Explore related tutorials & articles
19
Replies
I'll start right off by saying that I pretty much never ever use the pencil tool. A fully hard brush a few pixels wide, makes much nicer line work. As to whether you can save you existing lines, that depends if they are on a separate layer.
This is a pencil line zoomed to 200% so we can see the pixelation.
Gaussian Blur (value 2.0)
And Levels to harden up the line again.
It's common practice for line work to be roughed out, often with pencil and paper. Then scanned into Photoshop to Ink the line work. If you do the whole process in Photoshop then you need to use multiple layers.
If you are using Windows and have a tablet, or are considering getting one, then there's a wonderful plugin called :Lazy Nezumi Pro that turns you into an amazing artist. This video also demonstrates inking with LNP.
I'm pleased you are using Photoshop for drawing and illustration. That is so much more enjoyable than tarting up a few photographs. If you like books, get hold of Steve Caplin's 100% Photoshop. You'll be amazed at what you can create entirely from scratch just using Photoshop.
Semaphoric wrote
I use this method all the time. I sure wish there were a way to keep sharp corners.
This is something I kind of stumbled upon a few years ago, but I think I shared it here.
OK, the same layer copied a couple of times so we can try different methods. I have saved the selection so I can repeat it.
Gaussian Blur value 10 (the black background is 600 pixels square)
Lens blur with the same selection in place
And after OKing it
It works on things like skies where you want to blur a banded gradient without making the edges go thin. In fact it blurs the inside of a selection, but not its edges.
I don't know if that covers what you are talking about.
To fix the white object jaggies, if they are on their own layer, just use the steps I described back the thread. If they are not on their own layer, you'll need to select them like this
Gaussian Blur (value 2)
If the white and blue is on the same layer, you'll need to be very careful when using levels.
Or you will see an obvious change in the blue like this
The white on black is much easier, because black is black.
This section is the most tricky because the whites extend to the edges, so you will need to make an accurate selection. That means when you apply the blur, thew edges of the blue rectangle will go soft. They will come back partially after using Levels, but you would need to carefully paint over the edges to get fully sharp edges. Much easier to do it again on two layers.
christopherc92001052 wrote
I'm having some trouble hardening up the lines with the level adjustments though. They don't seem to be doing much besides changing the color. Could I be doing something incorrectly?
If the white and backgrounds are on different layers, then apply the levels adjustment directly to the white layers. If you use an adjustment layer, then you need to clip it to the white layer by clicking on the icon indicated below
Below we have a white circle on its own layer, above another layer filled with blue. The white circle is heavily pixelated — I achieved that by using Free Transform to up-res the circle with Interpolation set to Nearest Neighbour
With such heavy pixelation we need a high value blue to average it out. In fact we would not use this methos in this case, but we are exaggerating for the sake of demonstration.
Now we come to the step that you say is causing you problems. We have two options. We can apply Levels directly to the layer, or we can add a Levels adjustment layer, but we would need to clip it to the white layer as shown in the last screen shot in post #10. Whichever method we use, the properties for the levels adjustment will be the same.
Lets ignore the blue layer, and think about the blurred white layer. I'm going to make that blue layer black, because it shows the soft edge of the white layer better. Look at the histogram in the Levels Properties panel. The left side shows full black, and the right fulll white. The height of the histogram indicates the percentage of image pixels that fall at a particular tonal range. Right in the middle, for instance, you would have 50% grey, and we can see a shallow floor of pixels that are neither black nor white right across the histogram. We don't want that.
We need to have every pixel either full black or full white, with a tiny amount slightly off white and off black to provide what is antialias, which means blur, but a controlled amount of blur to fix the pixelation.
If we move the sliders all the way to the left, we make all of the off-white pixels full white, so the circle is bigger.
And conversely, moving the sliders all the way to the left makes all the blurred pixels black so the circle is smaller.
So we bring the sliders almost to the centre to keep it the same size. Note that we don't make the sliders hard up against each other or you'd still have pixelation. Also not the nasty indentations, which is why we'd never use this method on so large an object with so pronounced a blur.
Is that covering it?
What operating system are you using?
If Windows 8.1 or 10, then you have the Snipping Tool as standard
But I think most of us here use GreenShot
Greenshot - Screen Recorder download | SourceForge.net
It's an absolute godsend for anyone using forums or Reddit type sites. You can freeze the screen by hitting the Print Scrn key, and either copy to the clipboard, or send to the editor which allows you to annotate areas on the grab.
Adding to this forum is as simple as using Ctrl v (Cmd v) to paste it directly. The only times we use the forum widget is for animated giffs or videos.
OK I can see what the problem is now, and I am afraid I missed a step. Because you have the levels layer clipped to the white details layer, you are only working on white pixels regardless of how bright they are. You need to select the blue layer, and white details layer, and right click and chose Make a Smart Object Then clip the levels layer to that, and it will work OK.
AdChoices