-
Steve Kanter: What FCPX CAN Do
-
[Walter Soyka] “I’m suggesting that the design decisions that make the magnetic timeline possible make legacy import nearly inconceivable without some major changes in how the timeline is implemented. “
Walter,
“Some men see things as they are and ask why, I dream things that never were and ask why not.” – Bobby Kennedy
I might be wrong, but then again I might not be wrong… In any case, I can’t imagine that a translation app of some sort can’t be created. Heck, we put men on the moon with less computer than what’s in my phone.
David Roth Weiss
Director/Editor/Colorist
David Weiss Productions, Inc.
Los Angeles
https://www.drwfilms.comDon’t miss my new tutorial: Prepare for a seamless transition to FCP X and OS X Lion
https://library.creativecow.net/weiss_roth_david/FCP-10-MAC-Lion/1POST-PRODUCTION WITHOUT THE USUAL INSANITY ™
Creative COW contributing editor and a forum host of the Business & Marketing and Apple Final Cut Pro forums.
-
[Michael Hancock] “Audio dissolves spring to mind. I’ve seen a lot of complaints about that.”
Individually audio dissolves actually work very well. Expand Clip and then you have fader handle controls plus the curve options. The problem is when you need to do a lot of them and you really just need to use a “standard” dissolve added to a bunch of clips. Then your forced to detach audio
[Michael Hancock] “Apple failed to fix huge issues with FCP7. Gamma issues?”
Fixed in FCPX I believe.[Michael Hancock] “Pulldown when mixing frame rates in the timeline? “
Have you checked this in FCPX?[Michael Hancock] “There are too many examples of things being poorly done before. Or do they really care this time to get everything right, all the time?”
Some of the aforementioned issues may have been difficult to fix without a patchwork approach. That’s good reason to toss the whole codebase and start over.
I can imagine import presented a lot of issues. Although superficially one might imagine an import method which gave you options to make tracks either connected clips or secondary storylines but I can only guess they ran into so many issues and so many forks to travel down and the result was making a major R&D into an import interface with many options to run through and each creating a tangle of other options that it just didn’t make any practical sense.
-
[Craig Seeman] “[Michael Hancock] “Pulldown when mixing frame rates in the timeline? ”
Have you checked this in FCPX?
“I’m not talking about FCPX here – I’m talking about FCP7. I assume they fixed it in FCPX? Shame they never did for 7. It needed to be. And to be perfectly clear, I don’t have FCPX. It doesn’t have features I need and, while I’d like to try it, I’m not paying $300 for it. If they had a trial (like everyone else) I definitely would. I believe there’s some stuff there I’d love (and a lot I wouldn’t).
[Craig Seeman] “[Michael Hancock] “Apple failed to fix huge issues with FCP7. Gamma issues?”
Fixed in FCPX I believe.”Good. It desperately needed to be. Now if they hadn’t broken so much else in the process. 🙂
[Craig Seeman] “Some of the aforementioned issues may have been difficult to fix without a patchwork approach. “
Some of them, but certainly not all. I imagine the Gamma issue would have been a huge undertaking, but pulldown? They never should have gotten that wrong.
My original post was a counter to the idea that many things might be missing from FCPX because Apple has such high QC standards that some stuff just didn’t hit them. I’m saying their previous software is evidence that their QC standards aren’t as high as people may imagine. Stuff isn’t there for other reasons, and Apple is talking.
It will be interesting to see how FCPX develops.
—————-
Michael Hancock
Editor -
[David Roth Weiss] “”Some men see things as they are and ask why, I dream things that never were and ask why not.” – Bobby Kennedy”
My favorite variation on this theme is the old Chinese proverb, “The person who says something is impossible should not interrupt the person who is doing it.”
[David Roth Weiss] “I might be wrong, but then again I might not be wrong… In any case, I can’t imagine that a translation app of some sort can’t be created. Heck, we put men on the moon with less computer than what’s in my phone.”
I don’t think that it’s literally impossible to translate an FCP7 timeline to FCPX. I do think it’s a more complex problem than many have considered, and it would require some serious intelligence to look at the FCP7 edit, discern the relationship between clips from their placement in time alone, and encode those relationships for FCPX.
