HOME | DD

Description
body div#devskin0 hr { }

A caption is two parts put together: text and an image. The primary goal of a caption is to have these two parts complement each other.
So first things first: not every story needs to be made into a caption! I know, it sounds crazy to be telling people to not do the thing this guide is about, but hear me out.

Here are the two big indicators that a caption isn’t going to work:

    1.       The story is too long. It varies but I’ve found that once you get to around 3,000 words of text it’s very hard to format all of that into a caption and keep it legible.

    2.       The picture is too small or compressed. While tools like waifu2x exist to upscale images with amazing results, sometimes it just isn’t meant to be.

As an aside, the second is easier to deal with than the first. Tools like saucenao , iqdb , TinEye , and Google reverse search can help you find a better quality version of an image if you think yours is too compressed or small (by the way, if anyone asks me for an image source, I’ll direct them to these websites and tell them to use a cropped image I used instead of telling them where it came from).

My suggestion as for what to do when a story isn’t going to work as a caption is to upload the image into sta.sh (DO NOT link on another website), post the text in DevaintArt’s StashWriter , and insert the media you uploaded into the story. It’s quick and easy and also has the benefit of giving the reader raw text and the full image. For me, this actually what I usually do as my stories tend to get quite lengthy.

If you decide that you want to make a caption, never forget that a caption is worthless unless it is legible. Here are some quick ways to improve legibility:

    ·         Keep the background for the text plain.

    ·         Have high contrast between the text and the background. This applies to colored text as well.

    ·         Outline the text (white text with a black outline is legible on any background).

    ·         Keep the total caption at a 16:9 aspect ratio. This is what modern screens are at and this is the aspect ratio your eyes see at.

So without further delay, let’s make an example caption.

I have this image that I want to use in a caption but I have a problem: I got it from Twitter, which often compresses images uploaded to it. I want the highest-quality version of this image that exists online, so I’ll upload it to saucenao, which will tell me the source.

The pixiv source has been deleted, but the raw image exists on booru imageboards.

I snatch it (making sure to not download the SAMPLE size image that is scaled to fit your screen).

From there, I open the image in my image editor of choice. You can use pretty much whatever besides MS Paint (which is AWFUL), but for this example I’ll be using Adobe Photoshop CS6 (which is easy to pirate and sometimes free for college students). For those who want a free option that’s almost as good, GIMP is a great alternative.

Next, I check out the image resolution. It’s 922x1149 pixels, which is perfect for what I need to do. I generally make my captions no larger than 1920x1080 pixels, but sometimes if an image is too small I’ll make my overall caption smaller. NEVER EVER UPSCALE IMAGES OUTSIDE OF waifu2x!

Downscaling is fine though. Keeping my original aspect ratio, I resize the image to 867x1080. This part isn’t necessary, but I have a personal aversion to bloated image sizes.

Now, I do two things: I make the overall canvas 16:9 aspect ratio, and I make a new empty layer that I put behind the image. If you don’t know, layers make your life in image editing a million times easier.

From here, I have two options: I can either make the background plain white; or I can do a little image editing and do something fancy. Plain white is boring as all hell but it’s highly legible. I’m going to do the fancy option, but use your own personal best judgement.

Using the selection tools in the image program, I carefully cut around the image then delete the cut portions.

Next I copy the lovely gradient already in the image and apply that to the background behind the image with my paint bucket tool.

Huh. That doesn’t quite line up.

I decide to clean up the image and then redo the gradient (while also removing the Japanese text and paw print). It's close to what I want but not quite there.

I add a white outline and outer glow to the image.
Much better! Now, for the text. I can either wrap the text to the image or just do a linear edge to it.

As a mockup, here’s a 2,000 word Lorem Ipsum with the linear edge.

And here’s the wrapped edge (I put the image on the right because the left paragraph readings).

I like the second better, but first, let me show you how I made it.

I make a hard selection, then using the straight lasso tool I refine the selection for use as a text box and then contract it.

Using the wand tool, I save as a work path then I use the text tool and click on the new path I just created. Voila! I just made a wrapped text box.

Now comes time to actually make the text legible. I generally use Segoe UI Semibold for my captions because it’s a clean font that’s easy to read at many sizes, but you’re free to use whatever. Just make sure it’s legible and not too gaudy.

We’re almost there! I add a small and nearly transparent white outer glow to the black text to improve legibility, and then I call it done.

Now I save it as a raw .psd in case I want to edit it later. I also export it as a .png (use .png for flat images with sharp lines and .jpeg for complex images to keep size down).

And here’s the completed image!

And here's the .psd for you to play around with: 000 Start PSD

I hope this helps all aspiring caption makers out there. Let me know if it helped you and what I can do to add to this. If you have any questions, feel free to ask.