Easy-to-follow video tutorials help you learn software, creative, and business skills.Become a member
In this movie, I'll provide you with some insights for which Interpolation settings to apply when when downsampling an image inside Photoshop. So what I've done is I've created a diagram in which I went up to the Image menu and chose the Image Size command. And then, with the Resample checkbox turned on, I reduced the width value to 25%. And as long as I'm here, I want to provide you with little trick. Normally, notice this, if I switch from something like Percent to Pixels, then (LAUGH) I'm going to reduce the image quite a bit here.
Because I've changed its width to 25 pixels, but that also ends up changing the height. I'll go ahead and set that back to Percent, so we get the results I'm looking for. And if you want to change the unit associated with width or height independently, then you want to press the Shift key and click on this option. And then, I'll choose Pixels for height here, and we can see now that the height of the image is going to be 706 pixels. Anyway, the larger point is that I downsampled the image to 25%, and then I applied each one of these various interpolation setting. I got that already in a second file.
So go ahead and switch over to it here. And I'll zoom in as well, and I've got some Layer comps set up in advance so that I can just toggle through the Layers through the keyboard. This is the result of downsampling the image using nearest neighbor. So all that does is it just throws away pixels, and as a result, we get some very jagged transitions. Can't really think of any reasons why'd you want to use nearest neighbor for downsampling, but there it is and this here is Bilinear. So we end up getting the smoothest transitions between the various pixels.
This next one is Bicubic. And, let me go ahead and zoom in a little farther here, because the difference between these interpolations is a little bit subtle. So again, this was Bilinear. And now, this is Bicubic. We have a little more definition associated with the details this time around. If you feel like you are getting too much sharpness then you can back it off with Bicubic smoother, which actually does a good job of defeating noise inside of an image.
So if you have random variation between neighboring pixels, that are not really representative of the scene analogous to film grain, then you can help defeat those with that setting. And then if you want more sharpness then you go with Bicubic Sharper, and in that case, we saw pretty big increase in the sharpness of the detail there. And then finally, preserve details, which I have to admit, this is it. And if you didn't notice any difference there, it isn't because you're not looking hard enough, it's because there is no difference.
Pixel for pixel, this effect is absolutely identical. So, what's the upshot? Well, I'll go ahead and zoom out here, by pressing Ctrl+0, Cmd+0 on a Mac. Here's my advice, my rules for downsampling. For starters here, when in doubt, stick with Automatic, because that's going to go ahead and apply Bicubic Sharper. However, if you plan on later sharpening the image after you downsample, and I do this a lot. And I'll be devoting an entire chapter to sharpening by the way in the intermediate course.
At which point, we'll discuss these two commands right here Unsharp Mask and Smart Sharpen, both of which give you lots of control. And so, if you plan on going that route, then you don't want to sharpen on top of an already Sharp effect, because you'll be adding halos to halos. So instead, in that case, you want to use the standard Bicubic setting, which is called inside the Image Size dialog bo. It's called Bicubic, and in parentheses, smooth gradients. This is also a good idea if you think there's any chance that you're going to downsample an image more than once.
Next, if the image is noisy, as I say, random variations between neighboring pixels, then, go ahead and try Bicubic Smoother and see if that doesn't give you better results. And then finally, if the results end up looking jagged, I've had this happen on a few graphics that I've created for the web recently, where a very fragile detail ends up looking quite jagged, then you want to use Bilinear instead, because, it's just a straight averaging computation.
And it's going to give you the smoothest results possible, but of course, my biggest word of advice is that you never downsample and save over your original file. Always duplicate the image first, before you downsample so that the original file is safe. So there's my advice for downsampling, hopefully, it will help you achieve the best results possible when you're working inside your own images.
Get unlimited access to all courses for just $25/month.Become a member
117 Video lessons · 43083 Viewers
119 Video lessons · 54383 Viewers
65 Video lessons · 14610 Viewers
113 Video lessons · 82937 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.
Your file was successfully uploaded.