Consider an FCP7 timeline. This clip that starts on V2 at 00:07:24;10? Does it belong in the primary storyline? Or maybe a secondary storyline? Or maybe it’s a connected clip? Is it connected on the first frame? Perhaps there’s an audio hit that connects to a visual hit? Or perhaps it’s connected to a specific reference in the audio track? What role should be assigned to the audio track?
Now that’s just one clip. Those decisions must be made for every single edit, and you really need to understand the content as well as the placement in time to determine a clip’s relationship to the other clips. If any one of those decisions are incorrect, the magnetic timeline may not behave the way it’s supposed to when manipulated, because it relies on well-defined relationships between clips.
I think that magnetizing a timeline like this is a non-trivial task for humans, let alone computers. If the computer could read the EDL and figure out the relationships between the clips from their placement and content, would it really be that far off from being able to automatically edit the content from scratch?
The magnetic timeline encodes an awful lot of context at the time the edit is made for later use. The translation problem gets a lot easier to solve if you remove some of the constraints that the magnetic timeline imposes.
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 -
Interlaced 3:2 pulldown is fixed in FCPx.
The weird thing is fcp <= 7 understood 3:2 pulldown from v3 or so, as it could add it in real time on a dv output for chrissakes. It could just never render it properly in the timeline.
[Michael Hancock] “I’m saying their previous software is evidence that their QC standards aren’t as high as people may imagine.”
I don’t know if many people thought the fcs QC process was all that stringent. I think part of that had to do with a set of really, really old code, mixed with newer code (Color, Motion, FCServer etc) and the quilted patchwork was getting harder and harder to make work. In my opinion, this is one of the reasons why FCP has had such a major restart. Or whatever you want to call it.
-
[Craig Seeman] “Motion is very good and unlike FCPX has gotten a good reception at about 4 stars in the App store.”
Deservedly so! Apple did a very nice job with the update to Motion. If they had approached the FCP update in a similar manner, I have no doubt we’d all be jumping for joy.
_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl -
[Walter Soyka]FCP7’s timeline places clips in absolute time.[snip]
FCPX, on the other hand, has the relationship between clips built into its fundamental design. FCPX’s timeline places clips relative to each other.You are totally over-thinking this. The magnetic timeline is nothing more than a “gimmick.” A series of “If” “then” and “elses” that automatically groups clips together in a way that it thinks will be helpful to the editor. While FCPX may try to do its damnedest to hide time from the end user, it still has to relate to time at some point.
When you edit in FCPX, you’re still editing in the typical way. You lay down a “master” track, then you place tracks next to it or on top, and so on. There’s really no difference, and don’t let the pizzazz fool you. You can still select a clip and MOVE IT FORWARD IN TIME or even BACKWARDS IN TIME. Then, the logic functions of FCPX go “Hey, I think until editor specifically wants to move this clip again, I’ll just connect it to the master clip!” Remember, all of this stuff has to be done from scratch at some point. No footage being imported into FCPX has anything done to it to give any kind of pseduo-FCPX-mumbo-jumbo-magnetism that makes things work. It’s just footage, like it has always been. FCPX just has a lot of fancy processing functions that try to hide stuff from the end user, but at its core it is processing time just like every NLE on the planet.
Now, I’m not saying it’s easy, I’m sure there are some complicated edits that would require some translation to get across, but there’s absolutely no reason that FCPX could not import a FCP7 project on a basic level.
-
[Walter Soyka]Consider an FCP7 timeline. This clip that starts on V2 at 00:07:24;10? Does it belong in the primary storyline? Or maybe a secondary storyline?
It would go into a secondary storyline, by default. There are exceptions, but I would say that most edits have the meat of the footage in track 1 anyway. You can always move it down into the primary if it’s meant to be more than a “master clip,” there’s no reason why “accidentally” putting video tracks 2-infinity into secondary storylines would break anything. It’s all layers.
Or maybe it’s a connected clip? Is it connected on the first frame?
How does FCPX know to connect clips when you first drag raw footage into it? It makes a guess, and then you have to turn it off or select another option, so I fail to see how playing by the default rules that FCPX already has to do would be any different.
