FREELessons: 10Length: 57 minutes

Next lesson playing in 5 seconds

  • Overview
  • Transcript

2.4 Noise Reduction and Pre-Sharpening

Pre-sharpening and noise reduction are the last steps before you convert and export your RAW image to the raster editor. It's best to be conservative with this step and use a light touch.

2.4 Noise Reduction and Pre-Sharpening

In this lesson we're going to explore noise reduction and pre-sharpening. First lets talk about the concept of pre-sharpening. Pre-sharpening is a common practice used to restore the crispness that was lost from the original scene back into your raw image. It's a way of neutralizing and correcting your photo as opposed to adding style to it. You want to be careful that you don't over-sharpen your image at this stage, especially because you will most likely be sharpening areas of your image later, when you get to the stylistic phase of it. Next, let's talk about noise. Image noise occurs when seemingly random pixels become visible in your image. It can give your photo a grainy feel, and often appears in areas of solid blocks of color. Image noise can have random flecks of color in it, which is also know as chroma noise. Or it can appear as small grains of pixels, which is known as luminance noise. When excessive and unintentional, image noise can be distracting and potentially detract from the content of your photos. There's several causes of photo noise including using ISO settings that are too high or shooting in low light. Sometimes noise is caused by using a slow shutter speed or even when your image sensor heats up. Also, you may run into noise if you're using an inexpensive camera. As far as trying to prevent image noise, try to shoot with the lowest possible ISO settings without compromising the quality of your image. Shooting in Raw helps as well, because it does not compress the image, which a JPEG does. And just a note, you don't want to necessarily remove all of the noise from an image. Just remove enough, so that it's to the point where it's not detracting from the content and aesthetic focus of the photo. If you were to remove all of the image noise, it can begin look unrealistic or surreal. In this base editing process that we're going through right now, this is not our intention. We just want to bring the photo to a neutral level. And so first, let's sharpen this image. And then, for noise removal, we're going to move over to our second image. And so, here within our first image, let's first make sure that we are at 100% or the one to one ratio. And let's focus on our model's eyes since that is a vital area. And I'm going to go ahead and turn the toggle before and after view off so that I can see more of her eyes. And then we're gonna move to the detail tab, which is the second icon in. In particular, we'll be exploring certain settings within the Sharpening, Edges, Impulse Noise Reduction, and Noise Reduction dropdowns. So first, let's take a look at the Sharpening dropdown. And make sure that the little power icon is on. And here you're actually given two methods to sharpen your image. First, you have the unsharp mask method, and then you have the RL deconvolution method. Now, within the unsharp mask options, you have your radius slider, which reflects the size of the sharpening halo and details that you're sharpening. And so, just to demonstrate, I'm going to drag the slider to the right, increasing the sharpness of our image. And you see it updating in this window. And beneath that, you have the Amount slider, which allows you to adjust the strength of the sharpening effect. So as we increase the amount, you see it increases the sharpening effect. Now we'll go ahead and set these both back to their default values and we'll come back and sharpen it ever so slightly in just a bit. Beneath the amount slider, you have the threshold slider, which allows you to confine your sharpening to a specific tonal range. However, the default threshold values given to you will generally help you avoid over-sharpening your image and will allow only the midtones of your image to be affected. So we'll leave it as is. Within this tool, the vertical axis corresponds to the opacity and the horizontal axis refers to the luminosity, meaning you can select the tone range that you want to sharpen. Beneath that, you have a sharpen only edges check box. It will seek to detect and target the edges of your image for sharpening as opposed to the more uniform areas of data. When you select it, you get a radius slider to adjust how much noise you want it to detect. And so, just to demonstrate that, here's our Radius slider, and beneath that, you have an Edge Tolerance slider, which determines how sensitive it is to picking up edges in your photo. And we'll go ahead and deselect that box. Beneath that, you have a Halo control box, which helps you avoid getting a halo around light objects in your image. The higher you set the amount slider, the more aggressively it works. And so here we have the amount slider. And we'll leave that box unchecked as well. Now the second method that you have for sharpening your image is the RLD convolution method. It's named this because of an algorithm created by Richardson and Lucy. It's the most effective with an image that's blurred due to excessive motion. First you have your radius slider which allows you to set the radius of the blur that you're trying to remove. Next you have your amount slider which controls the aggressiveness of the effect. Beneath that, you have your dampening slider which helps you to keep from sharpening the noise in the smooth areas of your image. And beneath that, you have the iteration slider, which allows you to control the number of iterations of the effect that you apply. For this particular image, we're going to use the unsharp mask method, and we're going to increase the radius slider ever so slightly. And if you need to, you can navigate around the image. To see the effect, you can use the Navigator panel. Or, you can click and drag within the center window using the Hand tool which is shortcut H on your keyboard. And we'll increase the value of the Radius slider to a value of about 0.76. And so we're just restoring the sharpness from the original scene that was lost when the photo was taken. Let's move down to the Edges dropdown menu for a