[Paul Dickin]

[Walter Soyka] “AV Foundation… is a low-level framework. No one framework encapsulates everything that an FCPX project/timeline can do”

QuickTime was a low level framework.
But QT couldn’t do this:
AV Foundation uses… a mutable composition to add and remove tracks, and adjust their temporal orderings. You can also set the relative volumes and ramping of audio tracks; and set the opacity, and opacity ramps, of video tracks. A composition is an assemblage of pieces of media held in memory. When you export a composition using an export session, it’s collapsed to a file.
https://developer.apple.com/library/ios/#documentation/AudioVideo/Conceptual/AVFoundationPG/Articles/00_Introduction.html

QT elements didn’t exist only in memory – it took an application like Final Cut to take the raw QT assets and mutate them into a new entity, rendered into memory and ready to be exported.

But that description of editing from the AV Foundation document has subsumed most of what an old FCP timeline (recorded in a project) existed to do.

[Walter Soyka] “But this whole diversion is a bit of a moot point… “

Not entirely. Without understanding the context of the new entity that FCPX is, the discussion will still revolve around why aren’t there hay-net storage facilities in the new (motorised) vehicle.

Or “why can’t we import FCP 7 projects into FCPX?”
If the answer is that its the OS that now handles importing or other data exchange directly into the core Foundation engines then the scale of the problem assumes a different order of magnitude.

[Walter Soyka] ” I don’t think that makes any sense”
Fine, I’m cool with that. 😉

[Paul Dickin]

[Walter Soyka] “…how the app itself “thinks” has changed.”
Hi
Reversing your reply:
How the app itself “thinks” has changed. The new timeline model was a choice. Why (not) assume that this all goes back to AV Foundation?

I’m not saying that that was a ‘good’ choice. Just that that may be part of a simple explanation.

Until the OS’s CoreData functions have full XML I/O APIs FCPX can’t have any external interchange options.

[Paul Dickin]

[Bill Davis] ” …if you can’t open your brain up to the new possibilities… “
Hi
Good post Bill.
When I ‘retired’ from broadcast TV in 1993 I determined that I should boldly go into what was then a buzz word on the distant horizon – Interactive Television.

The prototyping tool of choice at the time (for an individual unsubsidised enthusiast like I was) was Macromedia Director. Director had an inflexibly timeline-based workflow with lots of ‘goto’ moments. In essence clunky and inelegant.

In the late 90s a new program called mTropolis was born – OS 9 only 🙁 (It was later bought out and killed by Quark!!!)
This had an extremely elegant and powerful node-based object-oriented workflow, one which allowed the essentially linear viewer experience of watching the ‘media’, but allowing very cool presentation in an ‘interactive multi-‘ way.

I think interactive television as a paradigm floundered because it neither gave the navigational freedom that say the Web does, nor the high-fascination (totally linear) experience that a good move or television programme provides.

BUT. The editing assembly process an editor goes through to aggregate his cut-together clips to eventually achieve the finely crafted finished movie or TV show is, at a lot of stages in its gestation, very much like the juggling-of-building-blocks that went on with Director or mTropolis.

At that stage of the editing process a Node-Based interface paradigm could be totally appropriate (and productive), in my view.

A question to David L.:
How much time have you spent using Shake or the node-based part of Color?

[Paul Dickin]

[Walter Soyka] ” The AVMutableComposition class that everyone keeps pointing to as the essence of an FCPX project does not provide any tools for clip connections, which are the foundation of the magnetic timeline, or for primary and secondary storylines. I think AV Foundation naturally models timelines the old-fashioned way.”
Hi
The linked document is dated 5th May 2010. A year is a long time in s/w development 😉

The AVF complex clips must be defined in essence in their own database structure. Wouldn’t it be possible in a ‘relational database’ context for the designers of AVF to have added hooks to the individual elements of the AVMutableComposition – to link to other AVMCs (via the database ‘hooks’) to form even more complex AVMC entities?

