We’re currently working with conventional exr’s (single file per channel) for most of our project, however we’re planning on using the exr camera data that can only currently be passed via a multichannel EXR for comp.
So, really two requests:
1. Would it be possible to enable the image format global to be overriden via right-click for render layers?
2. Would it be possible to pass exr camera header data in conventional (non-multichannel) EXRs?
An update to this: I’ve found that it’s possible to override the format text field on the vraysettings node to force a single render layer to do multichannel, so you can scratch request 1 unless you have a burning desire to make things more “Maya-like”
Number 2. is possible too in the latest builds - if you click on the Image format options button, there is a field called “Extra Attributes”; these are of the form int_attr=53;float_attr=3.14;vec4_attr=(1, 2, 3, 4);...
Values can be either integer or floating-point numbers, or lists of numbers - for example vec4Attr=(n0, n1, n2, n3) - for vectors or matrices.
I’d realized that you could manually insert a single frame’s values into the exr header, but I’m interested in each frame’s animated camera data being written per exr when rendered out to the farm (and dealing with stereo renders, so it’s not always the same camera).
I’m assuming you’d have to make a pre-frame MEL script that updates the header data in the vraysettings node before each frame is written? Or can you just plug variables in? There’s not much in the way of documentation on how those extra header attrs work.
Thanks for looking into that. It would certainly save us quite a bit of grief with regard to workflow/disk space/messy post-render hacks if we could just hand off our beauty pass with the cam data in it.