moment. We'll go ahead and select it. And then we'll select the power button next to it, to make sure that it takes effect. The Edges dropdown menu sharpens your image using a different algorithm than the sharpening menu that we just came from. It focuses on sharpening just the edges of your image. For the purposes of this course, we will actually leave this dropdown alone. And so we'll go ahead and close it. And we'll also deselect the little power button next to it, and we'll take a look at the microcontrast dropdown next. These setting compliment the edges tool, and they focus on the textures within your image. After the microcontrast dropdown, you'll see that we begin to get into the noise reduction settings. Next, you have the impulse noise reduction dropdown. And so, as you can see, this image does not have much of an issue with excessive noise. And so we're going to actually navigate to our other image, and use that one. And so I'm going to display the film strip at the top by selecting the Show/Hide top panel. And then I'm going to double click on the nature photo, and that is displayed in the center panel. We will apply noise reduction to this image. Let's go ahead and zoom in 100% to the image. And let's navigate to an area that has detail and background behind it. This photo was shot specifically with a high ISO setting so that we could correct the image noise in RawTherapee. When making image noise adjustments, make sure that you are zoomed in at 100%. If you're at any other smaller percentage, you will not be able to see the changes that you make. And we'll go ahead and close that top panel for now, just so that we can focus on this nature image. And as you can see, this image has quite a bit of luminance noise. Let's go ahead and bring up the before and after comparison windows. And we'll have them both at 100%. Just so that we can see the effect of the changes that we make. And so here on the right, we have our current step, and on the left is the immediate previous step. And so now let's open the impulse noise reduction dropdown. This slider focuses on single pixel noise, also known as salt and pepper noise. And so make sure you have the power button enabled next to it. And then, dragging the slider to the right, you'll see that it removes some of the salt and pepper noise that we had in the image. And so this is the previous step, and this is where we're at right now. And for now, we'll leave it on a value of about 85. Now, let's move on to the noise reduction dropdown. And we'll make sure and select the enable button. And you can see that once we did that, it actually applied some noise reduction settings for you automatically, further removing some of the noise that we had in the image. The first slider that we have is the luminance slider, which allows you to decrease the luminance noise in the image. This will be the main correction that we make to the image. And so, lets go ahead and drag that slider to the right, and as you can see it is removing the luminance noise from the photo. As I mentioned, we don't want to necessarily remove all of the noise from our images because we want to keep it looking realistic. And so for now, we'll leave it at a value of about 32.92. Beneath the luminance slider, we have the luminance detail slider which allows you to adjust the amount of detail in your image after you've applied the luminance noise slider. And so just to demonstrate, as we move that to the right, it adds in a little more detail to the image. For now, we'll leave this one on the lowest setting of zero. Next, let's move on to the chrominance section. For you, this may be set to automatic global adjustment, which means that the software automatically analyzed our image and applied a global adjustment to it in order to remove the chrominance noise. Just to show you the manual setting, if you select the dropdown menu and then select manual, it will give you control of the chrominance sliders beneath it. And for demonstration purposes, we'll go ahead and drag the chrominance master slider to the right, just to demonstrate the effect that it has on the photo. This adjusts the noise reduction in regards to the photo's color channels. If you were to set the slider to ten, for example, the delta sliders below this would not change the image. We'll leave the chrominance master slider on it's default setting and then move down to the Chrominance- Red-Green slider. Now with this slider, you can decrease or enhance the effectiveness of the noise reduction in the red channel. And so for example, we'll take the slider and bring it all the way to the left. As we do so, you'll notice you're seeing more red pixels. And if we bring in all the way to the right, that is reduced. And we'll put that back on its default setting. We'll leave this slider on a value of about 50.8 in order to get rid of the red-green chrominance noise. Next we'll move on to the Chrominance- Blue-Yellow slider, where you can decrease or enhance the effectiveness of the noise reduction in the blue channel. And so bringing this to the left actually more blue chrominance noise apparent in our image, and then bringing it to the right reduces it. And so we'll leave this on a value of about 51.7. As you can see, you can get very detailed when correcting image noise within RawTherapee. The overarching point is to remove the noise that is distracting in your photo while still preserving a realistic appearance. Finally, I want to point out the gamma noise slider. This slider varies the strength of the noise reduction settings you apply across the range of tones. When you slide the slider to the left and reduce the gamma value, noise reduction will effect the darker tones in your image. And if you take the same slider and bring it to the right and increase that gamma value, brighter tones will be affected. We'll go ahead and leave this on its default setting. That brings us to the end of this lesson. In this lesson, we pre-sharpened our image and decreased the noise of our second image as well. And you can see what a difference the noise reduction tools make within RawTherapee. That also brings us to the end of this chapter. In this chapter, we applied basic corrective edits to our images. In our next chapter, we'll evaluate our images, apply some style to them and export them.

Back to the top