FORUMS: list search recent posts

Towards a better NLE Part 2 - A conceptual framework

COW Forums : Apple Final Cut Pro X Debates

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
Simon Ubsdell
Towards a better NLE Part 2 - A conceptual framework
on May 18, 2018 at 11:57:07 am
Last Edited By Simon Ubsdell on May 18, 2018 at 12:36:04 pm

Most of the time, when talking about improving the NLE concept, we talk in terms of tweaking existing tools and modes rather than addressing fundamental design questions. But I think there remain major design issues that could usefully be addressed by different ways of thinking.

If I were designing a new NLE, the first thing I would look at is finding a way to break up the old-fashioned duality of the Browser/Timeline model.

Apart from convention, there is no good reason for the separation of these two functions, but all NLEs treat the separation as a given. The result of this model is that editors are encouraged to work serially - organising in the Browser beforehand then moving the organised material to the timeline.

What if instead we opted for a model that maximises parallel working and created a a holistic environment where these two processes were not separated?

Imagine the "browser" being inside the "timeline". (I'm putting the terms in inverted commas because I am imagining a completely different sort of environment that effectively supersedes both.)

I'm thinking of an editing space that's a lot more sophisticated than the kind we are used to and which is much more an interactive graphical arena, an "artboard" with deep contextual options.

Instead of thinking of one place where clips are sorted and another place where they are arranged sequentially, think instead of a master environment where clip selection and clip sequencing happen together.

Instead of a one-way flow from browser to timeline, think of a seamless sharing between the two.

For instance, every clip or subclip item would contain (via a contextual pop-up) the option to access any other similarly tagged clips or subclips and any or all additional information added to them.

Tagging and adding info can be performed anywhere including within a timeline.

Essentially each "clip" becomes a "container" regardless of where it currently lives. This is not to say that you wouldn't have higher level containers but rather to point to a different way of looking at what a "clip" might be in practical terms.

The key to this is a contextual model where any object can be a "container" for additional data and relationships.

Think also of a space where multiple timelines co-exist, master timelines as well as subsidiary mini-timelines, breaking down the arbitrary distinction between a clip that hasn't yet been "sequenced" and those that have.

Another area I would look it would be to have the interface reformat itself intelligently depending on the operation that you are performing. The fact that we still have fixed workspaces that you have to manually select seems antiquated. Depending on what you are doing you need certain interface items to be larger or smaller or hidden or unhidden and this should happen without you having to think about it. If something like this were implemented it would free up a lot more possibilities of the kind I am advocating here.

Obviously there are some features of this that will be partially echoed in your own favourite NLE right now (especially if you are familiar with Smoke), but it's not individual features I am talking about. It's a radical rethinking of the dual mode archetype that dominates most NLE design at present.

Moving towards a holistic model, however that is implemented in practice, surely has to be a desirable goal.

Anyway, just a few random thoughts for a Friday lunchtime. I'm sure you can all do a lot better so let's be having you!

Simon Ubsdell
tokyo productions
hawaiki


Return to posts index

Simon Ubsdell
Re: Towards a better NLE Part 2 - A conceptual framework
on May 19, 2018 at 9:00:33 am

Another fundamental issue that a reimagined NLE could address is the dominant clip>timeline model.

All conventional NLEs presuppose that editing revolves around moving clips from a container to the timeline - hence the rigid and immutable hierarchy of the two Browser/Timeline spaces (see above).

Not one of them is optimised for what is in fact a very common and crucial editing model - and that is moving clips and groups of clips between timelines.

Some of them do have features that partially enable this process, others make it really quite difficult, but none seeks to provide the maximum possible facilitation of the process.

I have talked before about the creative and operational advantages of thinking in terms of timelines for organisational purposes and I don't want to reopen that conversation here, but I want to stress that while there are many editors who apparently never edit between timelines (!), there are many others for whom it is a fundamental requirement of the kind of editing that they are engaged in. It is absurd that NLE makers haven't grasped how important this is and offered tools and modes to make it fast and easy.