Perhaps there’s an audio hit that connects to a visual hit? Or perhaps it’s connected to a specific reference in the audio track? What role should be assigned to the audio track?
Role? You mean like keyword? Assign it “Audio Track 2” or whatever they would like to designate it. Then you can rename it, can’t you? Sure, you’d have to go in and do housekeeping chores, but when else wouldn’t you have to when importing a legacy project? And an “audio hit” that is connected to a “visual hit” is an element of time. Video starts at 32sec 12frames and the audio must come in at 32sec 22frames in order to match the onscreen action. How could FCPX NOT know how to handle this?
and you really need to understand the content as well as the placement in time to determine a clip’s relationship to the other clips
You have, more than once, referred to FCPX needing to know things that, frankly, make it sound like it has some kind of Artificial Intelligence engine. It does not. FCPX is based on a programmers idea of how to group things in order to “second guess” the editor and do some stuff automatically. When the guess is right, it’s great and saves time. When it’s wrong, you have to fix things. Considering it does this just from a brand new project, why don’t you think it can be done in a project importer?
-
In the order of your post:
Until FCPX offers feature parity with Avid, FCP7, and PrP, many editors can’t simply choose one or the other because of the paradigm it offers.I find that analysis spot on. However, I’m simply advocating the wisdom of not spending over much time stuck in a mode comparing an 11 year old program which has evolved step by step from insignificance to world-wide professional acceptance – to one that’s been gestating for a couple of years at best, and only “live” for a couple of months.
The temptation is HUGE. And FCP-X clearly compares woefully if that’s the standard you must hold it to. My argument is that it’s clearly hard for many (tho clearly not you, personally to accept the reality that the designers NEVER intended FCP-X to be an evolution of FCP-7 – it’s it’s own beast – Accepting that is the required first stage to = judging whether it has unique value on it’s own.
This debate, largely driven by a palpable cloud of anger, loss and fear – (all reasonable emotions in anyone who sees a critical tool that feeds them in “jeopardy.”) has been way more about the LOSS of 7 than about the reality of X.
And I simply think that misses WAY too much of what’s going on here.
Your comment: FCPX forks off from previous NLEs in the fourth generation (software-only, new abstract editorial model).
I don’t think that there’s anything inherent in the self-organizing, trackless, magnetic, relative timeline in FCPX that gives it new capabilities that the previous standard manually-organized, tracked, absolute timeline had. Am I missing something here?Maybe not. Perhaps providing “new capabilities” is something they simply decided to rationally de-value – at least in the early days of the software – and instead elevated OTHER features that they felt would be potentially more valuable to a future superior sustainable business model higher on their “critical” list.
What might those be?
We’ve heard a LOT about the underlying relational over flat-file database underpinnings.
Those may be important if this turns out to be a “feature race” in the long run. If so, they merely valued long term profits over short term ease of development. It’s a RARE view in business these days, I’ll admit – but Apple is pretty singular in having enough return to make any shareholder happy – so they have room other companies simply don’t to hold a long range view if they so choose.But maybe there are simpler forces at work here. What about cloud deliverability? Apple is making HUGE bets (and reaping equally huge profits) by moving to an “all virtual” business model. Packaged goods are disappearing from Apple stores in favor of the App Store model. (Perhaps it’s instructive to note here that FCP-X was the very first product in the App Store that broke into the “over $100” price category. Perhaps code (and therefore feature) “slimness” was a factor in the FCP-X release strategy? It would be interesting to see the “features per lines of code” comparison between FCP-X, FCP legacy” and the other NLE products. (I doubt that metric even exists, but it WOULD be interesting)
Or perhaps they actually did that odd duck RESEARCH – and found that the growth curve is truly not in the high-end priests of the industry – but a larger body of day to day worker bees that are more and more tasked with doing functional video work in this new economy. To me that WOULD be sad. But maybe it’s truly an inescapable reality.
I don’t presume to know the answers to these questions, myself. I just know that FCP-X is a big bet. That bet is that the game is changing. I think that’s correct. The game is changing BIG TIME.
