Rotoscoping is one thing, animating using shapes is kinda another.
It's entirely possible to rotoscope in Motion. But I wouldn't recommend rotoscoping with a single layer on any program for complex shapes. Rotoscopes can be done with the pen tool and then those shapes used as a mask.This will create a great roto thats still adjustable.
I have a whole lesson on YouTube that covers this in apple Motion if you search for AV-Ultra Rotoscope.
Animating can be done the same way. If Im making a character- I would have a layer for basically any articulated points. Group them and change the anchor points to act as null positions for rotations.
Need displacements? Or want to animate a T-pose 2d character?
MotionVfx recently released mPuppet which can create a mesh form an image and you can animate the displacement mesh over time. Even add behaviors to the various control points to create keyframe-less animations.
Couple tips for roto work that have helped me:
1. Break up complex shapes into several pieces. Don't try and roto a big complex shape in 1 layer.
2. Unmoving simple shape? Track a feature on it and then make a mask from pen shape and tweak only where needed.
3. Start with making keyframes every 10 frames. Then after you get that, go to the 6th frame and tweak if you need to, and go to the 2nd or 4th. if you don't have to do every single frame, don't.
4. Make sure you are moving the point coordinate and not the whole shape. Motion unfortunately has a tendency to move a whole shape sometimes and that can throw things off.
5. if you are doing a LOT of roto work use silhouette- its not in Motion or FCP- but having onion skinning alone is worth the price of serious roto. Plus 2020 has some crazy cool features.
Its now part of BorisFX
Hope that helps!
As others mentioned. I think 10.5 will be very telling of what Apple is planning going forward.
With a push to services- I could see FCP being free but multiuser, iCloud proxy libraries, and better tracking being subscription services.
I don't know if Gamesack will be the same without FCP!
Also I really like your stacked slice transitions lately.
Personally I see Frame.io being a full online, multiuser realtime editor for newsrooms in less than 5 years.
Those are the people I'm keeping my eye on.
The future I see, doesn't require any kind of desktop or any specific OS.
Just a web browser.
Depends on the level of collaboration you are looking for.
Want to work on the same edit same time?
I don't see a lot of option out there -yet-.
I have seen a couple companies demoing doing a streaming service like Google Stadia but just for very high end workstations for 3D, editing, VFX. I guess it's just the next evolution in render farms.
Just need to review an edit? make notes?
Frame.io All day, every day.
Love it, and use it with clients on the other side of oceans daily.
Reliable, fast, and rarely any issues.
Realtime collaboration with multiple editors?
Not the answer you are likely looking for but this works great in Davinci resolve.
Why apple hasn't/cant/won't do collaborative edits like this, I don't know. Seems like a no brainer. (and maybe they have based on the multiple playheads found under the hood of FCP)
Today I had a chance to do something I've wanted to do for a while and test practical exports with the Apple official BMD eGPU (the original) and noticed some interesting times.
I have a 2017 iMac 3.4 i5 64 GB and rendering a 4k 25p timeline of just over 8:30.
Anytime I specify the eGPU my rendering of clips, and few simple graphics titles (no heavy animation) Im getting the following times:
Same export setting used for all exports (a compressor export I set up in FCP)
eGPU 580 using FCP- 21:22
eGPU 580 using Compressor 20:44
internal 570 using FCP- 9:18
internal 570 using Compressor 10:10
All other conditions are the same except for physically disconnecting the eGPU.
Anyone getting other results like this?
In earlier versions I believe I was getting better times.
my export is a h.264 for final delivery- did not test other codecs yet.
Should post the new one you just made too
It possible to do in Final Cut- but its faster/easier/better for a variety of reasons to build it Motion and then publish it to Final Cut.
Short answer in FCP- You'll need to use titles and masks to reveal the text however but since theres not really good layering and keyframe controls in FCP, its just a bit clunky of a progress.
I have a couple lessons on how to use masks and create mask reveals in Apple Motion on YouTube you can search and find.
Outside of building one, theres a few places that have some simple title reveals for final cut if you search for Final Cut Title templates online.
I personally think that might not be worth the trouble. Is the entire machine sluggish? Because that would likely be my first plan of attack. Is there something running in the background?
Does the 2017 have the t2 chip? If so it may not be able to be run from an external. But I may be wrong on that one.
Also can depend on what effects you have installed. I’ve had serious issues running certain plugins, or using templates that have a lot of 3D items.
Is it your footage? Sometimes it can take a bit to respond for certain codecs and editing with them.
The reason why I’m pushing so hard against it- is if you get a new machine, if it’s the same kinda of issues, you will have done a lot of work for nothing.
I have a 2017 that is incredibly fast and haven’t had issues with 4k and loads of templates/effects.
(After I cleared some plugins that I believe slowed things down)
Anyway- hope that helps.
Thats an interesting error.
While I dont have an answer for this exact question- I've found that sharing direct to a service (like YouTube in this case) is often prone to errors. I found when I was uploading straight from FCP to YouTube, sometimes it would error out, sometimes it would get all the way to the end and something else went wrong, and I'd have to wait all over again and try a new export.
The worst situation- it DID work and I would review it on YouTube and realize I made a mistake somewhere. Then fix the mistake and upload a whole new one again.
So instead I make it a point to always render out a local copy first.
I know its an extra step, but I've learned that by rendering a local copy first, I have a final to review before upload, and the YouTube interface for upload just works better all around.
It also has the added benefit of giving me a 'local master copy' just in case I ever decide to move my things off YouTube to another service, or in case YouTube changes its policy etc. I always have another copy of it.
Like I said, dont know if it helps,I believe having a local copy and uploading using the browser ends up working better.
And now I never have errors or have to wait for a new export to upload if/when it fails.
I guess I was just trying to start this as a fun wish list of things before the actual release.
See what matches up from user wants vs. release.
I’d also love a refreshed tracker- and port it to fcp.
To be honest if there was one feature I would wish for would be a front end to support csv, or the app numbers, to make titles and graphics driven from spreadsheets like dataclay templater and afternoon effects.
Can you use the image transfer app?
Or can you import into photos app?
So those ‘who know can’t say’ is the rule of the game.
But boy I’ve heard some interesting things lately.
If we are just talking rumors- what’s your personal favorite?
Mine is this: Fcpx on ipadOS
I would love the idea of fcpx on iPadOS.
And moving between desktop and iPad? Def what I’d like to do on a daily basis.
What’s your favorite rumor of 10.5?
I've found that FCP 10.4.7 and Compressor 4.4.5 don't use my eGPU nearly as much as the previous versions.
In 4.4.4 if I recall it was dependent on export format.
Some formats seemed to use both the internal and external gpus and I was getting great export speeds, but in the new versions its topping out around 15% usage.
Fill out a bug report?