NeilMyers wrote:Kruno -- I saw your detailed explanation about the "feature" of Planar Tracker not remembering stuff. I wasn't clear -- it was CornerPositioner (and, later, the poly mask tool) that was "forgetting" my keyframes. Is this also expected behavior for these tools?
I see. Well The polygon mask and B-spline tools have by default option turned on to animate key frames based on any changes you make the to mask. This is done so you can start using the tools for roto right away and get instant animation. Anytime you animate something by changing a parameter of points key frame is auto generated on the timeline and other keyframes are interpolated between the ones user adds. To get smooth animation with as manual keyframes as needed.
This is default behaviour, but it can be changed. Or keyframes deleted. Weather you did that or not, I can't say.
"A B-Spline mask is identical to a Polygon mask in all respects except one. Where Polygon masks use Bézier splines, this mask node uses B-Splines. Like the Polygon mask tool, the B-Spline mask auto-animates. Adding this node to the Node Editor adds a keyframe to the current frame. Moving to a new frame and changing the shape creates a new keyframe and interpolate between the two defined shapes. The B-Spline node can be used to generate a single smooth spline shape or combined with other masks for more complex shapes"
I personally prefer B-Spline over Polygon mask because it takes less points to roto an object. Less points, means its easier to make changes to less points, Other than that, its identical.
There is tones of options for each, but "losing" keyframes is not what I'm used to. I can only assume you might have not selected the correct node or something simple, because its possible to have one node selected and see the sliders of another and yet watch in viewers other nodes. Useful for complex process, but if you are new to Fusion, maybe its just that simple and you had one node selected and was thinking you had the correct one.
Anyway, you can see here that the B-Spline is not even connected to the main footage but I can work on the image and also you can see that the animation of keyframes is turned on by default. This can be unexpected if you are not aware of it, so maybe you just had the wrong one selected.
- sshot-3978.jpg (66.95 KiB) Viewed 1263 times
You can always check the spline editor to see if there is any keyframes used by the tools.
- sshot-3979.jpg (166.08 KiB) Viewed 1263 times
As you continue to work on your footage and roto, each time you change the shape to adjust it, new keyframe is added, in-between there is interpolation for smooth movement. You can change it, but that is default behaviour.
- sshot-3980.jpg (145.7 KiB) Viewed 1263 times
NeilMyers wrote: I will push back on trying to make Fusion work like After Effects. Not at all! In general I LOVE how different DaVinci Resolve is. So far, these differences have ALL been for the better (Color is the best example). And I don't mind working to learn the differences (I have been reading the manual, and the tool reference, and watching many videos).
My frustration is not that it is different. Rather, I am frustrated that it isn't working the way I *think* it has been described. Is it something I am doing wrong? I sure hope so! That would be FAR better than a bug.
Thanks again, I appreciate your generous (and patient) responses.
Well, its good to hear you are willing to learn. Yes, it must be just how you expect the tools to work rather than how they are expected to work. Once you get used to it, I think your frustrations will be gone.
NeilMyers wrote: And, here is a link to a Word document that contains the six page description in case there was something I did that could have been done in a better way.
1 Add PlanarTracker tool (output of my video footage goes into background of PlanarTracker node.
Good.
2. Drew polygon on my footage to define the area onto which I wanted to post the still:
Good.
...........................................
3. Set-option in PlanarTracker tool
Good.
Set and go buttons are important. Set simply defines the reference frame or where you can use other options to reference it, and go takes you to reference frame. Before you start tracking, other than being on frame 0, just click set button. It seems you have done that. Good.
"The Reference Time determines the frame where the pattern is outlined. It is also the time from which tracking begins. The reference frame cannot be changed once it has been set without destroying all pre-existing tracking information, so scrub through the footage to be tracked and choose carefully. The reference frame must be chosen carefully to give the best possible quality track. You choose a reference frame by moving the playhead to an appropriate frame and then clicking the Set button to choose that frame."
Tracker
There are two available trackers to pick from:
— Point: Tracks points from frame to frame. Internally, this tracker does not actually track points- per-se but rather small patterns like Fusion’s Tracker node. The point tracker possesses the ability to automatically create its internal occlusion mask to detect and reject outlier tracks that do not belong to the dominant motion. Tracks are colored green or red in the viewer, depending on whether the point tracker thinks they belong to the dominant motion or they have been rejected. The user can optionally supply an external occlusion mask to further guide the Point tracker.
— Hybrid Point/Area: Uses an Area tracker to track all the pixels in the pattern. Unlike the Point tracker, the Area tracker does not possess the ability to automatically reject parts of the pattern that do not belong to the dominant motion, so you must manually provide it with an occlusion mask. Note that for performance reasons, the Hybrid tracker internally first runs the Point tracker, which is why the point tracks can still be seen in the viewer.
There is no best tracker. They each have their advantages and disadvantages:
— Artist Effort (occlusion masks): The Point tracker will automatically create its internal occlusion mask. However, with the Hybrid tracker, you need to spend more time manually creating occlusion masks.
— Accuracy: The Hybrid tracker is more accurate and less prone to wobble, jitter, and drift since it tracks all the pixels in the pattern rather than a few salient feature points.
— Speed: The Hybrid tracker is slower than the Point tracker.
In general, it is recommended to first quickly track the shot with the Point tracker and examine the results. If the results are not good enough, then try the Hybrid tracker.
Motion Type
Determines how the Planar Tracker internally models the distortion of the planar surface being tracked. The five distortion models are:
— Translation.
— Translation, Rotation (rigid motions).
— Translation, Rotation, Scale (takes squares to squares, scale is uniform in x and y). — Affine includes translation, rotation, scale, skew (maps squares to parallelograms).
