website-featured-image

5 Clever Noise Reduction Techniques in Photoshop

Applying noise reduction to your images can be a real science. Any tool you use is faced with the almost impossible task of getting rid of the noise while retaining the details of your image.

In this article, I’ll feature 4 great noise reduction techniques, that come for free if you have Photoshop and one that requires a paid plugin.

Each one has its strengths and weaknesses, and there are a number of additional tricks that you can use to improve your noise reduction results.

So, let’s dive right in. We’ll roughly go from simple to sophisticated, and I’ll present five videos by other great photographers that explain each technique in detail.

Reduce Noise filter with a twist

In this video, Nathaniel Dodson shows you how to use Photoshop’s own Reduce Noise filter. He will also show you a trick that you can basically use with any noise reduction techniques as long as you apply it to it’s own image layer: He uses the Blend-If sliders in the Layer Styles dialog to quickly blend in the noise reduction only in the shadow and midtone areas, where most of the noise lives.

Noise reduction with Adobe Camera Raw

Adobe Camera Raw is Photoshop’s raw converter software. It provides the same raw development engine that you find inside Lightroom. This is simple and easy to use but provides a lot of power. However, you still need to understand the ins and outs of the tool to make the most of it.

In this video, Blake Rudis shows you exactly how to use the tool and what to watch out for when you do.

Noise reduction in Nik’s Define 2

In 2016, Google made the Nik plugin collection freely available. So, you can go and download the software for free. It installs as a plugin for Photoshop and Lightroom and can be used stand-alone. The software is a couple of years old, but it still works and it’s an excellent tool.

Define 2 is the noise reduction plugin in the collection. Unlike the previous two tools, Define 2 works with a noise sampling technology. This means that it looks at specific areas (that you can set manually) and tries to find the patterns specific to the noise. It then takes that knowledge and uses it to remove the noise from the entire image. The tool also comes with Nik’s control point technology that lets you increase or decrease the noise reduction effect in specific areas.

In the video below, Anthony Morganti explains in detail how to uses the software.

Noise Stacking

Noise stacking is a technique that requires you to take a series of photos of the same scene with the same exposure. You can then load all these images into a a single Photoshop file and blend them together by creating a smart object and setting its stacking mode appropriately.

By blending them in the right way, the noise patterns that are different in each exposure cancel out each other with fantastic results: Since there are multiple instances of each pixel, this is the only technique that can reduce the noise while retaining all the details in your photo. But, of courses, you need to know that you’re going to apply it while your shooting.

In this video, you see a demonstration of the technique.

Noise reduction with Topaz DeNoise

This last method requires the Topaz DeNoise plugin (not free). In my opinion this is the most capable noise reduction software out there. It gives you a lot of control over the noise reduction. But it is somewhat more involved than the other techniques.

In this video, I’ll show you how to really make use of all the potential of the software.

General tips for noise reduction

No matter which tools or techniques you’re using, there are some general things to keep in mind:

  • The best way to get rid of noise is, of course, to shoot at low ISO settings and expose properly in the first place. Do that whenever it’s possible and it’ll save you a lot of work. I know this is sounds trivial, but it’s true and important. So we got this one out of the way. 😉
  • Always reduce the noise on the original unedited photo (if possible). Great noise reduction software does not only look at your photos and smooths them out. It understand what kind of noise cameras typically produce, and it may even know the specific noise patterns of your camera. If you edit your image before you reduce noise, your going to alter the noise and make it harder for the software to do it’s magic. That’s why noise reduction is best applied to the original photo.
  • Two-pass noise reduction can improve your results. You should consider applying one round of moderate noise reduction at the very start of your workflow to give the subsequent editing steps a cleaner image to work with and to avoid that the noise is amplified in these steps. At the very end, before you do your output sharpening, you can apply another round to reduce the noise that creeped back in.
  • Don’t go too far. One thing that can be worse then having too much noise is the porcelain effect you get when you apply too much noise reduction. Be careful and try not to smooth out the noise areas completely, especially when your doing it at the start of your workflow.
  • Reduce noise selectively. Different areas and different tonal ranges will display different amounts of noise. In addition to that, having clean pixels can be more or less important depending on where they are in the image. So, applying your noise reduction selectively by using Blend-if (see above), luminosity masks or manually created layer masks is a good choice.

That’s it…

Do you have anything to add? How do you reduce noise in your images? I am curious to hear your opinions and questions. Just leave a comment below.

PS: Are you struggling to be creative?

Let me help you with the two video courses "Creative Photo Editing" and "The Inspiration Machine". Both are currently available with the Complete Photography Bundle until October 18, noon PDT with a 96% discount. Check out the other 40 products in the bundle before it's gone!

Leave a reply

1 Comment on "5 Clever Noise Reduction Techniques in Photoshop"


Guest
1 month 25 days ago

I think it was the worst mistake to give Nik Dfine the name it has. Everybody gets it wrong and uses Define.

Besides that thanks for this good article!