Adobe After Effects Help Thread

[_Anakyn_]

Sure, pichus, I'd like to try that. I'm looking forward to testing it

[pichu]

well, i got the script coded, but i need to test it soon -- and I've decided to create text selectors based on Karaoke style2 (hold keyframes are more reasonable to parse from there)

I'm not sure how do you use multiple range selectors in karaoke. The only time I used multiple range selectors is when I want the syllables to overlap each other on one layer... i.e. \k-negative timings (something that's not parsed here) [and because of this, I'm quite hesitant to furnishing one]. The multiple range selectors will have names like: Karaoke selector1, Karaoke selector2... each selector represents only one syllable to be safe. Do tell, if this will work?

So it might be a bit more challenging to get this to work correctly. :/

Commentary Draft:

/*

	Karaoke Auxillary Selector Generator based on Shinjinpierre's script
	
	Input explanations: [name]=[value]; [name2]=[value2]
	
	[name] Explanations:
	
	name = Keyframe style Name (default: Keyframe stylen, where n is an unused number)
	in = lead-in times
	on = Syllable ON times
	out = lead-out times
	[Notes] on the times:
		negative values are normalized to the syllable timing
		positive values are absolute timings
		Multiple times can be inputted, separating by spaces.
		You can define the percent values fpr the start and end at the keyframe using ^ and $ after the timings.
			Without these values, linear interpolation will be set for these values.
			Example: -0.5^50$75 says that at 50% of the normalized time, set the start to 50% distance and end to 75% of the distance.
	[Example]:
	name = Keyframe Cool Style; in = 0.5; on = 0^0$100 -0.5^50$100; out = 0.5
	In the example above, the syllable will have a lead-in of 0.5s with relative syllable effect at 0%, 50% syllable time, and lead-out of 0.5s.
*/

It's about a hundred lines of codes.

[shinjipierre]

You can't just copy that animator to every other layer, because then the timing is way off..

It would be used the same way as the usual karaoke text animators...
What I meant is to add that text animator within my script, you would only have to copy the properties, not the selectors, or make an expression.

That's why I said it has to be coded...

[pichu]

Ok, I wrote a script ... I never tested it, because I'm not sure how you guys really did this....

But, using them in expressions is possible, since you can just go through while loops starting from 1. The timing should be accurate, but overlapped effects may not work, and I have a feeling that using range selectors alone isn't enough, maybe I have to use text selectors instead? But, that's easy to take care of. Just want to know if this is feasible.

http://pichu.org/afx/SSA_aux_0.1.jsx

But I doubt this will work; it's just a draft.

Edit: Okay, I tested it just now with basic blurring; it works just fine with a basic text animator effect; I had to fix some parts that are wrong. The "Effect Selector" isn't needed, so I commented that out. Feel free to check if this script is of any use.

[dj_tjerk]

Originally Posted by shinjipierre

It would be used the same way as the usual karaoke text animators...
What I meant is to add that text animator within my script, you would only have to copy the properties, not the selectors, or make an expression.

That's why I said it has to be coded...

I usually copy animators, not properties.. copying properties and pasting it into the right animator is more work then copying an animator and pasting it into every other layer you want it to have.

[shinjipierre]

I usually copy animators, not properties.. copying properties and pasting it into the right animator is more work then copying an animator and pasting it into every other layer you want it to have.

well, make an animator, link it to the previous text animator with an expression and that's it ^^. anyway...

[dj_tjerk]

And then you get exactly what I already made with keyframe style 5 (that expression selector) in my script; I also link the values of the properties in the animator for that one because I have to..

[pichu]

so no one bothers checking my jsx script if it's of any use? :/
oh well, i guess i'm just wasting my time here.

[_Anakyn_]

I checked it a while ago and it seems to do the trick... partially, because, considering I apply a position property, syllables don't start at the same position. You'll understand it better here:

http://img365.imageshack.us/img365/1...sscriptbi7.jpg

Why does that happen? D: Or is it me?

This is what I do: I load Shinji's script, I select one layer, I load pichus' script and a brand new 'Keyframe style 7' pops up. I add a position property and turn off the Effect selector to see the changes. Then all the characters are displaced. However, the characters (why characters instead of syllables?) start moving 0.5 sec before the effect starts. The script gets the fixed length, but it messes up the whole line D:

[pichu]

Originally Posted by _Anakyn_

I checked it a while ago and it seems to do the trick... partially, because, considering I apply a position property, syllables don't start at the same position. You'll understand it better here:

That's the thing... I wasn't too sure what kind of effects you're trying to apply. If you need positional properties, you may need to create many text layers for each syllable, since the positions will be correlated if you use only one text animator. Then again, it could be part of your effects, who knows

http://img365.imageshack.us/img365/1...sscriptbi7.jpg

Why does that happen? D: Or is it me?

This is what I do: I load Shinji's script, I select one layer, I load pichus' script and a brand new 'Keyframe style 7' pops up. I add a position property and turn off the Effect selector to see the changes. Then all the characters are displaced. However, the characters (why characters instead of syllables?) start moving 0.5 sec before the effect starts. The script gets the fixed length, but it messes up the whole line D:

I noticed I made a fatal mistake in my codes after I posted that first post. I assumed you didn't get the updated one. It should be based on 'word' instead of 'character.'

http://pichu.org/afx/SSA_aux_0.1.jsx

Thanks for the reply.

---

Edit: you might also want to modify the input parameters. I just made those numbers up as an example.

