-
Step away from the Apple Keyer now! (And build a better one of your own.)
-
[Simon Ubsdell] “The question that intrigues me is why is the execution so bad?”
Well, I think that the keyer is remarkably fast. It is quite fun to scrub the keyer icon and see a near real time ‘key’ of the footage, and I think that’s Apple’s focus for the moment. Speed.
Quality will come later, or someone will bring one to market. For better or worse.
I’m not making excuses, I’d love to be able to do quality work in one program instead of multiple, but it’s not here yet.
Jeremy
-
[Jeremy Garchow] “Well, I think that the keyer is remarkably fast. It is quite fun to scrub the keyer icon and see a near real time ‘key’ of the footage, and I think that’s Apple’s focus for the moment. Speed.
Quality will come later, or someone will bring one to market. For better or worse.”
Because I’m a mean-spirited curmudgeon, I’d say that Apple’s focus is always “flash” rather than speed.
(Not that “Flash” obviously!!)
What looks good in a demo, that’s what really drives the development process.
As always, see the incredibly perceptive analysis by Ron Brinkmann who knows the inside story better than either you or me- and whose high-end credentials are something to weep over:
https://digitalcomposting.wordpress.com/2011/06/28/x-vs-pro/
“See, here’s the thing with how features happen at Apple to a great extent – product development is often driven by how well things can be demoed. Maybe not explicitly – nobody ever told me to only design features that demoed well – but the nature of the organization effectively makes it work out that way.”
I know I keep coming across as mean – but I really, really want Apple to step up and do better. They’re just not putting on a quality act right now.
But they could if they tried. Surely?
Simon Ubsdell
http://www.tokyo-uk.com -
[Simon Ubsdell] “Because I’m a mean-spirited curmudgeon, I’d say that Apple’s focus is always “flash” rather than speed.”
I think we are saying the same thing, just in different ways.
To be able to scrub a keyer and have it show up as keyed in real time, is amazing. It is great marketing, ‘flash’, and makes for great demos. I agree with you.
It makes feel good when I roughly composite a key. It only sucks when you want to use it just to find it doesn’t work very well.
[Simon Ubsdell] “As always, see the incredibly perceptive analysis by Ron Brinkmann who knows the inside story better than either you or me- and whose high-end credentials are something to weep over:
https://digitalcomposting.wordpress.com/2011/06/28/x-vs-pro/“
I remember this article. I do think that Apple tows the line between creative people and creative professionals. There is no doubt they want to appeal to the broad middle as it sells the most product.
[Simon Ubsdell] “I know I keep coming across as mean – but I really, really want Apple to step up and do better. They’re just not putting on a quality act right now.
But they could if they tried. Surely?”
I think so. I hope so. There are strokes of genius all over FCPX/Motion, but it is going to take a little while to stabilize and mature. I’m still hanging in there, and there’s nothing wrong with being critical as we have been more than patient with all of this.
I also think, and this is subjective, that Apple is in the middle of huge technological change. The very Quicktime engine that has powered the media capabilities in the OS for decades is slowly being retired, and that is why Apple killed everything off around it. The new system in place, is still being worked on and is not quite a full replacement. Apple moves slowly in this sense, they always have. They release capability gradually, and sometimes, it doesn’t feel fast enough. It’s never fast enough. On top of that there’s intel’s hardware delays, and it’s hardware on which Apple is betting the farm in the pro community, at least for now.
Jeremy
-
[Simon Ubsdell] “It was actually Conduit that go me into the wonderful world of compositing in the first place – it was an amazing discovery. There’s no better place to start learning and the results you can get with it are pretty much as good as it gets.”
Thanks for the fantastic work, Simon!
For me, it was Combustion (RIP) that got me into compositing, then Shake (also RIP), but I got into Conduit several years ago back when dvGarage was still selling it (version 1.5 and 2.2).
As brilliant and comprehensive as your tutorial is, I couldn’t help but feel that the project felt cumbersome to put together in Motion. I kept thinking how much faster it would have gone in Conduit or in any other node-based compositor. It’s difficult (for me, at least) to really see in detail what’s happening in a layer-based system like Motion compared to seeing a node tree. However, you’ve done a masterful job of achieving great results in Motion.
For anyone who is interested in seeing Conduit in action, I highly recommend Alex Lindsay’s excellent tutorial videos on the dvGarage site:
https://www.dvgarage.com/conduit-2-samples
-
[Alan Okey] “I couldn’t help but feel that the project felt cumbersome to put together in Motion. I kept thinking how much faster it would have gone in Conduit or in any other node-based compositor. It’s difficult (for me, at least) to really see in detail what’s happening in a layer-based system like Motion compared to seeing a node tree.”
Thanks, Alan. I completely agree that this stuff is more easily achieved in a node-based environment.
Like you I graduated from Conduit to Shake (and then to Nuke) – and Conduit is a great place to start learning about compositing and I can’t recommend it highly enough, especially now that the entry level package is completely free.
There were a few reasons I wanted to show how to do this in Motion.
First, I suspect a lot of users have opened Motion and never go beyond doing some basic titling work, and I’m keen to show that there’s potentially a lot more to it than that – which is why I started my Motion tutorials on YouTube back last summer with the aim of pushing the envelope beyond the kind of thing you usually see.
https://www.youtube.com/my_videos?o=USecondly, I suspect most editors would be much more intimidated by a node-based demonstration in an application they had never opened.
