David Cherniack
Forum Replies Created
-
[Walter Soyka] “I think the magnetic timeline is actually a really interesting idea, and I think that clip connections would work well in a traditional absolute non-collapsing tracked timeline.”
Wouldn’t this really amount to more or less the same thing as the Group Clips” function that exists in PrPro and by other names in other NLEs? The only potential difference that I see is that the resulting object has meaning in the relational database, but that can be worked around in other NLEs.
David
AllinOneFilms.com -
[Walter Soyka] “With this example, I think it becomes important to distinguish among different concepts in the FCPX “timeline” which FCPX bundles together into one big idea.
There’s the magnetic timeline, which relates clips to each other. There’s the storyline, which fundamentally relates clips to one main video track, instead of placing clips in absolute time. There’s the self-collapsing trackless timeline, which mitigates destructive clip collision.”
But how does the distiction address the specific example and its execution in an efficient way? I do believe this may be the tipping point where the magnetic timeline becomes comparatively inefficient.
David
AllinOneFilms.com -
[Craig Seeman] “I’m not quite sure I understand what you’re trying to do.”
I’ll try to expalin it another way because I think the example exposes a fundamental limitation in the efficiency of the magnetic timeline:
Say you have ten layers of video and ten layers of audio. At a single point in time you wish to create a gap, say to insert a new clip. But all the clips at that point overlap with some cuts and transitions happening to the left and some to the right of your insert point. The gap you want to create is not a vertical hole, a cut all tracks and ripple. Rather it’s going to maintain that overlapping nature. But which clips will move to the right and which won’t are editorial decisions.
I’ve found the most efficient way to do this is in non-ripple mode…to select all clips that intersect my insert point in time plus all those that lie to the right of them. (I can do that in Premiere Pro with a single macro-enabled mouse click, otherwise it’s 3 operations) Then I de-select the clips that I don’t want to move, and move the entire selection rightward. This leaves me with the overlapping gaps that I want. I then insert the new clip and extend whatever layers of sound and video that I decide are required to close up the newly created gaps on their various tracks.
This operation is necessary in some form or other in all complex timelines of any length. I will do it hundreds of times in my projects. And I’m sure there are alternative ways of doing it in all NLEs. The point is though, that it requires a lot of editorial decisions about individual elements, stuff that couldn’t possibly be done in ripple mode without the software reading my mind. And it seems to me that it will always be done more efficiently in non-ripple mode with a traditional vertical timeline.
The magnetic timeline appears to be designed to be better at automating simple tasks of assembly where it doesn’t require much built in ripple intelligence. Go beyond a few layers of overlapping video and audio and it’s efficiency will have rapidly diminishing returns. It’s clear that it was designed to be efficient at less complex forms of editing and thus appeal to a less demanding market. But unless the underpinnings of the interface will allow Apple to provide the alternative of a traditional multi-tracked timeline at some point in the future, I think using X for any complex editing task will be done more for the sake of proving that it’s possible rather than it provides a clear advantage.
As far as multi-layered audio the idea of designating tracks by metadata looks like an inherently clumsy workaround compared to the visually intuitive track based vertical timeline.
David
AllinOneFilms.com -
[David Lawrence] “Outstanding observations and a key question, David. Great stuff!
Yes — what does this interaction model mean for rhythmic modes of cutting? Cutting is always about rhythm and pacing. What does it mean to be cutting to a relative frame-of-reference?”
Thank you David.
I suspect that Craig’s point about rhythm & pacing being relative to other objects rather than absolute time is only partly true. It holds for rhythm but not so much for overall pacing which is often a combination of clock time and content, as any long form editor will attest.
Nevertheless, I’m sure the magnetic timeline can manage both. What’s not clear is how it deals with complex insert situations that I outlined in my post to Craig elsewhere in the thread…which may be the record for the longest wait for response by Craig in the the history of this forum…two hours and counting 🙂 (Ah well, I give the guy the benefit of the doubt…maybe his typing fingers have fallen off :).
David
AllinOneFilms.com -
[Craig Seeman] “[David Lawrence] “The question is, is this more efficient?”
It is when you’ve created stacks of items that are designed to happen relative to each other. The interview, cutaways, titles, fx, are relative to the words and actions in the interview and you can move that entire relative relationship anywhere in time and the relationship will be maintained.”
Craig, here’s a fundamental issue I’m wondering about with the grouped stack:
I don’t edit in ripple mode much at all. I frequently need to insert a video or audio clip in a timeline that may have 10 tracks of both video and audio. I select everything to the right of the insert point with a single click of my five button mouse and deselect those clips that overlap the insert point from the left that I don’t want to move. I then create a gap equal or longer than the clip I’m inserting and plunk the new clip on the correct track depending, if it’s video, on the layered ordering I want. But the clips that don’t move may be only 2 of 10 overlapping clips. I know which ones they are. I know the vertical layering order of the new clip. No software algorithm will or can.
It’s hard for me to see how a grouped stack of equal complexity can handle the insert situation I described as efficiently. This is where the vertical fixed track would seem to me to be more capable than going about placing and replacing objects before and after they’ve been shifted ‘magnetically’. But please enlighten me if I’m wrong.
David
AllinOneFilms.com -
Some observations that may fall out of your post:
FCP-X’s timeline is a creation from the minds of object oriented software engineers.
As you point out we experience the playback of a timeline in absolute time, not the relative time the objects and their containers exist in.
So in a sense the creators of the absolute time playback are creating in a relative time container. This leads to a significant question:
Is it appropriate for editing with a sense of timing and rhythm?
David
AllinOneFilms.com -
A question: what are the ramifications of this cutting segments and ‘projects’ to time. The broadcast slot is, of course, absolute ( in more meanings than one 🙂
David
AllinOneFilms.com -
Well that certainly shut everyone up….at least for a while.
Thanks David, I look forward to pondering the ramifications through with you and others here.
David
AllinOneFilms.com -
Hey Olof, uless it was someone high up in the company, the “tech” was just expressing his personal opinion. In their official statement there are no conditionals, ifs or buts in the sentence “FCP X does not support professional video monitoring”…a statement they haven’t seen reason to change on their site, on the off-hand chance that it was said in aerror.
It’s hard to believe they hadn’t seen the API or at least talked to Apple before making that statement. To me, and I could be wrong, that means that the API for FCP does not yet support 3rd party external monitoring. And that means it could be a longer wait than some here anticipate.
David
AllinOneFilms.com -
[Sean Thomas] ” Go to the Premier forum where all the PRO’s hang out.”
[Sean Thomas] “Yea – I’m just throwing some of the crap back.
You must have missed my :-)”
It seems I missed it for the reason that it wasn’t there.
[Sean Thomas] “n most forums that means: joking, funny, tongue in cheek. I’ll do a form search to see what it means here. :-)”
Apparently you think a smiley masks ‘snide’.
David
AllinOneFilms.com