-
FCPX and Fusion
-
[Bret Williams] “I’m no expert on nodes, but that’s what I always wondered. Looks great for routing and connections due to is visual nature of showing flow. But it does seem to be pretty lacking in showing the temporal. Keyframes, stops and starts, etc. Maybe better for color correction and compositing but less so for motion graphics?”
This is an argument I always have a lot of problems coming to terms with, although you hear it a lot.
Fusion (as an example of a node-based compositor) has a deep and powerful timeline structure where temporal manipulation is every bit as easy as it would be in After Effects or similar (as in this extremely simple example that I happened to have to hand):
Adjusting keyframes (and the timing of clips) in Timeline mode is exactly the same as it would be in After Effects.
And of course, you also have a Spline view where again there’s really no difference from Ae:
It seems to me that temporal manipulation in Fusion (for example) is every bit as fancy as anyone could want it to be. It’s not somehow less functional because it’s nodal – it’s every bit as functional a vast limitless amount more.
What am I missing? To me the “differences” seem purely semantic and subjective.
(And no, I’m not back on this forum – I’m only here because the discussion is about Motion and Fusion.)
Simon Ubsdell
tokyo productions
hawaiki -
[Simon Ubsdell] “It seems to me that temporal manipulation in Fusion (for example) is every bit as fancy as anyone could want it to be. It’s not somehow less functional because it’s nodal – it’s every bit as functional a vast limitless amount more.
What am I missing? To me the “differences” seem purely semantic and subjective.
(And no, I’m not back on this forum – I’m only here because the discussion is about Motion and Fusion.)”
I was hoping you might show up for this one, Simon. ☺
Maybe it’s just my ignorance – how would you (for example), call up all (only) opacity keyframes in the composition?
Shawn
-
[Shawn Miller] “Maybe it’s just my ignorance – how would you (for example), call up all (only) opacity keyframes in the composition?”
There are loads of sophisticated “view filter” options that would enable this sort of thing in both the Timeline and the Spline views.
Obviously they don’t match After Effects in every single respect but they are are pretty deep and flexible.
I come back to the point that I don’t think that there’s a fundamental conceptual difference between the two models, except that exclusively layer-based solutions have intrinsically fewer options … rather than the other way around, which is how I think most people think of it.
Simon Ubsdell
tokyo productions
hawaiki -
Obviously there are slight differences of emphasis as you would expect between any two compositing solutions but I’d suggest it’s much more true to say that After Effects is a compositor without nodes than to say that Fusion is a compositor without layers.
To me the real difference still lies in the fact that After Effects is geared more towards motion graphics and Fusion/Nuke are dedicated compositing solutions. The node/layer distinction is a relatively trivial difference by comparison.
I would add that transforming Fusion into a more motion graphics-friendly offering would be a much simpler proposition than converting After Effects to a compositor that could genuinely compete with its node-based rivals. That said, there is a vast overlap between what you can actually achieve in practice with a bit of flexibility.
But this is where we really need Walter to illuminate the bigger picture and we don’t have him any more.
Simon Ubsdell
tokyo productions
hawaiki -
[Simon Ubsdell] “[Shawn Miller] “Maybe it’s just my ignorance – how would you (for example), call up all (only) opacity keyframes in the composition?”
There are loads of sophisticated “view filter” options that would enable this sort of thing in both the Timeline and the Spline views..”
Yep, I think this is the more common way that animation packages approach graph editors and dope sheets (Fusion, C4D, Maya, Nuke, Lightwave, Blender, etc). I think the Adobe way of handling keyframes is somewhat unique to AE (single and double click access to keyframes in the layer stack)… I wish Premiere Pro operated same way.
[Simon Ubsdell] ”
Obviously they don’t match After Effects in every single respect but they are are pretty deep and flexible.”I don’t doubt that at all, however I think this is why so many motion graphics artists gravitate toward AE… they put up with a lot of crap for fast keyframe animation… well, that and the tons of available plugins, scripts, training, etc. But I get your point, Fusion et. al are every bit as powerful when it comes to mograph creation.
[Simon Ubsdell] ”
I come back to the point that I don’t think that there’s a fundamental conceptual difference between the two models, except that exclusively layer-based solutions have intrinsically fewer options … rather than the other way around, which is how I think most people think of it.”Agreed!
Shawn
-
[Shawn Miller] ” I think the Adobe way of handling keyframes is somewhat unique to AE (single and double click access to keyframes in the layer stack)”
Agreed.
There’s been a lot of optimisation in Ae over the years that makes some things like this a bit quicker, although I would argue that it’s very much stuff that’s specific to mograph rather than compositing.
But I think I’d stick by my contention that these are really (relatively) small differences of emphasis rather than major conceptual divergences. Wouldn’t you agree?
I find it very odd that many people still seem to think that applications like Fusion “don’t have track-based editing”. That seems to me to be fundamentally untrue.
Simon Ubsdell
tokyo productions
hawaiki -
[Simon Ubsdell] “There’s been a lot of optimisation in Ae over the years that makes some things like this a bit quicker, although I would argue that it’s very much stuff that’s specific to mograph rather than compositing.”
