Replies: 249 (Who?), Viewed: 112469 times.
Page 4 of 10
Field Researcher
Original Poster
#76 Old 27th Feb 2010 at 8:50 PM
Oops, my bad. Yes, the files you sent has no UV records. I'll fix it right now and upload 0.1.7.
Field Researcher
Original Poster
#77 Old 28th Feb 2010 at 10:33 PM
I have added an extra file to the first post which includes the importing of bones for testing. This is basically 0.1.7 plus the bone info. I'm not upping the version as I'm not sure about the bones. The bones are there, but since I've not worked with bones before I'd want the opinion of someone else before fixing it in.
Pettifogging Legalist!
retired moderator
#78 Old 1st Mar 2010 at 12:56 AM Last edited by plasticbox : 1st Mar 2010 at 1:07 AM.
Thank you =)

I'm not exactly sure what the bones/joints that I've seen in Milkshape really correspond to in blender .. are thy hooks? Are they skeletons with transformation groups attached? I've only just started to read the docs, links below mostly for my own convenience (but may be helpful for potential other testers with no clue on the subject, as well):

http://wiki.blender.org/index.php/D...rials/Animation
http://wiki.blender.org/index.php/D...res/Object_mode
http://wiki.blender.org/index.php/D...rs/Deform/Hooks

I'll try redoing the one object I made so far that has an animation, and see what happens.


Also, look: http://www.modthesims.info/d/393576 -- I think that's .. 36 meshes all in all? I imported all of them with your script, which was a huge improvement over having to go through MS 36 times. Thank you!

Stuff for TS2 · TS3 · TS4 | Please do not PM me with technical questions – we have Create forums for that.

In the kingdom of the blind, do as the Romans do.
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#79 Old 1st Mar 2010 at 1:07 AM
I tried it on a bed. The bones were there, and assigned to the correct vertex groups. Something is wrong though, and I can't quite tell you what it is. I attached 2 blend files: one is the bed imported with your importer and one imported by .smd importer. The bones are flipped or something. Also, it seems the groups have their own sets of bones, i.e., if group 1 and 2 have assignments to bone A, then there is a bone A for each group. So, a clock which has three bones, one being assigned to both groups, has four bones after import.

And can you change the bone size/shape? They're huge.
Download - please read all instructions before downloading any files!
File Type: rar bedbonecomparison.rar (153.6 KB, 8 downloads) - View custom content

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#80 Old 1st Mar 2010 at 8:46 AM
A big thank you to both of you

Blender does have something with flipping. I've had to flip one of the vertex co-ordinates, and one of the UV co-ordinates. I wouldn't be amazed if this needed something similar.

The extra bones would most porbably come in due to something Wes mentioned in his last post. I'll have a look in an object with a few groups where the object appears in more than one group.

Thanks also for the two blends. I'll have a look today.
Field Researcher
Original Poster
#81 Old 1st Mar 2010 at 6:16 PM
Quote:
Originally Posted by plasticbox
Also, look: http://www.modthesims.info/d/393576 -- I think that's .. 36 meshes all in all? I imported all of them with your script, which was a huge improvement over having to go through MS 36 times. Thank you!


Had a look and those are surely going to be downloaded
Field Researcher
Original Poster
#82 Old 1st Mar 2010 at 10:42 PM
cmomoney, I saw something else now as well with the bones which I'll need to fix as soon as possible. The way I have it at the moment it's creating multiple armatures (depending on whether each group file has a skin section) and then creating a duplicate of the mesh object for each extra armature. This is clearly wrong and I'll make some adjustments to the script. I'll have a look at the sizing as well.
Field Researcher
Original Poster
#83 Old 1st Mar 2010 at 11:48 PM
Looks like the sizing is sorted, the duplication should be sorted tomorrow. I'll have to add checks for existing armatures and bones with the same names. The bones however are still lying horizontal instead of standing vertical. Progress looks okay and I think we'll have a full working import script soon, then I can start on the export.
Alchemist
#84 Old 1st Mar 2010 at 11:54 PM
I gotta tell you Ceixari it's fantastic what you've accomplished so far. It's great watching this tool get made. Sorry to cheer from the sidelines but whatever makes it possible for more people to mesh is all good
Test Subject
#85 Old 5th Mar 2010 at 12:51 AM
Your a life-Saver i absolutely LOVE Blender. if i could i would marry it!

thanks soooooooo much
Field Researcher
Original Poster
#86 Old 5th Mar 2010 at 10:34 AM
Armature is coming along, just a bit slower than I thought. The other scripts out there does it a bit strange, so I have to play a bit with different scenarios until I get the right one. Will update as soon I have something credible again.
Field Researcher
Original Poster
#87 Old 5th Mar 2010 at 11:23 PM
Armature all in (from what I've been able to test on a single bed). I've loaded a new version of the s3asc_import_with_bones.zip file.