And once you acknowledge that, the REAL game is to try and figure out what the new game will be and how you can position yourself to best capitalize on it.
Every minute spent moaning about what’s lost (or even WHY it was lost) in the transition from FCP-7 to FCP-X is, IMO wasted time. The key is to understand what Apple saw coming that led them to make this huge break – whether it was driven by market forces, editing interface visions, database possibilities, economic forces, cloud deliverability, or just orneryness – the key is to focus on whether this upheaval presents possibilities for jumping forward over the ever increasing competition.
But that requires limiting distractions. And for me, this debate is remaining far too long in a place where people seldom interested in a rational discussion of the whys and wherefores – but rather just REACTING angrily to a changing game.
The game IS changing.
Walter, I truly appreciate your desire to look beyond the surface. That will be one reason I think you’re left standing when many who are so palpably resistant to even considering what might be better about FCP-X, and, more critically, what it’s revision might mean in a larger sense, will be left behind.
Peace.
“Before speaking out ask yourself whether your words are true, whether they are respectful and whether they are needed in our civil discussions.”-Justice O’Conner
-
[Walter Soyka] “I think they (Apple) simply chose to pursue a new editorial paradigm that needs more information to represent an edit than can be derived from legacy systems without reconsidering the magnetic timeline implementation.”
Hi
Thanks for your explanations Walter. Very helpful 🙂[Walter Soyka] “I don’t think it’s because Apple has done something that we just haven’t understood yet …”
I wasn’t really referring to FCP X, more to the overall (software/services) plan of which FCP X is just a part (Apple’s Big Plan) – which doesn’t include current broadcast workflows/professionals, about which they seemingly “don’t give a damn” 🙁[Walter Soyka] “I don’t think that it’s literally impossible to translate an FCP7 timeline to FCPX. I do think it’s a more complex problem than many have considered, and it would require some serious intelligence to look at the FCP7 edit, discern the relationship between clips from their placement in time alone, and encode those relationships for FCPX.
The magnetic timeline encodes an awful lot of context at the time the edit is made for later use. The translation problem gets a lot easier to solve if you remove some of the constraints that the magnetic timeline imposes.”
[Gary Huff] “You have, more than once, referred to FCPX needing to know things that, frankly, make it sound like it has some kind of Artificial Intelligence engine. It does not.”
The ‘choices’ that Walter refers to seem more to be a set of ‘rules’, rather than artificial intelligence.
With the new FCP X paradigm Apple has chosen to impose a set of rules that the GUI insists on, which seek to impose meaningful context on the layout of the timeline.It appears to me that FCP X puts this timeline relationship data into its own database, which sits above, and ‘controls’, the lesser. simpler, sets of data that give structure to the AV Foundation assets that are brought into the various Events.
AV Foundation assets – AV Mutable Compositions – can be an order more complicated than the QuickTime file architecture assets they replace. FCP X takes full advantage of the new power that this complexity unlocks.
It is at the lower level database that David Lawrence’s time-defined track-element structures exist and get defined – at what frame which track element in which basic asset gets played.The XML of an FCP 7 project maybe can only go as far as defining the basic data sets of the QT assets – which will need to be ‘translated’ into AVF assets.
Even an FCP 7 sequence probably just corresponds to an AVF data set.
Its only with FCP X’s ‘super’ database that all the lower level data gets coordinated into the edit as defined by the GUI and toolset that FCP X provides. And its ‘rules’.So dealing with a legacy FCP 7 project will require an ‘application’ to handle the allocation of various types of data from the XML into the two databases, as per the new rules. That application isn’t – by Apple’s deliberate choice – FCP X.
I know most editors aren’t interested in this infrastructure stuff, but I think its only through having a basic comprehension of the overall methodology – the reasons they’re doing things – that those ‘ching’ moments can happen.
Like in the thread where it was mentioned that maybe the reason why in/out point don’t stick in the Event browser is that this is could be designed (eventually 😉 ) for multi-user cooperative workflows, so editing metadata should only be affixed to the asset in an individual user’s ‘project’ – as a favorite or whatever. The shared Event assets should remained unaffected…
Reply to this Discussion! Login or Sign Up