---

Edit #2:

Try this parameter:

on=0^0$0 0.25^0$100 0.5^100$100

This will reach a full stop at 0.5s, and full syllable at 0.25s (required)

If you want the positions to stay after 0.5s, you can use the following:

on=0^0$0 0.5^0$100

(I could make a GUI out of this, but I'm a command-line person, so I'm used to this way :/)

[_Anakyn_]

Gomen, I didn't see your edit in the previous post. I think it works fairly well now. I only have to spend some time to unlock its misteries. Thanks to you, I'll be able to play quite a bit with karaokes. Thanks dude ;D Keep the hard work!

[shinjipierre]

Hi,

I've created a new version of my ssa -> after effects script.

1. There shouldn't be any problems with {\k0} anymore.
2. I've added fixed length text animators for the ones who want to add an effect before or after the syllable is sung.

You can find the new version on my portfolio : www.remipierre.fr

Tell me if there are problems with it or if there's something missing, I haven't made a karaoke for a long time anyway... I didn't really test it.

Shinjipierre

[FAiM]

Hey,

I typesetting Soul Eater and I stuck on logo effect...
so, if someone could me help I will be grateful...
SE-logo.mkv

[Yuudai-kun]

Where are you stuck?

I don't know what problems you have, but it seems fairly simple. (At least if you compare it to other signs).

For the "fire", i think Fractal Noise+Displacement Map+Color Correction+some Glow should be all you need.

As for the last seconds use a Radial Blur effect and be sure to change the method.

Most of After Effects is trial and error, so if you have any problems try playing around with plugins

[pichu]

Originally Posted by shinjipierre

Hi,

I've created a new version of my ssa -> after effects script.

1. There shouldn't be any problems with {\k0} anymore.
2. I've added fixed length text animators for the ones who want to add an effect before or after the syllable is sung.

You can find the new version on my portfolio : www.remipierre.fr

Tell me if there are problems with it or if there's something missing, I haven't made a karaoke for a long time anyway... I didn't really test it.

Shinjipierre

Also, does it work with negative timings? like \k-10... it's been used to time in aegisub to overlap the singing.

[jfs]

Actually the correct way (according to Gabest, if we should take his code for his word) is not to use negative \k timings, but to use the \kt tag. It explicitly sets the start time (in centiseconds) of the next syllable that starts.

[shinjipierre]

What the heck ? negative timings ??? overlap the singing ??? never seen aegisub

I don't really work for fansubs anymore... I don't quite follow the news...

I didn't understand the fucking use of a {\k0} already.... but a {\k-10}, *erm*....
I've just corrected the {\k0} problem since a lot of people were coming in my channel saying their script couldn't load, it probably doesn't work with negative things... don't even understand the use....

[getfresh]

Originally Posted by shinjipierre

What the heck ? negative timings ??? overlap the singing ??? never seen aegisub

I don't really work for fansubs anymore... I don't quite follow the news...

I didn't understand the fucking use of a {\k0} already.... but a {\k-10}, *erm*....
I've just corrected the {\k0} problem since a lot of people were coming in my channel saying their script couldn't load, it probably doesn't work with negative things... don't even understand the use....

{\k0} is useful what you have lines that through scene timing start after the actual start, or depending on the version of textsub you are using or if you are using {\K} only effects you do not always want non-alpha characters getting effected, like " ", ".", "?", etc... so you set a \k0 to them. Otherwise all you need to do is a {\r}. This is of course unless there is some new trick with aegis I have no clue about which is possible.

The diff between

{\K50}yo {\K50}man

and

{\K50}yo{\k0} {\K50}man

is quite noticable

[pichu]

Originally Posted by shinjipierre

What the heck ? negative timings ??? overlap the singing ??? never seen aegisub

I don't really work for fansubs anymore... I don't quite follow the news...

I didn't understand the fucking use of a {\k0} already.... but a {\k-10}, *erm*....
I've just corrected the {\k0} problem since a lot of people were coming in my channel saying their script couldn't load, it probably doesn't work with negative things... don't even understand the use....

Yeah I had to use negative timings in Macross Frontier First opening karaoke for Menclave. It's quite interesting how it was timed... But then I wrote my own script that handles it. It basically uses two range selectors for that, so I asked you that question since fixed timings also used multiple range selectors.

Example:

Dialogue: 0,0:01:43.78,0:01:48.47,Romaji,,0000,0000,0000,,{\ k32}i{\k166}ma {\k-75}{\k45}su{\k168}gu {\k-66}{\k35}TOUCH {\k164}ME

Dialogue: 0,0:01:43.78,0:01:48.47,Kanji,,0000,0000,0000,,{\k 198}今{\k-75}{\k45}す{\k168}ぐ {\k-66}{\k35}タッチ{\k164}ミー

so she sings "ima" first, and then the second voice singing "sugu" overlaps with "ima", and then another voice sings "touch me" while "sugu" is sung. It's two ranges, because "touch me" never overlaps with "ima." Then again, it can be done with ease if the script does multiple range selectors: one selector per syllable. But, you get the idea.

Originally Posted by jfs

Actually the correct way (according to Gabest, if we should take his code for his word) is not to use negative \k timings, but to use the \kt tag. It explicitly sets the start time (in centiseconds) of the next syllable that starts.

Well, jfs -- silverfire actually is the one wanting to use negative timings (he's the one timing it); I wasn't told about the \kt tag. I know you can time negative timings in aegisub.... So, is it possible to time \kt in aegisub? Or it's a WIP?

[jfs]

Uh... I've never been able to use Aegisub's karaoke mode to produce negative timings and I'm pretty sure it's explicitly impossible. At least using mouse dragging on the audio view...
Also it'd be hard to represent in a sensible way in the audio view I think. Maybe as a kind of Gantt chart but... eww.

Aegisub doesn't support \kt at all, and (at least officially) doesn't support negative \k timings either.