In other words, since FCPX doesn’t have a ‘project’ in the old way, the databased interaction of AVF and CoreData (= the OS) define the Project (= Sequence). The FCPX GUI designers have worked to service that ‘engine’…
That would be my guess. 😉

[Paul Dickin]

[Herb Sevush] “So tell me again about the greater flexibility of the relational database model….”
Hi
Sure 😉
This, O my Best Beloved, is a story–a new and a wonderful story–a story quite different from the other stories–a story about…..
…XML project management. Not just about FCP1-7 and the hundreds of COW media mangler threads…
Quote:
After you import your media, go grab some coffee or do some jumping jacks or something else interesting, and let the process finish. Now, if Premiere Pro continues to generate files, there is something wonky with the media database, and it usually means purging the PEK/CFA/etc. files and then cleaning the media cache database. Premiere Pro will have to regenerate the files (yeah, again), but this time, they’ll usually stick. I think it can be a little touchy if you continue to try to work during this process, and Premiere Pro can lose track of those files. That’s why most of us have griped again and again for this process to change or go away forever and ever.
https://forums.adobe.com/thread/871659
30th June 2011.

[Paul Dickin]

Hi
A piece of film hanging in a bin being transferred to a viewer doesn’t have a clock – only a frame rate.
The chinagraph marks are never applied with reference to a time-piece. No different in an NLE to my way of thinking 😉

[Paul Dickin]

[David Lawrence] ” I can only speak for myself. When I edit, any relationships I create between clips comes as a direct result of where I place them in time. My concern is always time first. This event happens now. The now part is very important. The relationships arise from this placement. I might not even understand them before I perform them! For me, this is the intuitive, creative part of editing. This is why I want my time foundation to be fixed. I want absolute control of what happens and when it happens.”
Hi
You have got me thinking…
I can’t remember a single instance – not one – since I started editing professionally for the BBC in 1965 where I’ve made a story-telling edit (or an edit decision) based on absolute time. Everything is always ‘relative’ to other A/V assets, and the ebb and flow of the action within the clips or the music, and the precise time-irrelevant cut points.

Sure I’d lay out a 93 feet 12 frames spacer in the 4-way synchroniser to time a 60 sec slot (PAL-land), but only at the end to get it to time – after the edit. Because the movieola or Acmade was too imprecise.

Similarly in a one-hour documentary the show would be assembled and then fine cut with only a very generalised notion of overall running time. Cutting down to make the exact timing would be the very last editorial process, long after the ‘story-telling’ had been locked down.

But I suppose for the BBC it was irrelevant and for ITV/C4 there was a relaxed policy of timing for commercial breaks.

[Paul Dickin]

[Walter Soyka] “What challenge was this redesign meant to address?”
Hi
i) As Craig has indicated, to find a way of graphically representing AV Foundation building blocks in an edit-friendly manner, and more importantly
ii) to make the ‘editing process’ more comprehensible to a (far) wider user base.

ie Not the answer you want to hear I suspect 🙁
But the whole caboodle is reverse engineered from the viewpoint of ‘where the puck will be’.
Not from existing custom and practice…

[Paul Dickin]

Hi
Shut up!!!! Not much chance of that here 🙂

i) For me, existentially, time is almost never ‘linear’ – except when I’m counting the seconds – ie never when I’m editing. So for me, whilst visual story,-telling seconds=centimeters can never be a valid metaphor.

ii) I used Premiere 3.1.3>6.5 almost as long as I’ve used FCP (1993-2001). With its A&B roll topography, and especially its precisely rulered timeline, always marked in meaningful seconds (unlike FCP for a PAL timeline), it was much more a layed down linear representation of time in a line, with meaningful overlaps for transition points. I know editors who still hold Premiere as a ‘better’ NLE for that reason.
Getting used to meaningless timeline time rulering was the biggest hindrance to me when transferring to FCP v2 – it took me all of two weeks to get my head around that…

iv) My favourite ever Mac program was mTropolis, and object-based nodal multimedia authoring program like Director. It had no problems with ‘time’, as any time-based object had its own timeline. So as a descendant of that philosophy I welcome FCP X. 🙂

iii) There was, briefly, a totally trackless Mac NLE called Montage made by Arboretum. Limited though it was it had no really fundamental problems for its being trackless – it was just too simplistic, and therefore limited, to compete with Final Cut Express. Complexity goes with the territory.
Complexity = learning curve.

[Paul Dickin]

Hi
Quote:
“While there are obviously missing features (like a comprehensive redo of Compressor)…”
Isn’t that a pretty myopic view of the missing features?