EVE Probe

I would be many happies

Alternately it would be cool if probe could output scenes in it's render layers with alpha channels, so if you want to put the ships on a different background you're destroying ship edges with a key :P

also, having a z-buffer output would make me a very happy man. because then I can accurately calculate depth of field in nuke :D

I understand that you guys might be averse to breaking apart renders into diffuse/spec/reflection/emissive/nebula/effects/ambient occlusion/zdepth etc as the point of probe is primarily for QA

There is no python coding required, just very basic YAML to set up the scene, which has been documented in probe. And there is no videos out from it yet because it only was released 2 days ago, and is effectively a bleeding edge internal tool, it's not feature complete yet. CCP are just testing the water to see what people can do with it, or even if people want to use it.

If you open up C:\ProgramData\CCP\EVE\SharedCache\probe\docs you will find 4 html pages with very simple to follow instructions :)




My scene relies on the custom camera move so I'll have to wait till its supported in the red files :)

...unless you're happy for me to send you an export of the camera to turn into a gr2 ;)

Sure! Here is the example animation after I've baked keyframes into it:

https://www.dropbox.com/s/sedikftp37mudf2/example-baked.dae?dl=0




Darn.

I was hoping that just doing this would work:

  - [camera, ((camera_name)), 0, 0, 0]
  - [select_camera, ((camera_name))]
  - [bind_matching_dynamics, '((scene_with_camera_name.red))']

When you get around to implementing it consider supporting curves on the camera's field of view.

Optical focus would also be interesting, but I don't see yet how to translate that into Collada (and previewing it in Blender isn't possible without a render).

edit: I played with the idea of driving the Z-scale of the camera (a largely irrelevant property) with the focus distance property, so that might be a way to export a focus curve.

Also, did you know that dae-to-red will accept a zip file? I tried breaking it with random files, and the first one I used was the .zip of the source code (which included the example collada files). I did a double take when it actually created an output. Parsing .zip files seems to be a feature of the collada package.

This is the way they have been creating trailers inside CCP for nearly a decade.

It may answer the question for you why they do not create a lot of videos using this method and partially why it takes a team to create them. It certainly answers the question as to why a lot of the trailers originally had lots of ships moving in a straight line.

I said this in the speakers corner thread; we are a very long way away from having a drag 'n' drop NLE system for cinematic/machinima purposes by the playerbase.

A side note, Ian Chisholm spent a week at CCP learning this - and I can only speculate the first thing he learnt was creating a one camera scene of one object not moving, probably followed by a brofist or high-five.

Like chess, this will take time to learn and years to master - waiting for a better solution/workflow is unrealistic because, frankly, if there was a better way, CCP would have deployed it internally 10 years ago and saved themselves 5 employees @ $X cost per year.

AK

PS: the DUST trailers were worse than this, they had to use UE3 and composite the footage with the EVE/Jessica renders. Now, that was a nightmare of monumental proportions. I cannot give that team enough kudos for what they create.

Ok, ship types, I get. What about planets? Is there a scheme to them as well?

Edit: Also, I can't seem to get the frames to capture using this command:

this would be so cool.

CCP could start up with certain set of actors that are taken in for recording from a live scene, over time the actor amount could be expanded to all active actors on screen. Simulators with a lot of actors are huge in my logistics field, no idea how EVE compares.

Make sure that your animation is keying on all XYZ channels.

Can you share your .blend file?

http://www.pasteall.org/blend/

Just started playing with this, great stuff CCP. Really happy to be able to play with stuff like this.
And many thanks to Valeo Galaem for making the DEA to RED process super easy.

One thing I'd really want to be able to do is export camera animation paths from blender. Above it was mentioned this can currently be done with gr2 files? Anyone willing to give me a short run-down on how that's done, and integrated in the scene file?