Jump to content
Sign in to follow this  
aesnakes

Camera and Camera Animation export to Compositing app

Recommended Posts

Hello,

I am coming from a maya workflow and like many people here just have come to find Houdini as the greatest thing ever. However I need to achieve a few things that I was doing in my old workflow.

So I am doing some camera projection and mapping inside a 3D app and then exporting to either After Effects or Nuke/Fusion for compositing of 2D Characters. I need to import the Camera Data with Animation and a placement Geometry so i can find the proper positioning for the 2D Characters. this is so the ground does not slide underneath their feet.

Now I have achieved this in maya-nuke and maya-after effects workflow using the MoCon_3_5_5 exporter. Please tell me there is an easy way to solve this in Houdini. I dont want to go back to maya PLEEEAASE NOOOOOO. lol

Share this post


Link to post
Share on other sites

for After FX CS3 (CS2)....

In Houdini:

...Windows-> Python shell

...type something like: execfile('c:/aecam.py')

select the camera and press enter.

In After FX:

File->Scripts->Run Script

jump to camera comp and adjust scene scale by scaling "Scene scale" null object

aecam.zip

Share this post


Link to post
Share on other sites

Did you write this? or get it from somewhere. Great job I wanted to thank you.

Does anybody know like the other plugin how I can put placement goe and export to after effects to know where I would place the characters in 2.5d space. I know it made white planes with the exported geo and I just had to replace the layers and the characters where hitting their marks without feet sliding etc.

thanks again!!

Share this post


Link to post
Share on other sites

Yes...this is my script...

i have not finished (polished) version which exports objects also (as AE null).

I will post it tomorrow

Share this post


Link to post
Share on other sites

Hey AndrewVK,

Have not tried your script yet... but this sounds great!!! It will be a very cool thing to export the Houdini camera to AE... and the objects as AE Nulls...

That was the last thing to make Houdini part of the motion graphics pipline :)

Share this post


Link to post
Share on other sites

Hi,

I want to transfer my camera and some scene information as well from houdini to Nuke so far this is what I have found out:

*) this thread has been really useful:

http://www.vfxtalk.com/forum/boujou-nuke-ummm-t10336.html

In nuke 5 you can create a new camera and add a quick expression to some of its channels as a test to see what is generated by for the chan files. In the settings for the camera you can import and export .chan files. The chan files are just ascii files so you can open them up and have a look inside.

The ones that Nuke requires are in the form of: FRAME TX TY TZ RX RY RZ FOCAL

If you look in the plugins folder of the installation of nuke you will find the following script: import_chan_file.tcl. If you open that up you get a little help file explaining some of the requirements. There are various other formats supported through tcl scripts. I first tried running these scripts, but then noticed that the option is already there on the camera for importing & exporting chan files. It is good to know where to find the scripts for other formats as well. There is more information in the nuke user guide on tcl scripts and adding menu options and changing Nuke's Ui.

In houdini you can create a new chops network and fetch the channels from your camera, this will give you tx ty tz rx ry rz focal. Now all you need is frame. So create a new channel (you can call it frame) and as a value give it $F.

Merge this with the fetched data from the channels. The order of the merge is important! First the new "frame" channel, then the fetched data.

Add a null underneath it (this makes sure all the above is definitely cooked/calculated, this might not be necessary), right-click on it and select: "save data channels". Make sure you write out your file with the ".chan" extension. Now just use that file in Nuke.

This works, but currently I am having a "problem" because chops only fetches the data in local space and my camera is parented under a null which has been scaled up. ( I had to scale up my scene for some dynamics I am using) I am now trying to get the global/world position data into chops and from there into nuke. Does anybody know how to fetch the position data from a child node in world space?

I have considered using an object merge in a new geometry node, but don't know which point represents the pivot of the camera exactly. (I can do a similar kind of parenting operation of the camera in nuke, but this gets a bit messy as you would need the scale value.)

Hopefully this was helpful. Please post back if you find something that might help.

Edited by pclaes

Share this post


Link to post
Share on other sites

Here is dirty version...i still learning python...

I have no time to test it well.

Looks like it works properly...but who knows...

You can export: objects, subnets, nulls, lights, cameras

Exported properties (can be animated):

Pure Transform,

Camera Zoom,

Light Color, ConeAngle, Intensity.

I have no idea how to change light type in AE via script...

Just switch it by hand

HoudiniToAE.zip

Share this post


Link to post
Share on other sites
Here is dirty version...i still learning python...

I have no time to test it well.

Looks like it works properly...but who knows...

You can export: objects, subnets, nulls, lights, cameras

Exported properties (can be animated):

Pure Transform,

Camera Zoom,

Light Color, ConeAngle, Intensity.

I have no idea how to change light type in AE via script...

Just switch it by hand

Thanks Andrew for that. Very handy.

cheers,

sy.

btw. It should be put to the "Others" subforum and/or added to wiki.

Edited by SYmek

Share this post


Link to post
Share on other sites

Yeah Thank you VERY MUCH. This is awesome and totally unexpected. :)

Now the last thing I need to learn is the Camera to Nuke export. I know people have successfully done it. If you know some good resources please post them.

Share this post


Link to post
Share on other sites
Sounds great, Andrew.

Have you considered making this into a SOHO-based ROP?

Apprentice limitation:

Any other SOHO outputs other than the shipped Mantra ROP are not supported. :tumbleweed:

Share this post


Link to post
Share on other sites
use the object merge and point to the camOrigin node in the camera

Hi Michael... what happens when you have a "Look At" object?

I mean... In my case the camera is looking at a null... so the position is ok with the camOrigin point being read into the chops... but the rotation is not correct since the camera is looking at a null?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×