There are some concepts here that it’s extremely useful to know about if all you’ve even done is push around the sliders on your favourite third party keyer and I think that knowing them, even in only the theory, can really help get better results in the long term.
Thirdly, I was interested to see how Motion would compare to After Effects at this kind of job and in my view it’s every bit as flexible, if not more so in this particular case. Again you often hear that Motion is unsuitable for compositing whereas After Effects is held up as an industry standard.
(As an aside, I don’t find doing this sort of thing at all cumbersome in Motion – but that’s probably just familiarity. I do think that Clones are a very powerful tool in Motion that actually help to confer some of the benefits associated with nodes.)
Personally, I would avoid Ae for compositing and go straight to Nuke or Shake as I don’t find it gives me enough flexibility and it can get pretty cumbersome for this sort of work. But just as I’ve wrangled Motion into doing the job here, there are legions of Ae users who will bend the tools to do pretty much anything.
All in all, though, it makes sense to use the application you’re most comfortable using – whether it’s objectively the best tool for the job is very often less relevant.
Simon Ubsdell
http://www.tokyo-uk.com -
Simon, what a great thread. Thanks so much for your contributions.
[Simon Ubsdell] “The question that intrigues me is why is the execution so bad?”
It is a fascinating question, isn’t it? It’s one of two that I would ask.
What I find equally fascinating is that this is really the first honest discussion of the keyer I’ve seen since its debut. Elsewhere, the M5 keyer is universally praised for its awesomeness.
I tried using the FCPX/M5 keyer for a heavy graphics and compositing job exactly this time last year. The first few shots were great. The interviewee was bald. The second shot featured an interviewee with wispy blonde hair and it sent me straight to Smoke.
I am part of the problem here. I assumed I was doing something wrong, just missing some critical aspect of using this awesome keyer that I had heard so much about. I didn’t reach out and ask for help, and I didn’t question whether the keying emperor was wearing any clothes because I thought I was being dumb. I just got the job done elsewhere so I could make the tight turnaround required.
In short, my question right now is not so much why is the keyer so bad — it’s why does everyone else think it’s so good?
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events -
[Jeremy Garchow] “Quality will come later, or someone will bring one to market. For better or worse.”
Sadly, I think there are plenty of issues where first-party quality never arrived at all. Constant QuickTime gamma/profile/swing confusion, I’m looking at you.
I guess there’s no need for keying to be a first-party solution, but the presence of what Simon has brilliantly termed “school boy errors” — and the fact that a better keyer can be easily built by a customer with the existing toolset as Simon shows here — is dismaying. They did 98% of what they needed to do, but not the last 2% to really make it work. If that last 2% never comes, it almost makes the 98% effort a waste.
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events -
[Walter Soyka] “In short, my question right now is not so much why is the keyer so bad — it’s why does everyone else think it’s so good?”
I admit to being just as guilty on this score.
I looked at it when it first came out and I thought it seemed pretty cool … and then I’m sure I’m not wrong in thinking I read loads of comments about how amazing it was – and ended up repeating that opinion to others.
All without really giving the tires a proper kicking. Now I find that if you kick the tires too hard, the wheels start to wobble!
I’m sure Apple knew exactly what they were doing with this – selling the sizzle first as always and not worrying too much about the quality of the steak. I think a lot of us, me included, took too much for granted.
Ron Brinkmann again: “See, here’s the thing with how features happen at Apple to a great extent – product development is often driven by how well things can be demoed. Maybe not explicitly – nobody ever told me to only design features that demoed well – but the nature of the organization effectively makes it work out that way.”
Simon Ubsdell
http://www.tokyo-uk.com -
[Walter Soyka] “Sadly, I think there are plenty of issues where first-party quality never arrived at all. Constant QuickTime gamma/profile/swing confusion, I’m looking at you.”
Oh it arrived. It arrived in the death of Quicktime as we know it! 😉
[Walter Soyka] “I guess there’s no need for keying to be a first-party solution, but the presence of what Simon has brilliantly termed “school boy errors” — and the fact that a better keyer can be easily built by a customer with the existing toolset as Simon shows here — is dismaying. They did 98% of what they needed to do, but not the last 2% to really make it work. If that last 2% never comes, it almost makes the 98% effort a waste.”
Except for the people who think it’s awesome.
I also think it’s the fastest rough cut keyer available. For that, I praise it. It is really easy to get a look roughed out, and content approved/recut, before moving on to the more detailed work in some other application.
In short, it’s not worth nothing, but really close to nothing.
-
[Simon Ubsdell] “I’m sure Apple knew exactly what they were doing with this – selling the sizzle first as always and not worrying too much about the quality of the steak. I think a lot of us, me included, took too much for granted.”
Have you tried the PHYX Keyer? Lots of sizzle there, too, but I haven’t sampled the steak. Maybe it has both.
A tangentially-related question for you, Simon: as a third-party developer, do you feel the low pricing of FCPX/M5 has sucked the oxygen out of the room, or has it created a larger user base that can make up lost margin in volume to fill holes like this poor keyer?
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events
Reply to this Discussion! Login or Sign Up