Got Landscape Photography Noise?
Digital Camera noise is one of those annoying, irritating, pain in the butt technical issues that can kill a great image. No doubt, top line full frame DSLR cameras are much better today than just three years ago, but noise is still a picture killer unless the photographer consciously works to minimize it impacts while shooting and during post-production sessions.
Here is the issue: If you are a Landscape photographer, you choose light for drama, emotional range, design and other narrative effects, but you do not control light. It is what it is, and you have to decide when the light is right for your purposes. You will often shoot in low light, at night, or intentionally underexpose an image for a creative purpose. We often use neutral density filters to lengthen exposure times such as when capturing waterfalls or working with fast moving clouds. Then there are those times when we wish to push the shadows or overall exposure in post-processing to enhance details in the darker areas while preserving highlight details. These creative processes can invite noise into our images.
Noise appears with varying degrees of intensity as an uneven grainy look, similar to grain in the days of film. However, grain was often a creative construct that could deliver an attractive tactile quality. In fact, film grain can be so attractive that photographers pay big bucks for plugins that create grain to mimic film grain. However, digital noise often creates unattractive milky, muddy, soft images, sometimes with weird artifacts and there are no computer-generated plug-ins to create digital noise— digital noise is just ugly.
The Nature of Noise
Men and women of a certain age may remember ‘snow’ on their analog television sets when the broadcast was a bit too far away, and the signal was weak. This snow was image noise and occurred when the background signal, “the snow,” was stronger than the broadcast image. As TV circuitry and antennas improved, weaker signals became much clearer and could compete with nearer signals. The same thing occurs with radio broadcast static and these relationships of snow and static to a signal is the Signal to Noise ratio.
In cameras, the Signal to Noise ratio increases with the sensitivity setting (ISO), length of the exposure, ambient temperature, sensor size, sensor quality and will vary depending upon camera models. You may not notice noise on the camera viewing screen, but when you increase the image to 100% (often much less) on your computer screen you may see noise, and if you print at large scale you will find noise can undoubtedly make a mess out of an otherwise excellent image.
Digital camera noise develops from one or two primary sources
- The camera sensor size and structure
- Sensitivity of the camera sensor, (sometimes both)
- Camera Circuitry (shielding, component quality, )
Secondary causes of camera noise are:
- Heat (usually associated with long exposures) but ambient temperature can also be a contributor.
- Individual image type. With digital cameras, darker regions contain more noise than the brighter areas. Brighter regions have a stronger signal due to more light, resulting in a higher overall Signal to Noise Ratio. This means that underexposed images will have more noise, even if you brighten them with post-production tools.
Sensor Design and Noise
All digital cameras use a sensor array with millions of tiny pixels to produce the final image. Each time you press your camera’s shutter button the exposure begins and each of these pixels has a “photosite” which is uncovered to collect and store photons in a cavity. The number of photosites is equal to the number of megapixels. Therefore, a 20-megapixel camera has 20 million photosites on the camera sensor.
But the number of pixels is not as critical as pixel size. Generally, the smaller the photosite, the noisier the image. So, noise (and overall image quality) is not about the number of pixels on the sensor but the size of the sensor.
For example, modern phone cameras may have 12 megapixels, and yet the same full frame 12-megapixel DSLR camera will yield a far superior image if we enlarge that image because DSLR’s have much bigger sensors, and thus they have much larger photosites than a 12-megapixel phone camera. Although phone camera circuitry and sensors have improved significantly in recent years, they still generate a great deal of noise even when shooting in bright sun. You just don’t notice it because you are looking at the camera viewer and not at an enlarged image. Consequently, smaller sensors with a large number of pixels generally produce noisier images.
Sensitivity Matters (ISO)
ISO is a standard measuring system, which describes a camera’s sensitivity to light with settings listed as factors of 2, (or 2 and fraction in some cases) such as ISO 50, ISO 100 and ISO 200, where higher numbers represent greater sensitivity. The ratio of two ISO numbers represents their relative sensitivity, meaning a photo at ISO 200 will take half as long to reach the same level of exposure as one made at ISO 100 (all other factors being equal). Today’s top line digital cameras have ISO ranges up to 25,000 plus, but for the landscape photographer, most of our work occurs at 100 to 6400.
As you increase the ISO settings in your digital camera, you increase the sensitivity of the sensor through amplification of the image signal resulting in progressively more noise at higher ISO speeds. This means that when you apply a high ISO setting to a small sensor and amplify the sensitivity you will not only see a lot of noise (luminance noise) but also bad colors (chroma noise) and all kinds of garbage artifacts.
Long Exposures mean that the shutter is open longer and thus amplification increases, which means more noise. How much and what kind of noise depends upon the quality of the camera sensor and some ambient factors.
The tradeoff is to increase ISO and decrease exposure or adjust the aperture to shorten the exposure.
How to Control and Reduce Camera Noise
Now that we know where and how noise occurs in our images let’s see how we can minimize it. We begin with the source, our camera.
Use Lower ISO Settings (when possible)
This is the easiest and most effective thing you can do to eliminate noise. Newer DSLR cameras (and phone cameras too) have an excellent signal to noise ratios at higher ISO numbers. I often work with a Canon 5D Mark III, and I shoot at ISO 100 to 6400. At 6400, image noise is very quiet in bright areas but does become visible in shadows or with very long exposures. Nonetheless, in most cases, I can quickly correct it in post.
