View Full Version : ultimate AE

12 December 2009, 11:37 PM

Ultimate AE is my attempt to influence the AE public, and therefore the development of AE CS5.

Like a lot of shops in the business of creative, mine relies on AE for itís design work. Because of the ubiquity of the software we can easily contract out to different artists as needed, and assemble shots in-house for final output. Workflow-wise AE is pretty much an unmatched creative tool on the desktop, but itís facing some new competition.

This year Autodesk announced Smoke on the mac. Smoke has some better tools than AE in several areas, but the main difference is workflow. Particularly three things: nodal compositing, disk based playback options, and 3D.

We canít expect AE to make up all the gaps. After all, a complete AE system will run you between $5K-$8K, and a Smoke is more like $30K for the same specs. But we can expect the good folks at Adobe to help make up some of the difference. If not shops like mine will start thinking $30K is pretty good deal.

Ultimate AE is a simple paint over designed to illustrate the workflow for the two biggest drawbacks in After Effects: lack of disk-based playback, and lack of good 3D compositing.

1.The Timeline

The first thing to add is an NLE style timeline that can be used to edit projects and render them to disk. Why would this be such a big deal?

Well, first off it lets you map out your projects using something like an NLE timeline, the best kind of cutting timeline for everything from music videos to feature films. This addition alone makes After Effects much more viable as a finishing tool. Add effects to your timeline comps and you can color correct to your hearts content. Sync that timeline to your layer view and you can easily track your animations against an audio waveform.

More valuable is the ability to render this timeline to disk. Premiere and Final Cut render preview files to disk using pre-determined codecs. You can render whatever your system will play back, and once youív rendered a timeline you can play it back whenever you like. You only every need to re-render the parts you change. This allows After Effects to reach its fluid creative potential.

Last, but certainly not least, would be the ability to commit the comps to disk. That would allow you to save all of your animation out, and work with speedy files for finishing.

More than nodes, more than 64 bits, more than a proper 3D workspace, the integration of the NLE timeline into AE is the single biggest thing Adobe could do to close the gap between AE and Smoke.

Note: this is exactly where combustion was headed, so itís not surprising that Autodesk has let it slip into obscurity.

2. 3D Workspace

A real 3D workspace is a necessity for a motion graphics application. And AE is, primarily, a motion graphics application. Those of us who make our living in this field know how competitive it can be, how short the deadlines are, and how efficient you have to be to succeed. Sure, you can lay out camera paths in a 3D program and bring them into AE, but why do that when you should be able to accomplish this inside one package?

Fusion and Motion both have pretty solid 3D environments. In Fusionís case this includes support for 3D objects without the use of Photoshop as a bridge. This is really a requirement for modern motion graphic design, where effects, graphics and live footage are often integrated. Making the 3D environment as complete as possible, with adjustable motion paths and predictable behaviors, is also a necessity in a world where changes come often and turnarounds are short.

The Rest of the Shortlist

Those are the two most important things. Here is a quick list to round out the top ten.

Tracking and Stabilization - everything, and I mean everything, is better at this than AE.
Rotoscoping and paint - shake has about the best roto and paint on the desktop, and AE would do well to copy and paste. Combustion is a good template too.
Schematic view - AE is not a nodal compositor, it probably never will be, and it probably doesnít need to be. But a functional schematic view would be very handy (combustion is once again a good model to follow).
Graphics acceleration -slated for the next version, and very welcome.
Nucleo - tied right in to the guts, with true multithreading.
64 bit - again, itís coming, but it wouldnít be anywhere near as big a deal if disk based workflows were in place. In fact, it probably wouldnít be a big deal at all.
Colorista - donít bother trying to make it. Just license it.
32 bit everything - itís long past due.

As a professional I have no problem paying $3500+ a seat for a version of AE that does whatís on that list. I also have no problems switching to another platform, and Iíd certainly consider paying the $20,000 that Autodesk is asking for a license of smoke. Why? Because that money is an investment that helps my company maintain itís billing rate and saves production time.

These are the features that would make commercial, music video, and title production quicker and more efficient. Thatís where I think AE is strong and needs to be stronger. What do you think?

Debate away.

12 December 2009, 04:50 AM
As interesting as forum conversations may be, they aren't the best way to actually make requests. The best way to influence the After Effects teams is to submit feature requests (

12 December 2009, 09:34 AM
Ultimate AE is my attempt to influence the AE public, and therefore the development of AE CS5.

