Cineversity

One thing I have always had an issue with is exporting or saving a rendered still scene from C4D with all the mask data that I might possibly want/need. Specifically, If I have an object in a scene file, I would like to have multiple masks so I can edit those different areas in a 2D editor like Photoshop. I know about Object Buffers and how to apply and render with them. However, I can only get an object buffer to mask an entire object and not different portions of one object.

I am including a link below to simple scene file of a cube with 2 extrusions on a background plane. I have an object buffer assigned to both the cube-like object and the plane. I also have a saved set selections for the polygons on the large extrusion and on the small. What I would greatly appreciate knowing is how a still render can be performed so that there will be four individual alpha masks in one rendering execution.

Why can’t the set selection be used to create masking or alpha channels at render time? I have asked this question on another user forum and the recommendation I got to create duplicate portions of an object seemed inelegant and/or awkward. I am vaguely familiar with the work flow procedure of setting up very distinct saturated color materials that can be applied to selections of an object so that when rendered, selecting those areas in another imaging editing package is easy. If this is the recommended workflow perhaps there are some resources here at Cineversity that might show me how to set up such a render?

http://we.tl/V5C0COGBWQ

Hi Pablum Picasso,

The Problem why this option is not given is not clear to me.

With the Take System this will be a simpler task, as you can set up a “version” on top of all that takes the selections into account.

Selections as in “all polygons” are covered by the Compositing tag. Then you have the option to use the same color channel or not, which makes an total of eight (___, RGB, RG_, _GB, R_B, R__, _G_ and __B [_=0% and the e.g. R=100% for red] To work with gray values might work, but in the moment motion blur is part of the game everything will be trickier) This is the cleanest way. The RGB has less limitations. ;o) With clean RGB channels, no “Key” needs to be pulled, just use/combine the channel information…

The material on the most right, the real material, can get an X in the selection (or any other letter not used to name a selection) as shown in the image, which allows to leave the “mask” material on the object.

Besides that, I need to explore more the Variations Shader for this target. I’m not ready to say it works for most cases, it is to get the idea, but solid information needs tests. ;o) (see image)

There is always the Suggestion-Page {the more ask for it, the more it might happen
http://www.maxon.net/support/suggestions.html

All the best

Hi Doc,
I made my suggestion to the Maxon team (I also requested a revision of the BP UV edit tools)

Have you seen this:
https://vimeo.com/117248701

It doesn’t address the multiple selections of one object but it does make quick work of making unique masking info for multiple objects (I think I am saying that right).
Thanks!

Hi Pablum Picasso,

Thanks for making the suggestions.

To give you a little idea why I have some concerns with selection tags: I have set up a plane and demo one problem if some selections have multiple selection tag records.

Yes that can be solved with channel operations in PS or Ae, NUKE, FUSION, etc. but there is always a situation that might interfere. Yes, a Buffer that reads the Selections Tags might be good, but it causes other things, as mentioned above, e.g., the Motion Blur stuff. Case dependent—I agree. To manage information is your core request and I’m with you on that.

Here is a one minute clip with the R17 options to do that on the fly, especially useful with “Take System” (not included here)

As a side note, I have used the RGB channel stuff in the ‘90s, but not really since. I hope the Take System will help here with other new options, as mentioned.
Besides that, the Multi-Pass UV channel output can be organized for a key option as well, but I know, most people avoid UV. UV has no Anti Alias, so it needs to be oversampled, four times the render-size, but it is really fast. No lights, no reflections, no GI nor anything…

Enjoy

https://www.amazon.com/clouddrive/share/K4jQpheZXb1For0EhHW8K11h33Z22I9ZTwWdUfGYPul?ref_=cd_share_link_copy

Scene file is R17, demo of R17 is available!

Thanks Doc,

You have given much for me to consider and I want to go through it thoughtfully and thoroughly.

I think the release of R17 is right around the corner and I have already paid for it so I may show some patience for once and wait for the announcement.

The takes system looks like a time saver for certain (and the variation shader looks great too), and I don’t think there would be a lot of advantages with more masking capabilities of selections, especially with the other issues you bring up. Maybe I should take back my request on the Maxon site and ask for something else… like a pony?

I really am grateful to have your thoughts on these things.

Thanks for the reply, Pablum Picasso!

Yes! Get your copy of R17 and explore! Great that you have already everything lined up. Fingers crossed that it will arrive soon!

MAXON has a good idea what people like to have, and they certainly listen. I have read this kind of request from time to time, which means that there are more people in need of more control, but perhaps never voice it.
What the answer is finally, that will show the future, as usually one idea pushes the next. We will see. :o)

I have recorded another of my “one minute clips”. It shows how simple one can set up an alternative UV-Tag, and how to employ it for an alternative render. In short, the UV space is represented by Red and Green, U and V: in other words—two gradients—90º cross and mixed. So each area on the surface has a unique combination, if not used with overlapping “uv-polygons”.

The idea is to move the groups (polygon selections) to a different color combination. The uv polygons can be very little for that.

Keep in mind that you must oversample for that pass, but if only the needed parts in the scene are working, as mentioned above, these passes a fast to render. Then sample it down to the same size of the other footage, to get a similar Anti Alias.

The Object Buffer of an Object works as Frame, and the UV inside then as separation. Just an idea, and I hope I have made you aware that it is not without limitations ;o)

Perhaps, this idea pushes another, and you find your very own workflow. Perhaps share it then ;o)

Here is the clip:

https://www.amazon.com/clouddrive/share/KwFwusw3JfUPXttxC4JWlUoXHdM6AZB8IbQsZtKu2Wk?ref_=cd_share_link_copy

My best wishes

R17: Mask - Variation Shader - Take System

I was just curious how the idea that I have introduced above, would work in conjunction with the Take System. Since I have expresses some concerns already about double selections, I leave them out here.
I explore this currently, so I will report back if I find some other ideas or concerns. ;o)

Here is a one minute clip and the file of course [R17]

You can download the clip:
https://www.amazon.com/clouddrive/share/4BgjkVaea02TFIiYAiSkBoQSCVGZOvNPSv72hQpKupg?ref_=cd_share_link_copy

Enjoy