How to merge Japanese characters into Roman letters (or the reverse) -- looking for advice

[stevens70907226]

Hi,

I'm new to AE and am teaching myself with the After Effects Apprentice book and video series (on LinkedIn Learning). One thing I want to (eventually) do is merge Japanese hiragana/katakana characters into Roman letters and/or vice-versa. I don't want to reinvent the wheel, so I was wondering 

1) Does anyone know of any tutorials that can teach that kind of skill or have some suggestions for keywords so I can better find something on my own? 

2) Does anyone have any font suggestions as to what would be the best fonts to work with for this purpose? I'm guessing sans serif small caps, becasue they'd be a uniform height, but I'm not sure and am hoping someone here might have an opinion...

Thank you,

[Rick Gerard · Accepted solution]

Trying to morph one font into another isn't going to be easy if you stick to After Effects. You can play with displacement maps, liquify, mesh wrap, and other tools but the fonts can be very complex, especially if you are going from Japanese Characters to Roman (Latin) characters. Asian characters don't really have serifs so I would stick with Roman fonts that had a similar look.

I would open up Illustrator, add your Japanese characters, add your Roman Characters, convert both of these layers to outlines, then line up one shape with another and enable the blend mode to merge them. Play with the paths and settings until you get a few shapes merging. Start with the fonts, convert to outlines, select colors, set up blends, release to layers, save, import as a comp cropped layers, open the comp, set all layers to be one frame long with some room at the front of the comp, group layers by color, sequence layers and adjust timing. Here's a sequence of screenshots.

Add Text

Convert to outlines and color

Set Up Blend:

Align and release to layers:

Import as a comp:

Organize layers and set up for sequence:

Sequence layers:

End up with this:

Here's a quick tutorial that I did a long time ago. It shows the steps to go through. I've been using this workflow for almost 20 years. It's still the easiest way to morph shapes with predictable results. When you have the comp set up the way you want it, embed the comp in another comp (or Pre-compose everything) and apply time remapping to adjust the timing of the morph. I usually pick between 30 and 60 steps so that my default transition between shapes takes 1 to 2 seconds in a normal comp.

 

[stevens70907226]

Super complex, yet you distilled it into a video less than two minutes long. Thank you so much the advice and help. This is going to be my pet project for the rest of this month. 

[Mylenium]

Not really a good way to go abhout this, given how the languages and glyphs are constructed to begin with. Rick has explained his go-to technique already and I would just add that if you can, simply avoid such complex morphs. They simply tend to look extremely weird and it's usually much more efficient and more visually pleasing to have "invisible transitions" like dispersing letters into particles and reassembling them, using a blur whip/ cross fade while they move and similar.

Mylenium

[Rick Gerard · Accepted solution]

For this specific transition, I think I would probably do something a lot simpler. Even a simple cross-dissolve using a transparency fade and a blur or a spinning layer would probably look better. This is just a gaussian blur, opacity, and a short adjustment layer with Simple Choker to blend the two characters. I think I like it and it took about 3 minutes to do.

[stevens70907226]

That is EXACTLY what I had in mind when I posted this afternoon. My long-term goal is to be able to, for example, have とうふ morph into 'tofu' and then into a photo or illustration of actual tofu. 

You did it with the same technique you outlined above in three minutes! Wow. 

[Rick Gerard]

I copied your とうふ text from your post, added tofu to another text layer with the Hiragino Kaku Gothic Pro font I had on my system, created a simple opacity transition between the layers, created an adjustment layer with animated Directional Blur, Gaussian Blur, Simple Choker, and added and adjusted Bevil Alpha to get this:

If you convert the text layers to shape layers and animate the position of a couple of the extraneous parts of the Japanese characters you could eliminate the disappearing bits, but this looks pretty good to me.  Here's my project file. It was fun.