-
FCPX and time code
-
“In any case – pixel interpolating metaphor aside – you seem to be implying by your argument that 7toX is not a “proper” translation.”
Going either direction is a problem, but even worse going from X to 7. IMHO, all translations from FCP X XML are at best a hack at this point. The information exposed to the developers is minimal and often requires them to “read the tea leaves” as to what Apple intended. At this point it’s a very immature standard and often requires that media files are still online. For instance, if you need to be able to have reel numbers in an FCP 7 translation or an EDL (generated by EDL-X) that reel info is read directly from the media files by FCP 7 or EDL-X, as it’s not passed through FCP X XML yet.
Oliver
Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com -
Well — PIOPs are here!
https://www.philiphodgetts.com/2012/10/final-cut-pro-x-10-0-6/
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 -
[Franz Bieberkopf] “As I pointed out in my ProTools / OMF example above, I am not necessarily (ever?) looking for a translation to “look as if it had been originally created in” the new tool. Different tools are just that, and creating in them often brings different decisions or different ways of expressing those decisions (ie stereo tracks vs. stereo clips). It is one thing to present the limitations of a translation, quite another to say that it can’t be done.”
Continuing your translation metaphor, a literal translation is easy, but less valuable. An idiomatic translation is not so easy, but vastly more valuable.
I really don’t see what the fuss is about. Apple more or less says that 7’s data model doesn’t contain all the information necessary for representing the edit in X’s data model. Intelligent Assistance says attempts to make best guesses at what that missing data should be.
The two theses are perfectly compatible.
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,
… considering this was originally posted as a cheap shot a Apple regarding the persistent PIOP issue, it would seem the point of wit has been rather blunted.
Franz.
(It still raised an interesting and unexpected exchange though, and I am still rather surprised by your take – so I’d still like to hear you talk about project translations using other examples.)
-
You thought my post that unfortunately derailed the thread regarding TC into a PIOP battle was a cheap shot? It wasn’t intended as a cheap shot.
-
Keith,
My post was the cheap shot.
Franz.
-
nice! sorry to derail anyways. it was a good TC feature discussion before. trying to download 10.7.5 so I can upgrade and dig on some PIOPs…
-
[Franz Bieberkopf] “Walter… considering this was originally posted as a cheap shot a Apple regarding the persistent PIOP issue, it would seem the point of wit has been rather blunted.”
My apologies. I don’t know if I was being dour or just plain dense, but it whooshed right over my head…
[Franz Bieberkopf] “(It still raised an interesting and unexpected exchange though, and I am still rather surprised by your take – so I’d still like to hear you talk about project translations using other examples.)”
There aren’t really other good examples to use, because in general, project translation happens within the same linguistic family. Translating from Legend to Pr is pretty straightforward, because they both fundamentally represent the data in the same fashion. Translating from Legend to ProTools via OMF is pretty straightforward for similar reasons.
FCPX describes an edit with a totally different data set than other NLEs. This is perhaps more like translating between layer-based and nodal compositors, but I’ll be the first to point out that you can represent a layer-based composite entirely with nodes, and a node-based composite entirely with (often duplicated) layers. However, just like with a 7toX translation, you’ll get a project file that renders the correct output; just like with a 7toX translation, it’s a very hard problem to provide organization within the project file that is meaningful to the user because some missing data must be guessed.
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] “… just like with a 7toX translation, you’ll get a project file that renders the correct output …”
Walter,
I think this speaks to my question. Restricting ourselves for the moment to the question of timelines, I’m speculating that most people would be satisfied with a timeline that looks the same on output (even with a few qualifiers), as well as replicating keys structures of the timeline, and would not expect it to look “as if it had been originally created in” the software in question. I think the underlying model is of less interest than the practical implications.
Again the OMF example – it is in some ways quite limited both in what it brings from the original project, and in the way it exploits the possibilities of the importing software (like ProTools), but it is a very useful translation nonetheless. One might even say “proper” (though limited).
The FCP7 to PProCS6 example might illustrate the point, because if you created a project in PPro you might exploit features not in FCP7 (like some of the audio features), where in FCP7 you would not use these features because they don’t exist – therefore the project would not, strictly speaking look “as if it had been originally created in” PPro.
Where this may falter is – as you’ve pointed out above – the (significant) issues of organization outside of timelines.
I take it you feel there are translations (like FCP7 to PPro?) that fulfill the “as if it had been originally created in” criteria?
Franz.
-
[Franz Bieberkopf] “Again the OMF example – it is in some ways quite limited both in what it brings from the original project, and in the way it exploits the possibilities of the importing software (like ProTools), but it is a very useful translation nonetheless. One might even say “proper” (though limited).
The FCP7 to PProCS6 example might illustrate the point, because if you created a project in PPro you might exploit features not in FCP7 (like some of the audio features), where in FCP7 you would not use these features because they don’t exist – therefore the project would not, strictly speaking look “as if it had been originally created in” PPro.”
Let’s go Chomsky. FCP7/PrCS6/ProTools/OMF/whatever have different surface structures, but the same deep structure.
FCPX is an anomaly. It has a different deep structure that cannot be expressed by the data structures in other applications.
[Franz Bieberkopf] “Where this may falter is – as you’ve pointed out above – the (significant) issues of organization outside of timelines.”
With FCP7/FCPX, I think it also significantly falters at organization within the timeline itself — and that’s a big problem. If the translator doesn’t guess the relationships right, then the editor can’t exploit the magnetic timeline out of the box. Magnetic maneuvers will work exactly as the timeline reads, but garbage in, garbage out.
It’s the difference in organizational structure — on a deeper level than differences in features — that makes FCP7 to FCPX translations so hard.
[Franz Bieberkopf] “I think this speaks to my question. Restricting ourselves for the moment to the question of timelines, I’m speculating that most people would be satisfied with a timeline that looks the same on output (even with a few qualifiers), as well as replicating keys structures of the timeline, and would not expect it to look “as if it had been originally created in” the software in question. I think the underlying model is of less interest than the practical implications.”
I agree with you that a “wrong” timeline that yields the right output is vastly more valuable than no ability to import legacy projects at all. Practically speaking, an imperfect translator is better than no translator, and I can only assume that the market is proving that with Intelligent Assistance products.
I said “as if it had been originally created in,” but I meant that to be very specific to FCPX, because an FCPX edit is so built differently in terms of method and stored differently in terms of structure than an output-identical Legend edit. My phrase has less meaning for me with ideologically similar software.
I think that Apple chose not to pursue a translation feature because by definition, it cannot be perfect, and because an imperfect translation would make the magnetic timeline feel broken.
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