[Bf-committers] Masks are not only for Mattes

[Troy Sobotka]

On Jun 4, 2012 7:02 PM, "Daniel Salazar - 3Developer.com" <zanqdo at gmail.com>
wrote:

> We got to rethink how to present the masking tools not as something
> strictly attached to a footage but something of regular use in the
> compositor for any scene, even fully generated
>
> Part of the solution can be to allow a compositor viewer to feed the mask
> editor, this way we can see the composited results under the masks.
Similar
> to my simple design from more than a year ago that used the image viewer
to
> both edit masks and also load the viewer result

I must say this is a timely post. I agree 100% here.

I will say that I think the new spline code is wonderful at the core, and I
preface this with tremendous thanks to Peter and Sergey for the hard work.

However, it would seem that a spectre is looming on the horizon.

I would humbly call into question our design choice to now tag much of this
new rotoscoping functionality into the newer Movie Clip Editor space. What
is the Movie Clip Editor space for? Is it for tracking? Now rotoscoping?
What is the internal logic of this new editing space?

If we look at grading alone, we can already see how the decision to create
a new working space will adversely impact our workflow.

As Daniel pointed out, mattes have far more uses than rotoscoping, and so
too does tracking.

How can we create a simple point and track a grading region to it? How
would we be able to isolate a region for a particular grading effect? Etc?

It would seem to me that a large part of this stems from a lack of
interactivity within the UV Image editing window, much like Daniel
suggests. We need look no further than the Backdrop view toggle to see how
the user interface vision wandered.

We have already seen how we have had to resort to rather unfortunate
workarounds via a "Tracks to FCurve" button in the Movie Clip Editor, where
it would seem to beg for a tracking node.

I cannot help but wonder if a stronger focus was placed to make the UV
Image Viewer space interactive would dissolve the current design problems
looming on our horizon and create a greater cohesion within the workflow.

It would seem like a wonderful juncture to perhaps analyze our design
choices and consider future limitations at this juncture...

With utter respect,
TJS