Open EXR

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

THREAD CLOSED
 
Thread Tools Search this Thread Display Modes
  06 June 2013
You could try the following if you have trouble with cinema's generated z-depth channel.

Create a position pass(remember to activate post effects in multi pass, since this is the chnnel this data will be written to) and set the space to camera. Dont forget to use 16/32 bit per channel or this pass will be useless.

When it's done rendering you go to your comp app and extract the Blue Channel from the Camera World and use it as your new depth channel. This way you get an depth pass which has no AA and absolute depth values.

Don't know how hard it would be to do this in AE, since i haven't worked with it for ages but it's really fast in nuke.

EDIT: LucentDreams made the same suggestion allready, I am too slow

Last edited by Tyrus86 : 06 June 2013 at 07:03 AM.
 
  06 June 2013
One thing that's always confused me about the depth pass coming from c4d is that by default I get what I would call a depth of field map. So rather than a depth gradient (near is black, far is white), I get a white, black, white image where black represents camera focus.

To solve this I've always jiggled the camera blur settings to give me something similar to a depth gradient.

Now it appears that the position pass would solve this completely, thanks for pointing that out.

But was there another solution hiding in there all along that I missed?

Thanks,

Will
__________________
My portfolio site

@WillMacNeil_VFX
 
  06 June 2013
Hey guys, up until now I've been unaware of the different multilayer options in RPF format (never used it).
I can't seem to get Frischluft DOF to recognise the z depth info, although I see AE does with its depth effects. I presume it is possible though?
 
  06 June 2013
Originally Posted by Kokosing: But was there another solution hiding in there all along that I missed?


i always thought that just activating a single DOF map blur check box gave you a single black-white gradient in the depth pass?

i'm assuming that is what you meant, if so, then your method is the correct way to do it.

personally i just use an environment object with fog or ambient light with fall-off depending on my mood.

although the position pass to get y fall-off sounds interesting.

cheers, simon w.
 
  06 June 2013
"It will be interesting to see if C4D or VRay start supporting Deep Images, I'm not holding my breath"

the very newest vray core supports OpenEXR 2.0 already. so we will be able to support it if rendered via vray frame buffer or standalone you can save it. to save it via c4d picture viewer Maxon needed to add support for it though.

Stefan
__________________
V-Ray on FB:: http://www.facebook.com/vrayforc4d
http://VRAYforC4D.net
http://www.3dtools.info
http://rhino.io
 
  06 June 2013
Originally Posted by sneather: Sorry to disagree, but the z-data is not the same as the depth image matte generated by C4D.

The simplest way to point this out, is as follows:

Render a .RPF image with z-depth data enabled in the options. Import into AE. Apply the basic "Depth matte" effect, and click anywhere on the image. In the "info" palette, you'll see X, Y but also Z position data listed.

Try the same thing with the basic grey-scale "depth" pass from C4D, and you'll see what I mean. Nothing. The data doesn't exist.

There are filters (upon which I rely) that must have that real z-depth data. Without such, they simply do not work.

So, unless I'm missing yet another render option in C4D, the OpenEXR test files I'm generating do not have that necessary z-depth data.


The grayscale data from the depth pass provides the Z data to Lenscare, so the only effective difference is the aliased edges, which cause problems in some cases. Per is correct in saying that RPF is providing the same information, but without AA. Kai is correct in saying that the main difference is in native AE's inability to translate the depth pass into Z depth info. I think you are imagining that the RPF file includes actual geometry data, which it does not.
 
  06 June 2013
Originally Posted by LucentDreams:
That said you guys have learned how to use position passes right? Set it to camera and use the Zpass and a non antialiased Zdepth with amazing precision, and get both Z and why depth gradients simultaneously. Far more useful if you ask me.


Can you explain how to do that in AE?

Edit: never mind. I needed to adjust the scale setting in the position pass. Works great.

Last edited by AdamT : 06 June 2013 at 03:31 PM.
 
  06 June 2013
Originally Posted by LucentDreams: that fact that pixel sampling the zepth pass doesn't give you a zdepth value is really more of a flaw with After effects than any thing else, although C4D's fairly plain storing of channels and most image formats being layer based doesn't help either.