I would also observe that the way NLEs are traditionally conceived creates an unconscious bias towards thinking that certain ways of working are "the right way" to work, and this is very much true here. You see a lot of people who fetishise the inefficiencies of the clip>timeline model as if they were virtues. It is hard for them to see that there are other options because of the conditioning engendered by the design of the tools.

And this is why it's valuable to keep thinking of ways to disrupt the conventions of how the NLE is conceived.

Simon Ubsdell
tokyo productions
hawaiki


Return to posts index

Oliver Peters
Re: Towards a better NLE Part 2 - A conceptual framework
on May 19, 2018 at 11:53:25 pm

[Simon Ubsdell] "Not one of them is optimised for what is in fact a very common and crucial editing model - and that is moving clips and groups of clips between timelines."

While not necessarily optimized, Resolve 15 has formalized the "pancake" timeline structure. That's in addition to its timeline "changes" comparison view.

- Oliver

Oliver Peters - oliverpeters.com


Return to posts index


Oliver Peters
Re: Towards a better NLE Part 2 - A conceptual framework
on May 19, 2018 at 11:59:38 pm

BTW - anyone remember xm|edit Traffic?

http://www.studiodaily.com/2006/04/xmedit-traffic/

- Oliver

Oliver Peters - oliverpeters.com


Return to posts index

Simon Ubsdell
Re: Towards a better NLE Part 2 - A conceptual framework
on May 20, 2018 at 1:06:10 pm

[Oliver Peters] "While not necessarily optimized, Resolve 15 has formalized the "pancake" timeline structure."

Good point. But I don't think even the most enthusiastic adherent of the pancake method would think of it as anything better than a kluge dictated by the current limitations of NLE design.

To continue with the theory behind this ...

Editing, and indeed all film-making, is about the journey between shots. Every cut is not just a cosmetic change of perspective, it moves the viewer further forward into the story.*

Every film is itself a journey from the first shot to the last, and within that overarching journey there are many sub-journeys.

And as editors we need to be building those sub-journeys with every bit as much care as we build the master journey. The sub-journeys are the true bulidng blocks of our edit.

An individual clip is essentially lifeless and inert - it's only when it is brought into conjunction with another clip that it acquires the dynamic energy that makes it a meaningful part of the story (or acquires the potential to become a meaningful part of the story).

The notion that clips themselves are the fundamental building blocks of story-telling is a very unhelpful one and it masks this important insight.**

An editing model that sees editing as merely the transfer of clips to the finished timeline (without recognising the essential importance of sub-journeys) is a profoundly misleading one.

But this is the dominant NLE model and it's one that many editors buy into unthinkingly.

What I am looking for is an NLE that maximises the ease of creating and manipulating the sub-journeys that are the lifeblood of the edit.

(* Obviously even a single shot film is a journey between "shots" (or visual states perhaps?), with the simple difference that it is conceived in camera.)

(** There are many editors who miss this aspect of the job and whose work is essentially just a process of wallpapering over the timeline until it's time to stop. It's as dull as ditchwater to watch the results.)

Simon Ubsdell
tokyo productions
hawaiki


Return to posts index

Steve Connor
Re: Towards a better NLE Part 2 - A conceptual framework
on May 20, 2018 at 1:24:09 pm

[Simon Ubsdell] "The notion that clips themselves are the fundamental building blocks of story-telling is a very unhelpful one"

Is that an actual notion? Clips are just containers, they're NOT the building blocks. The shots IN the clips are the building blocks surely? Clips are a convenient way of storing shots and in fact in FCPX with the skimmer, my entire library of clips becomes one single "clip" with all my shots within. In other NLE's a stringout timeline does the same thing.


[Simon Ubsdell] "What I am looking for is an NLE that maximises the ease of creating and manipulating the sub-journeys that are the lifeblood of the edit."

How would this be achieved apart from using AI to actually perform some or all of the editing for us?


[Simon Ubsdell] "An editing model that sees editing as merely the transfer of clips to the finished timeline (without recognising the essential importance of sub-journeys) is a profoundly misleading one.