You mean CS7. Sorry to burst your bubble, but software development is just as slow-paced as any other product development. New car models don't fall out of the sky overnight, too... Based on various blog-posts and info snippets from Adobe (presentations at MAX and other shows), you can be reasonably sure that CS5 is done.

For the rest - well, "engineering by comparison" is not getting anywhere and just citing the ever same shortcomings over and over again doesn't help anyone, either. Honestly. Seriously. The Adobe people are not that shut off from the world to not realize this on their own. Sorry if I put on my ugly software engineer mask, but what does "support for 3D objects" mean? This is at best vague, at worst a useless platitude. That one feature alone could be disected into a 2000+ page feature description and implementation spec and keep a 100 coders busy for years. So where does it start, where does it end? In order to really help, you'd have to be much more specific. Annoying as it is, it's all in the details and let me tell you this: Once you start thinking about such things real hard, you quickly realize the intricacies involved. What seems obvious can quickly lead to a dead end road and having one feature can completely destroy another.

That aside, threatening/ pondering about the potential mass-swing to other softwares is utterly pointless and I think it nicely contradicts your own claims - if Fusion's 3D really were that essential, then why don't we all use that program? Yes, some 3D is needed by everyone, but not everyone needs FBX import when he has no way of creating those files in the first place. See my point? Even the price discussion is missing the mark - tomorrow someone could wipe the floor with smoke* by offering all your features in a 3000 USD program and then that whole discussion is going nowhere...

If you allow me to say this: You are making too many assumptions in some areas and those are colored and biased by your own work experiences and specific needs. That's not wrong - we all suffer the same and don't know the world outside our own work - but individual priorities vary considerably. Even your "single biggest thing" doesn't give me a tingle in my left toe - as long as AE cannot preview fields, it would be utterly useless to me. First and foremost AE needs a better 2D rendering pipeline and that needs to happen long before we can get lost in fancy features... Back to basics, if you will.


12 December 2009, 05:42 PM
I'm just trying to sprout a seed, not push the whole development in a new direction. CS5 is done? OK. Maybe later.

Fusion isn't bought up for four reasons: price, platform, target market, and current user size. As fabulous a program as LWCore might end up being, no studios are going to start dumping their Maya pipelines when it's finally finished. As great as Pipeline Deals might be, no one's dumping Salesforce. Features don't cause tectonic shifts, and I never suggested that they would.

There is, however, a market change that's opening up as a result of recent economic pressures combined with ever more demanding specs in broadcast commercial production. Shops have to do more with less, and the economics of the pipeline point to the need for smaller numbers of faster seats. Where we used to have one finishing suite supported by five other stations, now we just have the one finishing suite (and that's not my company, that's anyone doing commercial work...except maybe The Mill).

When you look at pipeline choke points, the features I've highlighted deal with the most time-expensive. More or less in order. 3D object support is not the most important thing. Interactivity is.

For feature request forms to work I'd assume that there needs to be a decent number of requests. There needs to be some kind of consensus that lines up with software development budgets. That means that users need to be aware of their options: if all you know is RAM cache, all you're going to want is more RAM. AE users are varied, and have different needs, and not everyone is going to see the same solutions. Pointing out workflows that solve problems doesn't hurt, and if more people see that as an improvement for them, they'll ask for it (surfing shows some others have made similar comments here ( and here ( me at the second one).

Disk based playback improves workflow for indy film makers, commercial post, compositing, and visualization. I'd guess that's the largest improvement for the largest potential number of users.

And if you're hinting that you've made a $3000USD program that can mop the floor with there a demo? Can we sign up for a beta somewhere? ;)


12 December 2009, 08:36 PM
Shops have to do more with less, and the economics of the pipeline point to the need for smaller numbers of faster seats.

I don't necessarily agree. Many shops get themselves in this under-bidding spiral of death and then naturally you can only try to rationalize and outsource the hell out of it. A lot of those "pipeline economics" are simply shifted down in the chain.

Interactivity is.

A rather ambiguous term just the same as performance, stability and reliability are. Everybody uses them, but nobody really ever explains what they mean to them. There's just a million ways that can be seen and when to one person it means interruptable processing so people can interactively scrub through the timeline every time, to another it merely means better 3D navigation. Even then it becomes a question of where's the tradeoff - would calculating textures at lower res still be acceptable just to move your 3D camera more quickly? I'd be wary to use such terms. The one thing we all can agree on is that AE sucks in maxing out system resources. That alone would go a great way toward the aforementioned buzzwords having more meaning.


CGTalk Moderation
12 December 2009, 08:36 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.