[Greg Sage]

OK, thx. Will give cc Wide Time a shot.

Any other general approaches to smoothing out video filters like cartoonizers with sharpening or others that tend to get jumpy due to exaggerating changes from frame to frame?

[Greg Sage]

or… is there perhaps some other sort of time smoothing pre-filter approach I should also be investigating?

[Greg Sage]

Well, the only audio footage is 44.1, so I figured I’d just change the project to match, and then do the conversion on output via ME.

If there’s a specific reason not to do it that way, it’s no big deal to switch.

I’ve exhausted the tricks I remember. 1/8 adaptive, tried changing buffers, tried matching or not matching audio sample rate, tried just wireframe, cache settings, and probably a few other things. Currently got 32GB of ram with 6GB allocated to other programs, and nothing else running.

The GeForce 980ti is new in the equation too, and obviously not mac native. Looks fine as far as I can tell from checking hardware settings, but possible it’s not being properly utilized. Again, though, there’s literally no video yet… and it’s still choking.

The only timeline item is an “audio track” which actually a Premiere project I imported that was just an audio track with markers.

Is this inefficent somehow, like maybe I’m better off importing it in some other way (not as Premiere project) where the markers are still readable via metadata.

[Greg Sage]

I’ve worked with a few of them before, but it’s been a while. After some digging, it looks like perhaps the Marker Sync function in Edit Monkey is what I’m looking for.

The vid description is a bit light on that function, but it looks like it auto generates layer segments (pages in monkeyspeak) according to the marker names.

If this is correct, I’d like to make sure I get my markers set up properly first, so…

I currently have a Premiere project with the song audio as the only item in the timeline, and there is a comment sequence (not region) marker at the beginning of each word.

Is it optimal to then just import this Premiere project into AE, and launch the Marker Sync function in Edit Monkey, or would it be better for some reason if they were region markers, or if it were imported into AE in some other way, etc.

If the markers should be done differently, like region instead of sequence markers, for instance, is there an automatic way to take what I have and convert it into whatever format will work best for the import into AE and auto text layer creation?

[Greg Sage]

Dug this up:

https://forums.creativecow.net/docs/forums/post.php?forumid=227&postid=32398&univpostid=32398&pview=t

Seems quite similar, though not sure if that was a different type of markers as he’s talking about within marker duration when mine have no duration, as they’re just indicating the start of each word. So, seems, maybe closer to this where he’s talking about “hold” type behavior (marker triggers text layer which persists until next one is triggered)

https://aenhancers.com/viewtopic.php?t=847

Answer given for that was:

m = thisComp.layer("movie layer");
n=0;
if (m.marker.numKeys > 0){
n = m.marker.nearestKey(time).index;
if (m.marker.key(n).time > time) n--;
}

if (n > 0) m.marker.key(n).comment else ""

I’m not quite following, though, where such code would be pasted. I’ve done some basic js in to drive various parameters before (like driving the scale of a layer by entering code in the field for scale on that layer). Not clear where the code here belongs, though, as it’s driving something outside the layer.

[Greg Sage]

I’ve only used the typemonkey stuff and done some vids manually.

I see there are a number of solutions out there, but I need to pick one quickly and get up to speed, so hoping to hear some feedback on any of them from anyone who has used them.

Also, any tips/tricks for optimizing the workflow that may not be immediately obvious.

[Greg Sage]

Ah. OK. Much less confusing.

[Greg Sage]

Sure. Here it is.

I’m using the magic symmetry plugin, but it instances the symmetry object along with it’s controls, so I think it will show up fine for those who don’t have it. Not used to sharing c4d files, so not sure if there are texture path issues or something I can fix on the sending end.

EDIT: I might as well ask. I thought I just hadn’t had enough coffee last time. What is up with the system to post a file? It basically forces me to create a new post for the file… and never even brings me back here. Very weird. I end up having to backtrack through history and tabs to get rid of that “post”, copy and paste the link manually into (this) original post already in progress. It’s either broken or just the most anti-intuitive file posting system I’ve ever encountered on a forum. Does it just somehow work out despite appearing to force me to make a separate post that’s detached from the OP?

Anyway, I’ll manually post the link here:

12916_trumpprebake.c4d.zip

[Greg Sage]

OK. Will look into that. Thx.

[Greg Sage]

It’s specifically to stop them from being on top of each other so I get individual balls coming out rather than blobs. You are correct that self collisions are the only thing at issue… but while I can’t avoid them occurring due to the emitter size, I’m trying to avoid SEEING them once they clear the emitter.

After extensive trial an error, I believe it was changing the “time constant” to 0.1 that more or less got the post collision speed to match pre-collision. It also stablized the crazy mismatch in rotations pre and post that becomes obvious once the spheres are texture wrapped. That much is stable, but I still haven’t been able to fully fix the issue of post-collision objects breaking the emission plane.

Since it’s such a simple scene and just being used as one layer in an AE comp, though, I faked it by just putting lots of distance between the objects so it effectively keeps the balls from breaking the foreground or background planes. Good enough to move forward, though I would always like to learn why something happens for next time.