MtI 3.2.5

Announcements, requests and support regarding the MAYA exporter
User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Tue Mar 06, 2012 6:08 am

fixed: Scale of a group objects is not working, the objects stay with the original sizes.

Hierarchies should result fully supported. I just tested this with one mesh grouped twice, each level being transformed independently (translate, scale, rotation, shear), with success.

This is not related with the persisting issue where instanced groups result in empty meshes.

Back on the tracks, update within two weeks max.
obsolete asset

ior
Posts: 166
Joined: Tue Aug 24, 2010 11:47 am

Re: MtI 3.2.5

Post by ior » Mon Mar 12, 2012 8:26 am

Object disappears on angle -720 degrees on X axis.

When animation is exported at the frame at witch the object angle is -720 degrees the object disappears because in the igs file it looks like this:

Code: Select all

<model>
	<mesh_name>objectShape</mesh_name>
	<scale>1</scale>
	<rotation>
		<matrix>1 0 0 0 0 -1 0 1 0</matrix>
	</rotation>
	<keyframe>
		<time>0</time>
		<pos>0 -0.1419999968 2.557999943</pos>
		<rotation_quaternion>
			<axis>1 -0 0</axis>
		

	<angle>-0.0004188790205</angle>
		</rotation_quaternion>
	</keyframe>
	<keyframe>
		<time>0.5</time>
		<pos>0 -0.1419999968 

2.557999943</pos>
		<rotation_quaternion>
			<axis>0 0 0</axis>
			<angle>0</angle>
		</rotation_quaternion>
	

</keyframe>
	<keyframe>
		<time>1</time>
		<pos>0 -0.1419999968 2.557999943</pos>
		<rotation_quaternion>
			<axis>-1 -0 0</axis>
	

		<angle>-0.0004188790205</angle>
		</rotation_quaternion>
	</keyframe>
</model>

<model>
Spotted what is wrong?

In this code:

Code: Select all

		<rotation_quaternion>
			<axis>0 0 0</axis>
			<angle>0</angle>
		</rotation_quaternion>
It should be:

Code: Select all

		<rotation_quaternion>
			<axis>1 -0 0</axis>
			<angle>0</angle>
		</rotation_quaternion>
Or something like that, or else the object disappears.
I have manually fixed that frame, but I do not understand why it gave that error.

CTZn, what do you think.

BTW, thanks for implementing the stuff I proposed. It helped me a lot. :wink:

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Tue Mar 13, 2012 2:09 am

In all cases the axis can not be 0 0 0, that is a bug.

But that's what the method returns currently if the keyed object is "not changed" (in terms of rotation 720°=0°, the method MtI is using doesn't like that case) and not moving.

I don't know yet what the best solution could be (detect axis without length or change whole keying method) but I'll look into it, thanks for reporting ior.

And, hum, is the formatting this bad really !? :oops:
obsolete asset

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Wed Mar 14, 2012 6:45 am

ior, the error with nurbs has been identified and is fixed locally, my mistake.

I may also delay the next release as I am changing some methods in my working environment, in order to allow branching, merging and other stuff usefull to programers I figured. This should grant me a greater control overall on my work, augmenting MtI's reactivity and robustness in the future.

An embryo of a pdf documentation should also make its way into the package soon.
obsolete asset

ior
Posts: 166
Joined: Tue Aug 24, 2010 11:47 am

Re: MtI 3.2.5

Post by ior » Wed Mar 14, 2012 7:11 am

CTZn wrote:I don't know yet what the best solution could be (detect axis without length or change whole keying method)
It should happen rarely because I have an object rotating and it passes 0 degrees 3 times, and on the 360 and 1080 degrees frame the result was fine.
CTZn wrote:And, hum, is the formatting this bad really !? :oops:
Just copied pasted from notepad, maybe that is the cause of bad formatting, I usually do not give importance to formatting, my code is always a mess, but I do it only for myself...

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Tue Mar 20, 2012 10:19 am

ior wrote:
CTZn wrote:I don't know yet what the best solution could be (detect axis without length or change whole keying method)
It should happen rarely because I have an object rotating and it passes 0 degrees 3 times, and on the 360 and 1080 degrees frame the result was fine.
In the MayaToIndigo Options window, Animation tab, raise the "Meshes Steps" to an higher value until the problem resorbs eventually.
obsolete asset

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Fri Apr 06, 2012 2:24 am

Since I've been at this, I believe that I can safely announce an update before the next week.

I am currently adding the basics of an automated materials type conversion to types MtI does support. While this first implementation will be relatively trivial it should already be of great help for beginners, and when dealing with big, imported scenes.That will be for a later time, another announce I'll have to regret *facepalm*.

In the future I will simply avoid announces and deliver. Good intentions are not what matters, even though they would be based on granted facts as recently.
obsolete asset

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Tue Apr 10, 2012 2:23 am

I'll put it that way: contact me if you want a test version.

The cool stuff is not stable enough for a minor release yet.
obsolete asset

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Sun Apr 15, 2012 8:20 am

Added aim at velocity for particle instances, finished nurbs implementation. Consolidating code where this wasn't possible before MtI used Maya matrices (holy fix).
Attachments
flocks.jpg
flocks2.jpg
obsolete asset

User avatar
OnoSendai
Developer
Posts: 6241
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Re: MtI 3.2.5

Post by OnoSendai » Sun Apr 15, 2012 9:19 am

Cool! Looks like those arrows need some motion blur though :)

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Sun Apr 15, 2012 10:32 pm

Yes, particle instances lack motion blur. The method MtI is using for keyed transforms is not relevant for dynamic particles, probaby the simulations would have to be baked first. I am reviewing the keyframe method, and more than that actually; I want to add robustness to the many "proof of concept".

I forgot to mention previously that unknown shaders are now converted to indigoShaders, defaulting to diffuse though. vRay scenes and other mr should render at once.
obsolete asset

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: MtI 3.2.5

Post by CTZn » Mon Apr 16, 2012 5:30 am

On the topic of particles, I want to add support for custom per particle attributes, rotation to start with. On a second thought it should be no big trouble to add an "arrow" motion blur based on the very velocity vector (linear motion blur) of dynamic particles.

Similarly dynamic attributes will make their way into shader code, through regular expressions I assume. This would allow for instance distance based shading and time dependent shaders, preserving the procedural chain.
obsolete asset

Post Reply
27 posts

Who is online

Users browsing this forum: No registered users and 8 guests