— Perspective (maps squares to generic quadrilaterals).
Each successive model is more general and includes all previous models as a special case.
Translation, Rotation, Scale seems to be good for drone footage. If the perspective does not work.
...........................................
4 - Pressed “Track to End”, and watched it create the tracking data. NOTE: It got wonky as the drone footage got to an oblique position with a giant tree blocking the plane I was focused on:
By wonky I mean the bounding area was jumping all over the place So, I stopped the operation and deleted the tracking data after the frame just before this started
Yes. Its generally good idea to watch the footage before tracking for any possible occlusions like the tree. If you have one, you can create occlusion mask to tell the track not to pay attention to the thing you have designated as occlusion mask. Occlusion mask can be anything, from simple shape drawn with the polyligon mask to using color selection keyer and using that as a mask.
Depending on the footage, sometimes if you have a red car driving trough the forest, its best to color select the red car and make it only thing to track, the car paint. That is automatic exclusion of everything else.
With footage like some random stuff on a building and a tree in front, probably you will have to rotoscope a tree with a lose garbage mat, B-Spline or Polygon tool will do fine, and than hook that mask to Planer Tracker in the occlusion mask input.
- Occlusion Mask: The white occlusion mask input is used to mask out regions that do not need to be tracked. Regions where this mask is white will not be tracked. For example, a person moving in front of and occluding bits of the pattern may be confusing the tracker, and a quickly-created rough rotomask around the person can be used to tell the tracker to ignore the masked-out bits.
— Effect Mask: The blue input is for a mask shape created by polylines, basic primitive shapes, paint strokes, or bitmaps from other tools. Connecting a mask to this input limits the output of the Planar Tracker to certain areas.
Also "The point tracker" option possesses the ability to automatically create its internal occlusion mask to detect and reject outlier tracks that do not belong to the dominant motion. Tracks are colored green or red in the viewer, depending on whether the point tracker thinks they belong to the dominant motion or they have been rejected. The user can optionally supply an external occlusion mask to further guide the Point tracker.
So if you have anything other than a footage with no distractions, I would choose point tracking, to ignore any disruptions. If its not enough, use occlusion mask you made yourself. Hybrid Point/Area is more steady way to track with less micro jitters, but needs clean source to be successful.
........................................
5. Next, I added my image, followed by a color correction node, followed by a corner position node
This step is not needed since all that is done inside the planar tracker node, with the corner pin option. Making corner position node redundant, and you would not use it like that anyway. Its more of a node to use on static footage when you want to change perspective of something and put it back into perspective after a change. Also it can be used as Planer Tracker cornor pin option when using ordinary tracker, or more precisely four trackers, and you have enough data to corner pin something, but unlike planar tracker, ordinary tracker, has no option to corner pin, so this node you used could be used. With planer tracker its actually counter productive, since its not as efficient.
6. I then created the Planar Transform using the“Create Planar Transform” button:
7.I then placed the Planar Tracker between my media segment and the footage segment. Like this 6.I then created the Planar Transform using the“Create Planar Transform” button:7.I then placed the Planar Tracker between my media segment and the footage segment. Like this:AND ... it works! Why is this working today, and not yesterday? I think because I had done some serious video watching last night to clear up some misconceptions I had about where to place
While its possible to make it work like that. Its like using bazooka to kill a fly, and not worrying about collateral damage. Its just not needed. Corner pin option inside the Planer tracker will do all that plus more. You have blend modes and merge node all in one, plus tones of other option.
One thing about Fusion to keep in mind is that almost every node, is actually a whole toolset that has more options than some applications. So usually its a multitool. Paint node for example is more versatile than whole of Photoshop tool set or about that much. Its crazy what you can do with one node.
By the way, about planar tracker. If you need to do lot of work on the stablized area. Another way to work is to use steady/unsteady workflow. You track with your planer tracker. You put it to steady mode, and that will stabilize the tracked area. Also its a good way to preview how stable the track is when you are done. Whatever you set your refernce time to be in the track mode, do the same for steady mode. Set the same refernce time.
Anyway, you set tracker to steady, and you copy the same tracker node, place it a bit further down the line and you check invert steady mode, to bring back the original motion. Now you place as many nodes or anything you like between steady and unsteady nodes. And you do your work there. That way you work on stabilize footage and you can paint, retouch, add graphics, anything you like and than you use unsteady copy of the original tracker to bring back the original motion. Its a good way to work when you have multiple operations on the footage.
................................
8 - Here is a frame from the result:You can see Bruce Springsteen composited onto the roof of the Flamingo hotel (stock image of Bruce).I fade the image in and out so I don’t have to deal with the wonkiness that happens late in the Planar Transform. That actually works for the storyline anyway.
There you go. Kudos.
9.So, now I have to do the following:a.General clean-up (like blend mode to multiply, a little work on color).b.Create TWO masks:i.One for the tall column (the FLAMINGO sign)ii.One for the roof itself (so the image doesn’t spill beyond the surface of the mask).c.Repeat the process for adding images on top of theother roofs
Yes, that would be the next steps. You can use polygon or B-spline to roto in one of two ways. Actually more than two, but these two are best.
Either use planer transform from the tracker to steady the roto work, so you roto the sign on first and last frame, and than see if there is any need to roto in between or if its linear steady movement, because of planer transform, usually there is only one or two frames to adjust.
Or if you have bunch of stuff to do , use steady / unsteady approach with does the same, but you don't have to use planer transform node. And its more flexible way to work where there is more stuff to do.
I'll post a video I did to illustrate this.
Sun May 05, 2024 12:18 am