Changes in this file:
1. I've had to move the per file armature outside the file processing loop as we're handling the armature as a single object rather than an armature object per file. The reason being that bones can be duplicated across files. The export script is going to be interesting.

2. I've renamed the mesh and object. Previously they both had the group name, i.e. group00, group01, etc. I've just added an 'm' and an 'o' in front to differentiate in the code so I don't pick up the wrong thing.

3. The bones are now upright instead of lying on their sides and they're shorter as well.

As soon we've had a little test of this, I think we can release the import script as Version 1.0.0. Then I can start on the Export.

Please note:
In some cases I've seen this not showing the bones, but they are there. Selecting another layer, then coming back to layer 1 seems to do the trick. This doesn't happen if you run the script through the text editor. There it looks fine. I'll keep checking to see what I can do about this.
Field Researcher
Original Poster
#88 Old 6th Mar 2010 at 8:38 AM
I forgot to say that due to the single armature, it will always be attached to the group00.
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#89 Old 6th Mar 2010 at 6:20 PM
I tested on the modern double bed, it gave me this error:


It still imported, but it only had one bone(root) even with switching layers.

I tried it on a clock:


This from the side view. The bones are pointing upward, but the whole armature is rotated -90 degrees.

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#90 Old 6th Mar 2010 at 7:35 PM
Thank you again for checking these out. Should be a simple enough fix for the rotation. I'll do that first and then check the bed error.
Field Researcher
Original Poster
#91 Old 7th Mar 2010 at 11:36 AM
I'm attaching comparisons of the ClockWallContemporary from MilkShape and Blender. The distance from the mesh to the bone seems to be further away in Blender (is that a problem?) I'm guessing the bigger problem is that the bone, when animated, goes backwards then downwards, instead of to the right and down as a normal clock?
Screenshots
Field Researcher
Original Poster
#92 Old 7th Mar 2010 at 3:14 PM
Ignore my last post (don't want to actually delete it, helps me think). I've been playing in Blender with toggles and I can now better see what's going on.
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#93 Old 7th Mar 2010 at 6:06 PM
How the clock should look (minus the duplicates):



I moved the bones from the 12 o'clock position so you could see. The duplicate bones were named the same with .001 added.

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#94 Old 7th Mar 2010 at 7:36 PM Last edited by Ceixari : 7th Mar 2010 at 7:53 PM.
I've now got the bones at the right places, just need the rotation and they're lying down again. But I'm happy with the placing now.
Field Researcher
Original Poster
#95 Old 7th Mar 2010 at 10:06 PM
Uploaded a new version of the bones test file. Placement and standing upright should now be okay, but I'm still playing with the rotation. It might be right by accident, but I promise nothing

I'm planning on adding extra features with regards to storing text and file names to make the export part easier.
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#96 Old 7th Mar 2010 at 11:59 PM
Tried the clock again. Imported bones with no duplicates, rotated correctly, and correctly placed.

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#97 Old 8th Mar 2010 at 8:35 AM
Thank you I've started looking at the bed and will be debugging some more during the day.
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#98 Old 8th Mar 2010 at 4:04 PM
It happens in other objects too, though I haven't been able to find anything in common.

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#99 Old 8th Mar 2010 at 5:15 PM
I think I know what the problem is, working on a solution at the moment.
Field Researcher
Original Poster
#100 Old 8th Mar 2010 at 10:32 PM
I've replaced the 'with_bones" file with a '-WIP' one. I've run quite a few objects through the import and they mostly look okay, even the modern double bed loaded without much fuss. The bed's bones look wrong, I think it's rotated, but all the other's looked okay. If you can confirm this, I can check for obvious differences in the text files we're importing to see. So far I haven't seen anything with rotation and have had to add my own rotations to get them to be proper.
Page 4 of 10
Back to top