Animation Set Translations not Exporting

I have created a couple animation sets which translate an object from a starting position to an ending position and those animate just fine when triggered with a script.

Additionally, I have configured a timeline which includes these animation sets at their key places and when I run the simulation everything appears on screen as expected. The objects appear to "fall" into position as opposed to just appear in position.

My problem occurs when I attempt to record/export this animation. My workflow is to open the animation ribbon, click the record button then start the simulation from the beginning and finally hit the animation stop button once the simulation has finished. When I export the animation made in this way the objects simply appear in the starting position for the duration of the time they should be moving through space. At the point the objects should be in their final positions they disappear from the scene altogether. If I make a shorter animation by clicking the record button then triggering the translation animations with a scripted key press the resulting animation only includes 2 frames, the start and end positions of the object with the translations lost.

How do I get an exported animation to include object translations?

Using Manage 2014.

Object animation and script triggering doesn't work with the standard camera (viewpoint). You have to create a camera in the object animator and time the object animations yourself.

Quote:

---

Originally Posted by dgorsman

Object animation and script triggering doesn't work with the standard camera (viewpoint). You have to create a camera in the object animator and time the object animations yourself.

---

Excellent information. I may be revisiting this thread should this prove insufficient but it gets me in the right direction for now.

I went into Animator, added a blank camera for the scene, set the slider to each keyframe in the animation (there are only 2; begin and end) and captured those same keyframes for the camera. There has been zero change in results however. I do not doubt that a camera is needed in order to export the object movement but just to be clear, I want the vantage point to remain static throughout the animation.

I have attached an example utilizing the exact strategies I am using in the production model. In this case the final position of the beam is at the top of the column but I want an exported animation of the beam raising into position as it will do if run from either Animator or Timeliner/Simulate or by enabling scripts and pressing "r". Note a camera is added to the Animator scene but is not necessary to accomplish what is currently happening.

Don't have the time for detailed review... when you export the animation, there is a dropdown for the source. That needs to indicate the current Animator Scene or Timeliner source; selecting Current Animation is using the viewpoint camera which doesn't reflect any of the object animation.

Quote:

---

Originally Posted by dgorsman

Don't have the time for detailed review... when you export the animation, there is a dropdown for the source. That needs to indicate the current Animator Scene or Timeliner source; selecting Current Animation is using the viewpoint camera which doesn't reflect any of the object animation.

---

That didn't help. If you get a chance to check the attached file and explain what I missed it would be much appreciated.

Quote:

---

Originally Posted by dgorsman

Object animation and script triggering doesn't work with the standard camera (viewpoint). You have to create a camera in the object animator and time the object animations yourself.

---

Still no resolution but it occurred to me that in my issue the starting frame is represented in the export. That is, if it were an issue of the "wrong camera" then I am not so sure I would see the starting frame of the animation since the starting position of the objects exists only in the animation while the ending position is the actual modeled position of the elements.

Just some rumination on what the problem might actually be.

Seems I identified the issue; I was using the Autodesk renderer.

I switched to using the Viewport renderer and got the expected results but at drastically reduced visual quality. One thing is for certain, the Viewport renderer processes orders of magnitude faster. Using the Autodesk renderer it was taking about 2 hours to produce a 20 second 640X480 10fps uncompressed video. The same video at 60fps rendered with Viewport in under 5 minutes.

Why does using the Autodesk renderer not work as expected and how might I get better quality while also maintaining animations in the exports?

PS I tried using Presenter understanding that it should render the highest quality and thus be the slowest render. It alaso took about 2 hours and the animation worked correctly. Presenter clearly uses lighting with shadows and things which were largely annoying for this purpose.

PPS, considering the lack of multi-core support and overall flakiness of this process I will be happy to export frames and compile in a real video editor program. I hear NW 2015 at least gives multi-core rendering support but my IT Dept. hasn't gotten around to getting the license server updated for that regardless of having the license since April.

Autodesk rendering in pre-2015 releases may still be multi-threaded, so you might be running into a peculiar hardware issue in that instance. When different threads are spun off they don't share RAM with the other threads so its usually a considered a good idea to check RAM numbers per core. Each core can end up with a *very* small amount of RAM which makes processing its part as difficult as when you have low total RAM with single-threaded processing. Consider that a hex-core processor (6 cores, effectively 12 with hyperthreading enabled) with 16 GB RAM total would be around1.3 GB per core, with hyperthreading disabled (not very common) it would be 2.5 GB per core. With dual-processor Xeon set-ups and modest RAM numbers you can see the numbers dropping even further. Typical minimum target values I've seen are around 2 GB per core.

And no matter what, I'd still recommend rendering direct to frames. The more common CODECs have some built-in limitations on frame size and run time which can get in the way.