Just took DF for a test drive, first impressions and a few questions

#1

Wow…

This is an extremely powerful tool. I’ve used After Effects, Combustion, and XSI’s compositing tools extensively and am really impressed with DF. This is an amazing program… I’m very surprised that it is one of the lesser used compositing apps on the win32 platform. For it’s price range it’s pretty much the only compositing application with scripting capabilities, which I think is an essential feature.

Very fast, very stable, 3d particles, great image processing tools, clean timeline, awesome flow view, and a unique interface. It also allows for a great level of control. Super file support as well.

The UI and workflow was a bit baffling for me, but only for the first 5 minutes or so. After that it almost becomes second nature.

Now for the questions.

#1: is there some sort of spreadsheet editor for changing shared attributes? i.e. if I wanted to enable alpha-premultiply on all of loaded footage at once, etc…

#2: Concerning depth. I use XSI at work, and unfortunately, XSI’s z-depth files are not rendered anti-aliased. This leads to undesirable artifacts when using the depth blur. I sometimes also render out custom depth passes for purposes of focus pull and other reasons. So my question is, can I specify alternate footage that I want to be used as the Depth for my depth blur and 3d fog? Namely, the custom pass I rendered out of XSI. This would be a HUGE bonus. I can currently do this with Frischluft’s Lens Care DOF, specifying the layer I want to be used as the depth channel.

(seriously considering purchasing this awesome package)

That’s it for now!

Thanks

#2

Ok, figured out how to swap in a Z-channel with the channel boolean, awesome tool by the way :slight_smile:

One problem though, I’m still getting some artifacts for some reason, even though my Z-channel was rendered with nice anti-aliasing. Any suggestions?

When using the Boolean tool I left the mode at copy and just remapped each channel to itself, then enabled aux channels and mapped my FG luminance to Z-depth. Am I doing anything unsavory here?

Thanks :slight_smile:

#3

Another couple of quick questions,

Is there a similar function to “commit to disk” a la combustion? And then setting up a switcher node to switch from the original tree to the committed footage?

Can you group/nest nodes? like put selected nodes into a container? And then open that group/container at any time? Would just help to organize and maintain unruly flows.

Thanks again everyone.

*Edit: figured out grouping, I guess I should read the courseware before posting my questions :slight_smile:

The questions about a common attribute spreadsheet editor and committing to disk still stand though :wink:

#4

Thanks for the tip on using alternate source for Depth Blur and Fog. I was just trying to figure that one out. I have to go try that now. If you find out another way to do it, don’t hesitate to post it. (Just letting you know your not just talking to yourself here).

#5

Be sure to go through the courseware… It should cover most of the basic bits that you need to know about DFX+.

You can use the Cache to disk option to commit a branch of a flow to disk. This will (up to 4.02) lock the branch, and disable any editing of it. It is possible to unlock it, but this can open yourself to headaches later unless you really know what you are doing. I recall writing something up while working at eyeon, so take a look at their site, I think it’s posted somewhere.

There is also a spreadsheet option in the timeline view, but it doesn’t currently offer access to things like motion blur settings unless they are curently animated.

Also, be aware that an anti-aliased z-depth is wrong. By antialiasing you are essentially pushing the edge of objects further or closer to the camera as they blend into the surroundings.

If you have access to the RPF file format it has a layer in the format set to pixel coverage, and this effectively combats the problem of z-depth compositing and DOF.

#6

Hi Jay,

Thanks for the great response. I’m using XSI exclusively right now, and it doesn’t offer any of the file formats like rpf, rla, iff, etc… The only way to depth out without setting up a seperate pass is with .pics and .zpics… which works fine for certain things. But sometimes I want to create a custom depth pass. So I should have no antialiasing when I render this custom pass out? perhaps a setting like -1, -2 for sampling? Ant ideas for the filtration method? Perhaps box or mitchell.

One feature I would love would be a keyframe divot at every animatable paramater like in XSI and DS… makes keying fun :slight_smile:

Thanks again,
Brad

#7

Fusion/DFX+ should automaticly load the zpic portion of the pic sequence in the background and apply the z information for use in DF.

#8

Hi Jay.

Problem is, when you create your own custom depth pass n XSI, you cannot make them in the zpic format, as in a floating point info for the diffuse pass. XSI just can’t do it.

#9

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.