For example in our software at work if we sample the zdepth we get a Zvalue, but the zdepth is a float image. Thing is that the exr knows it's a zchannel, not a black and white image in an rgb layer or a generic alpha as photoshop stores it in a psd. Now how can it get that accurate Zspace informaton from an image. Because a float image is simply a agrid of cels that each store a value, nothing more. The fact that we decode these pixels into red green or blue is a rendering process we do to that data. We can store anything we'd like in there. SO yes a Zdepth image as a float is capable of everything that the z data in rpf is. The antialiasing does screw up the data at edges, which is a concern, and something C4D should allow one to rectify but it isn't a game stopper by any means.

That said you guys have learned how to use position passes right? Set it to camera and use the Zpass and a non antialiased Zdepth with amazing precision, and get both Z and why depth gradients simultaneously. Far more useful if you ask me.


Thanks for the detailed explanation. I only wish I understood more of this.
You final paragraph has me scratching my head. Can you please break it down a little for my feeble mind? Where are the options for position passes in the render output settings?
 
  06 June 2013
Originally Posted by AdamT: The grayscale data from the depth pass provides the Z data to Lenscare, so the only effective difference is the aliased edges, which cause problems in some cases. Per is correct in saying that RPF is providing the same information, but without AA. Kai is correct in saying that the main difference is in native AE's inability to translate the depth pass into Z depth info. I think you are imagining that the RPF file includes actual geometry data, which it does not.


Okay. Thanks.

How would you use AE's Depth Matte effect without the z-data as interpreted by the .RPF file? That's a must for me.
 
  06 June 2013
Originally Posted by sneather: Okay. Thanks.

How would you use AE's Depth Matte effect without the z-data as interpreted by the .RPF file? That's a must for me.


Unfortunately you can't, AFAIK, but only because AE doesn't have the function built in, and I'm not aware of a plugin that would do it. Kind of surprising because I don't think it would be that difficult. You can effectively do it by comping Luma Keys -- one using key out brighter and another using key out darker. Not very practical, but possible.
 
  06 June 2013
Originally Posted by AdamT: Unfortunately you can't, AFAIK, but only because AE doesn't have the function built in, and I'm not aware of a plugin that would do it. Kind of surprising because I don't think it would be that difficult. You can effectively do it by comping Luma Keys -- one using key out brighter and another using key out darker. Not very practical, but possible.


Yeah, that kind of extra work (just to get the depth matte to work) is out of the question for me.

There's definitely an issue with the way C4D writes the OpenEXR file. It's not a limitation of AE. I know this, as I'm playing around with the demo files from ProEXR, and the included still EXR image of the X35 jet works perfectly for me in AE. Lenscare works perfectly with the file, and so too does the Depth Matte AE plug.

I'm guessing that test file came from Maya or Max? Who knows. But clearly, not from C4D...
 
  06 June 2013
Originally Posted by sneather: Yeah, that kind of extra work (just to get the depth matte to work) is out of the question for me.

There's definitely an issue with the way C4D writes the OpenEXR file. It's not a limitation of AE. I know this, as I'm playing around with the demo files from ProEXR, and the included still EXR image of the X35 jet works perfectly for me in AE. Lenscare works perfectly with the file, and so too does the Depth Matte AE plug.

I'm guessing that test file came from Maya or Max? Who knows. But clearly, not from C4D...


Okay, some success here. I can't figure out how to set it up in render settings directly, but if I render to EXR with no compression, and then save the file from the picture viewer, I get a layers option in the save dialogue. If I check that and then resave the file, AE sees the layers and I can extract them with the EXtractoR plugin. That way I can get the position pass and/or depth pass data to use with Lenscare. Still no depth matte, though.
 
  06 June 2013
Thread automatically closed

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
__________________
CGTalk Policy/Legalities
Note that as CGTalk Members, you agree to the terms and conditions of using this website.
 
Thread Closed share thread



Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright 2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 12:00 PM.


Powered by vBulletin
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.