14:49

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Letterbox » 14 Jul 2009, 13:21

Hi Eric,

I look at it while I was uploading, what else are you going to do?

The vids look impressive, but it was a long upload, so i watched it a few times.

Now dont take this the wrong way, I don't know the answer, so I'll ask. The SPH demo's still looks a bit "thick" for fluid, is the the sph not having enough particle density, or is it due to the fact of lower settings on emPolygonizer, or is it like I have a HP? Dell? laptop, and it hurts it :)

Cheers
Terry

User avatar
Mootzoid
Administrator
Posts: 243
Joined: 03 Jun 2009, 18:48
Location: Germany/France
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Mootzoid » 14 Jul 2009, 14:04

@ Hirazi Blue :
ah, come on, just a teensie-weensy sip !

@ Terry :
I do not assume responsibility for the movement of the particles, only for the creation of the mesh ;) Just kidding, the SPH simulation has about 10.000 particles, in other words: not very many => the particles are rather big => everything is a bit too thick.
I planned to make another SPH demo tomorrow with really many particles this time. Let's see how that looks...

User avatar
Mootzoid
Administrator
Posts: 243
Joined: 03 Jun 2009, 18:48
Location: Germany/France
Contact:

Re: emPolygonizer - metaballs for SOFTIMAGE

Post by Mootzoid » 16 Jul 2009, 23:42

Man, oh, man,

I now know why you guys talked about overstrained networks and compressing those cache files... they get really big really fast :-o
For a start I implemented an own format available in ascii or binary. It is very close to the internal XSI format what made it very easy to code. I will add some optimizations to the binary format (including a LZW compression), so that the files get smaller and reading/writing the files (especially via network) goes faster.
Then comes the Wavefront format. Collada will have to wait a little.

cheers,
Eric

Post Reply

Who is online

Users browsing this forum: No registered users and 32 guests