Converting Images to Vectors

I’ve long been a beginner-intermediate user of Photoshop, but I have no experience with Illustrator. If I’m not mistaken, Illustrator is the program people use to create vector images (are there others?).

Anyway, here’s my question…

Suppose I find a tiny image of a photo of a person. Let’s say it measures just 200 pixels wide. If I enlarge this image to 1,000 pixels, it would obviously look like crap.

However, is it possible to convert tiny images to vector formats, then enlarge them, creating images that look OK?

If so, then I assume the next step is to acquire Illustrator. I’m assuming I would simply open the image in Illustrator, then follow some command to convert it to a vector image.

Thanks for any tips.

Illustrator’s “live trace” function will make any photo into a really bad paint-by-number-look piece of crap. The smaller the photo, the more crap-like it will look. If you enlarge a paint-by-number look, you get a BIGGER paint-by-number look.
From Adobe’s tutorial:

If that’s the style you’re going for, start with a really large image and play with the number of colors and snap settings (but usually the small photo is the problem that prompts this question.)

Kicking myself for not stopping to take a photo of a realtor billboard I saw some time back. Realtor obviously did not supply designer with a proper billboard-sized photo so designer decided to use live trace. Yeah, unrecognizable at billboard size. And gone the next day. Bet designer was out about $5000 for that mistake. Billboard prints ain’t cheap.

1 Like

I’d imagine you need to have at least a working understanding of vector and raster. Not knowing where you want to go will never get you there.

You can scale vector graphics up or down (as long as it is still in vector format) and convert it to a raster image but you can’t do it the other way around. It’s just like Eriskay mentioned, you’ll need a good understanding of both of the types. There isn’t a simple way to scale up normal raster images. You’d have to trace it like Printdriver explained or do it by hand.

The best way to always go about is to always use a very high res image, preferably much bigger than you will probably need. If someone doesn’t own that image, there’s probably a good reason for that too. Either they aren’t allowed to use the image for free or they don’t know what they are doing. (fe. taking bad quality photos with their phone and thinking they can be used on a billboard is unrealistic) Raster images can be scaled up a little bit, but not too much because it’ll be pixelated and lose quality.

This page probably explains it better than I do: https://modassicmarketing.com/understanding-image-file-types

As already explained, no, that’s not doable unless you’re working on a project that will benefit from that paint by numbers look that @PrintDriver described. Vector applications, like Illustrator, are great for creating various kinds of graphics composed of shapes, lines and fills, but they’re almost never useful for manipulating continuous-tone photos.

With some notable exceptions, like raw images, there is no additional image information hidden in a photograph beyond what you see. There are algorithms that can infer how certain features in an image, like an edge, should look, then sharpen or exaggerate those features. But as a general rule, it’s just not possible to enlarge an image and have additional detail added to the image that wasn’t in the photo to begin with — this would require magic.

If this were possible, there would be no need for telescopes or microscopes. All one would need to do is take a photo of, say, the moon with a regular camera, then enlarge that photo using special software until it showed the grains of sand in the craters there. Obviously, this is impossible since the camera never captured that information to begin with. The same is true for much less extreme examples. It’s just not possible to get a recognizable image of Fido’s face from a blurry image that doesn’t contain that information. The best that can be expected with enlarging that photo is a larger version of that same blurry image.

2 Likes

It’s all pretty much been said.
All I would add is; it’s great fun!
turning a sketch into vector then playing with it can be quite fantastic; the potential is impresive. Also, once you pass the initial learning curve, it’s fairly intuitive.
Playing with images, and colors through the scope of vector can indeed be a rich experience. Althout sometimes demanding on the computing capacity for large images of multiple colors… extra RAM makes all the difference!

©2019 Graphic Design Forum | Contact | Legal | Twitter | Facebook