Easy-to-follow video tutorials help you learn software, creative, and business skills.Become a member
I've got a demo set up here that you can duplicate if you want to follow along. I'm going to use a horizontal gradient with the ZBlur node to increase the blur from black to white, and apply it to this checkerboard. So first of all, we have our CheckerBoard node. Next is a Grade node that I just use to increase the contrast to make the blur easier to see. Following that is a Ramp node, and what I did here was I set it up so we'll put a ramp in the alpha channel that goes from 0 black over here to 1.0 over there.
It also have has a broad strip of 0, so we'll have a nice section of unblurred image, and you just gradually roll into our more blur as we go to the right. If you need to see what that setup looks like, I put the output in the alpha channel and .0 is on the left and .1 is on the right, just a little bit short on the right, so I would have a narrow strip of exactly 1.0. Okay, we'll close this, switch the viewer back to RGB, and let's go get the ZBlur node. So from the Filter tab, ZBlur, and that hooks right in under the Ramp node.
Now immediately we get a blur. The reason is the ZBlur node is looking at the depth.Z channel, which happens to be black or 0 and then combining that with a focus plane and the depth choices, it's giving a bit of a blur. So first we'll do is we'll set the channel to RGB--always want to get into the habit of setting those channels. Now my "depth data" is in the alpha channel; it is not in the depth.Z channel, so we're going to direct the ZBlur node to look into the rgba.alpha channel for its business.
Now we can set it up any way we want, but I wanted to have the black on the left with no blur and the white on the right with maximum blur. So to set that up, the math function that we want is far= 1, so the right edge where the alpha channel is 1 is going to be the farthest distance from the camera and get the most depth-of-field blur. We set the focus plane to 0, meaning the 0 black over here will be in sharp focus. We'll set the depth-of-field to 0, meaning it's going to immediately start blurring the images as soon as the code values get above 0. And we'll set the Size for 10 to give it a nice blur.
Okay, so right off the bat, you can see it's sharp on the left and blurry on the right. That is to say, where the gradient is black, it's sharp. Where the gradient is white, it's blurred. Okay, let's push in and take a look at what we've got. I'll set a zoom factor of 1 in the viewer. We're at the left edge and here the alpha channel is 0 black. You can see the code values down here. Back to RGB.
So we've got no blur here and as we move to the right, you can see the blur starting to pick up. It's not just a mix back of some blur radius. It is in fact increasing the blur diameter as we move to the right. In fact, you can see the yellow line getting thicker and thicker as we move to the right. Here we are. To increase the blur more, we could increase the size, but don't forget to increase the maximum along with it, or you won't see any change at all.
I'll re-home the viewer. As you can see, the trick to using the ZBlur node as an image-based blur is to load it's Z channel with a one-channel blur mask image. This blur mask can be created any way you want, through a Paint node, a Luminance key, or any other source, and it could be put into any available channels, since you can set the ZBlur node to look at any channel for the Z information.
Get unlimited access to all courses for just $25/month.Become a member
100 Video lessons · 8549 Viewers
107 Video lessons · 37412 Viewers
94 Video lessons · 26739 Viewers
147 Video lessons · 6182 Viewers
Access exercise files from a button right under the course name.
Search within course videos and transcripts, and jump right to the results.
Remove icons showing you already watched videos if you want to start over.
Make the video wide, narrow, full-screen, or pop the player out of the page into its own window.
Click on text in the transcript to jump to that spot in the video. As the video plays, the relevant spot in the transcript will be highlighted.