Definitely agree. I think for better or worse, Adobe knows what AE is… a great multi-tool that specializes in mograph. Unfortunately, that seems to mean that they put much less effort into making it a better compositor and color management tool. Can you explain to me why AE doesn’t support compressed CinemaDNG? Adobe can’t seem to…
[Simon Ubsdell] “But I think I’d stick by my contention that these are really (relatively) small differences of emphasis rather than major conceptual divergences. Wouldn’t you agree?”
Absolutely! I almost said something similar in my original response to Bret Williams.
[Simon Ubsdell] “I find it very odd that many people still seem to think that applications like Fusion “don’t have track-based editing”. That seems to me to be fundamentally untrue.”
I might be tempted to say that Fusion and Nuke have layers but not tracks. ☺ Mostly because I see tracks as immovable time-based object containers, and layers as containers for time independent objects. How do you see them?
Shawn
-
[Shawn Miller] “I might be tempted to say that Fusion and Nuke have layers but not tracks. ☺ Mostly because I see tracks as immovable time-based object containers, and layers as containers for time independent objects. How do you see them?”
Yes, it would be completely wrong to say that Nuke/Fusion/AfterEffects/Motion have tracks rather than layers – although of course I just did exactly that ☹
None of these applications allow you to have more than one source image on the same Timeline “lane” – what defines tracks is exactly that they allow multiple image items sequentially on the same “lane”.
(Of course you could create “virtual tracks” in any of these applications by grouping (or in Ae by precomping). You can then manipulate multiple sequential events on the timeline …)
I suppose it’s easy to elide tracks and layers when thinking about layer-based compositing because layer-based compositing feels the same as track-based compositing – what is above is composited over what is below. This is not true of node-based compositing where the vertical timeline order is not relevant. But yes, layers are definitely not tracks.
[Shawn Miller] “I see tracks as immovable time-based object containers, and layers as containers for time independent objects.”
That’s a nice description of tracks, but I’m not sure I quite understand what you mean by a “time independent object”.
In Nuke/Fusion each node (tool) is a layer. The consequence of this is that not only can you change the timing of a clip (in and out points and position in time), you can also move the overall timing of any operation up and down the Timeline as well as its individual keyframes, without having to invoke a Graph Editor.
In Motion you can do the same – filters are objects in the Timeline that can be manipulated in time just like any other layer.
In After Effects to achieve the same result you need to use Adjustment Layers. (Or the Graph Editor, but takes us into different territory.)
In each of these cases moving a layer on the Timeline allows you to move the timing associated with it, namely its animation. The layer contains timing information that travels with it.
But I’ve probably not grasped your definition properly.
Simon Ubsdell
tokyo productions
hawaiki -
One other point that might be worth adding to this discussion is that it’s not necessarily a good idea to focus on one specific application to extrapolate general rules about what is and is not possible for applications of that type.
We tend to use After Effects as the benchmark for what a layer-based compositor can look like, but in doing so we miss things.
And we can see what those are by looking at Motion (originally conceived as a competitor to Ae although that’s an ambition long since abandoned).
Most After Effects users are so accustomed to Precomps that they probably assume this is the only away of approaching grouping. Motion by contrast uses Groups in place of the Precomp concept and it’s a much more open way of working. A Precomp is essentially a closed container and we have to work around that in various ways, whereas a Motion Group is an open container and that offers considerable benefits. In other words, in Motion you can access (not simply view) the overall architecture of the composition without having to dive into closed containers.
You could say this architectural “openness” is a lot closer to a nodal way of working in that the cross-routing of layers and tools is much more instantly achievable.
Another obvious example of this sort of difference is to compare Track Mattes in Ae with Masking in Motion. In Ae a Track Matte source layer has to be adjacent in the layer stack, whereas In Motion you can access any layer or group as the source of the “matte” regardless of its position in the layer stack. It’s a massive difference and much more node-like in practice.
But there’s an even bigger and more powerful difference and that lies in Motion’s use of Clones. These are very little understood even by many Motion users so here’s a basic primer:
A Clone can use both layers and groups as it source, it will update to reflect any changes to anything that is routed to it, and its contents can be reassigned at will.
That might sound a trivial thing but it’s extremely significant in practice. And it’s a very node-like concept in that it allows for very powerful and flexible routing that very much transcends the limitations of the layer-based model.
There are many things to admire about Motion (still, despite its practical obsolescence from a development point of view) but the Clone concept is genuinely a stroke of genius.
All of which is to reinforce the point that there is much more of a continuum between After Effects/Motion and Fusion/Nuke than is generally recognised.
Simon Ubsdell
tokyo productions
hawaikiSome contents or functionalities here are not available due to your cookie preferences!This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.
-
I find it interesting that each clip on Resolve’s edit page timeline could have a Fusion comp associated with it and which can be ‘called’ instantly just by switching to the Fusion page. So, the Resolve timeline is actually a ‘string-out’ or a sequential collection of Fusion node trees.
Besides, the Edit page of Resolve provides a basic track-based compositor for anyone wishing to composite with tracks.
So Resolve-Fusion seems to be the ‘convergence’ (not sure if its the right term) of tracks and nodes. And one can work in either, or both but within the same software.
———————————–
Neil Sadwelkar
neilsadwelkar.blogspot.com
twitter: fcpguru
FCP Editor, Edit systems consultant
Mumbai India
Reply to this Discussion! Login or Sign Up