MtI 0.9 (19 nov 2007 release)
Rev. 143
Small Revisions of Icons / GUI (Camera, Sun Icons now updating properly)
Moved "string to bool"-functionality to mti_ModuleTools
Preparing to add Inverse-Transforms / Transforms
I'd prefer we do not export new .obj for animations, except if we've checked against Vertex-Animation, if that's possible.
I expect most animations will be Camera/Particle animations or simple transform-animations (rigid things moving).
Dougal, how do we fix the direct-render stuff ?
Small Revisions of Icons / GUI (Camera, Sun Icons now updating properly)
Moved "string to bool"-functionality to mti_ModuleTools
Preparing to add Inverse-Transforms / Transforms
I'd prefer we do not export new .obj for animations, except if we've checked against Vertex-Animation, if that's possible.
I expect most animations will be Camera/Particle animations or simple transform-animations (rigid things moving).
Dougal, how do we fix the direct-render stuff ?
yeah that's a good idea, but I don't know if it's possible.bkircher wrote:I'd prefer we do not export new .obj for animations, except if we've checked against Vertex-Animation, if that's possible.
Fix what, exactly? I know there's a couple of things to tidy regarding animation export, and the XML code display - I plan to fix that this afternoon. Other than that, I'm not sure what you mean.bkircher wrote:Dougal, how do we fix the direct-render stuff ?
REV 146
mti_Gen.mel:
moved saveFolder/saveName into mti_Gen()
- Added animation sanity checks
- Improved console output
- Enabled halt spp in renderer_settings, and removed haltsec and haltspp from the command line.
(should haltX be outside of AdvOptions?)
- Moved model export into mti_ExportModuleModels.mel
- Made ShowXMLOnRender work, added check to disable showing code on animation export
It's looking pretty tidy now, and everything seems to work pretty smoothly.
Can you guys test please?
mti_Gen.mel:
moved saveFolder/saveName into mti_Gen()
- Added animation sanity checks
- Improved console output
- Enabled halt spp in renderer_settings, and removed haltsec and haltspp from the command line.
(should haltX be outside of AdvOptions?)
- Moved model export into mti_ExportModuleModels.mel
- Made ShowXMLOnRender work, added check to disable showing code on animation export
It's looking pretty tidy now, and everything seems to work pretty smoothly.
Can you guys test please?
The issue requiring the instances object to be at the origin for it to work properly.
The issue of having rot=pos=<<0 0 0>> in the model and translation/rotation data in the obj itself. We're not really using <model> correctly.
Anyway, remember that SVN downtime I mentioned a while back... well I'm going offline now for a couple of hours. See you later, if all goes well
The issue of having rot=pos=<<0 0 0>> in the model and translation/rotation data in the obj itself. We're not really using <model> correctly.
Anyway, remember that SVN downtime I mentioned a while back... well I'm going offline now for a couple of hours. See you later, if all goes well
Ah yes that's right, I noticed that.The issue of having rot=pos=<<0 0 0>> in the model and translation/rotation data in the obj itself. We're not really using <model> correctly
As for original object in instances, as even without instances it is drawn usin the model tag, then just dont write a model for the original but the copies, or that's going to be something complicated and error prone IMHO.
If the mesh is instanced:
users: leave it untransformed, place instances;
devs: don't write it's own model.
The simpler the better.
obsolete asset
Back online...
REV 147
better implementation of <model>:
meshes are translated and rotated to <<0 0 0>>, exported, and then restored to their originl positions.
<pos> and <rotation> are now entered properly in the <model> element.
Split out the rotation matrix conversion routine into it's own proc (is now used in ExportModuleModels as well).
Temporarily disabled IESLight helper, new <model> stuff broke it.
I realise that there are some problems with grouped/parented/nested objects, and that their transforms are relative to their parents, they will not be exported to their correct locations.
We're getting closer, but there's still a thing or two to think about.
(perhaps just instructions to user to un-parent everything prior to exporting?)
REV 147
better implementation of <model>:
meshes are translated and rotated to <<0 0 0>>, exported, and then restored to their originl positions.
<pos> and <rotation> are now entered properly in the <model> element.
Split out the rotation matrix conversion routine into it's own proc (is now used in ExportModuleModels as well).
Temporarily disabled IESLight helper, new <model> stuff broke it.
I realise that there are some problems with grouped/parented/nested objects, and that their transforms are relative to their parents, they will not be exported to their correct locations.
We're getting closer, but there's still a thing or two to think about.
(perhaps just instructions to user to un-parent everything prior to exporting?)
- Attachments
-
- Indigo render overlayed with Maya viewport + text
- new_models_instances.jpg (87.43 KiB) Viewed 2965 times
Could do, but I found a better solution... it's possible to find the World location of a child object without unparenting it.
REV 148
Fix for exporting correct location/rotations of parented objects.
Fixed cd command in RenderDirect.
Re-enabled IESLocator.. NB, do not duplicate IESLocators, get errors about duplicate mesh names.
REV 148
Fix for exporting correct location/rotations of parented objects.
Fixed cd command in RenderDirect.
Re-enabled IESLocator.. NB, do not duplicate IESLocators, get errors about duplicate mesh names.
- Attachments
-
- Each sphere is a child of the sphere to the left. Same with the instance locators.
- untitled.xml.png (99.45 KiB) Viewed 2959 times
edit: replaced with flash version
http://gallery2.hamsterfight.co.uk/v/in ... 1.flv.html
http://gallery2.hamsterfight.co.uk/v/in ... 1.flv.html
REV 151
more work on mti_ExportModuleParticleInstances.mel for rotations... not getting anywhere
added TestScenes/particleInstance.ma
Can one of you instance/rotations experts have a look at mti_ExportModuleParticleInstances.mel please? I can't work out which values I'm supposed to be using for rotations, or even if my conversions are correct.
Indigo renders don't match the instance rotations in Maya. I've included my test scene in the SVN so you can see my particle setup.
more work on mti_ExportModuleParticleInstances.mel for rotations... not getting anywhere
added TestScenes/particleInstance.ma
Can one of you instance/rotations experts have a look at mti_ExportModuleParticleInstances.mel please? I can't work out which values I'm supposed to be using for rotations, or even if my conversions are correct.
Indigo renders don't match the instance rotations in Maya. I've included my test scene in the SVN so you can see my particle setup.
Hi Doug, amazing work ! Am I keeping saying that ? Well that's true then ! For some reason I'm still reluctant to coding. Please bear with that.
I'm no expert in Particles but I know that in some cases you got to create a PP (per particle) attribute before you can use it ! Well if you know that already then the problem is otherwhere (?)...
I'm no expert in Particles but I know that in some cases you got to create a PP (per particle) attribute before you can use it ! Well if you know that already then the problem is otherwhere (?)...
obsolete asset
Who is online
Users browsing this forum: No registered users and 34 guests