So, quick question, how exactly do you decode glitches? Everyone seems to use a similar method but nobody mentions how. I spent about 20 minutes trying to figure it out with my limited (read: nonexistent) knowledge of photoshop last night before falling asleep at my laptop.
I included most of the shorthand details and caveats in the second image here, but to stepways it out:
- Download/save the glitched-text gif.
- Find a way (likely outside Photoshop) to explode the gif into its component frames. (For me on a Mac, I have to load the gif in Preview and copy each frame, pasting each into the same new Photoshop document as a new layer each.)
- Put the frames in Photoshop as separate layers, and toggle visibility until you get the cleanest, most aligned one. (By ‘aligned’, I mean some of the frames make parts of the text uneven; you want everything in straight, even lines for typing over.)
- Create a new text layer on top of it all using 14-point Courier New, Bold, with anti-aliasing set to “None". Use a color substantially different from the text colors in the log. (I’ve been using a Jade Harleyish lime green.) Make sure Smart Quotes are turned off in Photoshop’s preferences/options.
- On the layer’s blending options (of the text), switch from “Normal" to “Linear Burn". Type some of the first line of the log and move the text until it aligns with the source image.
- Keep typing along with the source image, thus decoding it! If any pixels show up in your pure chosen text color (lime green in my case), it means you typed a wrong letter and need to revise it. If there’s enough glitching in an area, you may get a number of potential matches letter by letter, but they’ll usually only add up to a whole word in a couple ways, narrowing it down to a single option that actually makes sense in context.
- Not all the lines are spaced the same amount apart! As you move from line to line, you may need to highlight a misaligned one and adjust the “leaning" a pixel or two up or down, nudging it until it fits.
- When you have the full log, just duplicate the text layer and shove a filled-in layer the same color as the background of the pesterlog beneath it. Set the blending of this duplicate layer back to “Normal", and set the colors throughout it to those of the original pesterlog to get a clean, matched version of all the text. :D (You might want to set the new background to invisible and the text layer’s opacity to 50% while you’re coloring, just to make sure while you’re working.)
That’s about it! Once you’re used to it, it’s simple and fast; I’m not at all surprised that more people are doing it.
If you’re still confused, try opening the original PSD file I linked in the first decoding attempt and use that as a guide?
(NOTE: Here’s my Glitch Deobfuscation tag for all my previous/future decoding attempts; you can find this post under there in the future, too.)