But this is the dominant NLE model and it's one that many editors buy into unthinkingly."


Return to posts index


Simon Ubsdell
Re: Towards a better NLE Part 2 - A conceptual framework
on May 20, 2018 at 1:32:38 pm

[Steve Connor] "Is that an actual notion? Clips are just containers, they're NOT the building blocks. The shots IN the clips are the building blocks surely?"

The nomenclature has become ever more hazy since the introduction of non-linear editing (was that last year or the year before? I can't remember) but you are introducing an unnecessary level of confusion here.

Anything sitting alone in the Browser is a "clip" - lifeless and inert until it is actually joined by another "clip".

That's the point I am trying to emphasise.

Skimming across clips doesn't render them any less lifeless and inert, for the simple reason that no conjunctions exist.

Simon Ubsdell
tokyo productions
hawaiki


Return to posts index

Steve Connor
Re: Towards a better NLE Part 2 - A conceptual framework
on May 20, 2018 at 1:39:25 pm

[Simon Ubsdell] "Anything sitting alone in the Browser is a "clip" - lifeless and inert until it is actually joined by another "clip".

That's the point I am trying to emphasise.

Skimming across clips doesn't render them any less lifeless and inert, for the simple reason that no conjunctions exist.
"


So what would improve this? Do you feel somehow constrained by the current NLE models?


Return to posts index

Simon Ubsdell
Re: Towards a better NLE Part 2 - A conceptual framework
on May 20, 2018 at 1:48:52 pm

[Steve Connor] "So what would improve this? Do you feel somehow constrained by the current NLE models?"

Yes, see above.

The best kluge that current NLE models offer is the pancake method and it's a crappy kluge.

There's obviously a better way of doing it ...

Can think of it yet but give me till teatime and I'll come up with something.

Simon Ubsdell
tokyo productions
hawaiki


Return to posts index


Oliver Peters
Re: Towards a better NLE Part 2 - A conceptual framework
on May 21, 2018 at 11:47:18 pm
Last Edited By Oliver Peters on May 21, 2018 at 11:56:30 pm

FWIW - In this video starting at about 13 min. you can see how Avid handles the bins visually, which is quite a bit different than FCPX or Premiere Pro.







FWIW - this is the same Josh:

http://www.fcp.co/final-cut-pro/articles/2070-josh-beal-reconsidering-final...

- Oliver

Oliver Peters - oliverpeters.com


Return to posts index

Herb Sevush
Re: Towards a better NLE Part 2 - A conceptual framework
on May 25, 2018 at 11:56:26 am

Simon -

I've been following this thread but I've been out shooting for awhile and haven't had time to put my thoughts together, but this morning we have a late call, so here goes.

Unlike some, I'm quite comfortable with the "bin and timeline" paradigm that has evolved in NLE editting. I have a long list of irritating bottlenecks to my work, but all of them could be addressed within the current methodology, so I'm not looking for any new workflow revolution.

What I am looking for is a totally different editing environment, where editors like me have direct and continuous feedback to system designers, where both designers and users are part of the same community. I need to feel like someone is listening when I'm talking and where not only do I get answers, I get explanations when the answer is "no."

I've had something near this type of relationship a few times in the past and I can't really quantify how much pleasure it brought to my work-life but I would give up a lot of technical innovation to be working within that sort of environment again.

I've spoken of working with *edit before, but during the turbulence of it's demise many of the members of the Cow's *edit forum made a concerted effort to form a co-op with the system designers to buy the software from Autodesk, an effort that failed quickly and quietly, but was the kernel of an editting "revolution" I would gladly sign up for.

For now I am content to anonymously utilize whatever tools the market spews out, separated from the source of my code by layers of bankers, market researchers, lawyers, and other assorted twits, where my biggest choice is whether to use bins or keywords.

Herb Sevush
Zebra Productions
---------------------------
nothin' attached to nothin'
"Deciding the spine is the process of editing" F. Bieberkopf


Return to posts index

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
© 2018 CreativeCOW.net All Rights Reserved
[TOP]