To avoid a high ISO setting, you must make some choices. Open your aperture to its widest setting that will allow you to create the image you want (assuming you have the desired depth of field at that aperture) and select a low ISO. Alternatively, use a tripod and longer exposure, or increase your ISO if none of the other choices fulfills your needs. It is a good idea to experiment with your camera’s ISO and shoot images at various high ISO settings to determine your upper level of noise generation and acceptance.
Shoot in RAW format
RAW images are, well raw, and chock full of great image data, whereas a JPEG image, is the result of pixel compression, and in-camera process. High ISO setting with JPEGS often creates some nasty, noisy artifacts. Moreover, post-production tools remove noise much more efficiently from RAW images.
Today’s high-end full frame (big sensor) cameras are much more efficient at managing the highlight shadow ratios (known as dynamic range), and you can make excellent exposures without compensating for highlights in most cases. In fact, you can now overexpose slightly so that the shadows are a little brighter than normal and then you can pull the highlights down in Lightroom. This is a very effective and part of our Fool Proof Develop Module Workflow.
However, landscape photographers often experience very broad dynamic ranges in the field wherein the proper exposure for highlights and shadows can be many f-stops apart and beyond the camera’s dynamic range. Thus, photographers adjust exposure for the highlights and try to avoid clipping those precious white details, which often means we lose detail in the shadows, or clip the blacks. To compensate we attempt to boost the darks and shadows in post-production and this usually works well but at high ISO settings post-production tools often introduce noise into those shadows. For landscape photographers, it is far better to take two exposures and blend them together in post (not covered in this blog).
Long exposures can be cool or funky
Shoot a long exposure of a waterfall, ocean waves crashing on a beach, or clouds scudding across a sunrise sky and you will no doubt introduce some amazing drama into your image. But long exposures can cause your sensor to heat up and render wrong, weird colors and faulty exposures. As an example, I’ve done some long blue hour exposures and captured some saturated blue colors that I did not think existed— and they didn’t.
This is where you need to understand the limits of your camera and previsualize your compensation workflow. In case of the strange blue colors, I simply made a hue and saturation adjustment in Lightroom, and the real-world sky returned.
Use of In-Camera Noise Reduction
In most DSLR cameras, you may have an option to turn on High ISO Noise Reduction (HNR) or Long Exposure Noise Reduction. In general, this tool helps reduce noise at high ISO settings, long exposures, or both with circuitry that looks for any incorrect pixels. This is an in-camera post-processing step and often takes a long time to complete. For example, if you do a 20-second exposure of a waterfall the HNR process will take 20 seconds to complete. If you are doing 3 minute captures this time issue can be impractical. I have not found that the trade-off of time versus image improvement yield to be very positive. You may find this works for you and your camera and I recommend that you test the results and decide if this feature fits your shooting workflow.
Post-Production Noise Reduction
Lower noise reduction processing at high ISO is a ying and yang process that balances the loss of detail in high ISO settings. Major camera manufacturers have noise reduction algorithms in their sensor circuitry that create a low noise appearance by smearing and softening details along pixel edges. It is a tradeoff between sharpness on the one hand and noise on the other. Noise reduction is a complicated issue because noise reduction can be applied in both Chroma and Luminance dimensions collectively and individually and some of the better programs can attack noise by frequency which allows the software to target only areas where noise is an issue.
The Lightroom Noise Reduction tools are excellent, and you will find them in the Details panel (also available in Camera Raw). The panel has several adjustment sliders, and each has a distinct and useful purpose. As you adjust each slider, you are making trade-offs between Noise Reduction and preserving details or sharpness.
This is a visual process, and you need to zoom into 100% of the image to see how your adjustments are affecting the noise/detail ratio.
The noise reduction sliders in Photoshop Camera Raw, are identical in Lightroom
The Luminance Slider
Start with this slider which reduces luminance noise which occurs from over or underexposed pixels. As you move the slider to the right, you reduce noise but increase softness.
The Luminance Detail Slider
Use this slider next, which controls the luminance noise threshold. As you move to the right and achieve higher values, you will recapture or preserve more detail but can also produce noisier results. Lower values produce cleaner results but also remove some detail.
The Luminance Contrast Slider
This tool can be useful for noisy photos, but gentle; careful moves are needed because higher values preserve contrast but can also produce noisy blotches or mottling. Lower values produce smoother results but can also have less contrast.
Reduces color noise. This is often noticed in the underexposed shadow areas of an image. Color noise occurs more frequently in JPEG images than RAW.
Controls the color noise threshold. Higher values protect thin, detailed color edges but can result in color speckling. Lower values remove color speckles but can result in color bleeding.
This slider controls the smoothness of the colors in the image. This is useful if you still have some unusual color noise in your image after you have made all your adjustments above. Use this to finish off the noise reduction workflow.
This is a tedious process with no standard setting for all images. Proceed slowly, and zoom in and out so that you can see the adjustment results after each adjustment. I tend to start with higher Luminance, and Color setting then retreats to the point of sanity. Once I am happy with these adjustments, I‘ll move to the next, starting with a higher setting then retreating until I am happy with the results. With patience and a bit of practice, you can eliminate noise or reduce its effects to an acceptable level.
Noise is inherent in all digital cameras, and some noise is necessary to eliminate the plastic look that occurs in digital images, but excessive noise ruins good work and should be dealt with while shooting or planned for in the post-production workflow. Today’s newer DSLR cameras manage noise quite well even with a high number of photosites or pixels. None the less it can occur when you are least expecting it and knowing how your camera performs at various ISO setting and how your camera processes noise is critical to reducing noise to an acceptable level.
Creative landscape photography requires essential detail and a wide dynamic range to present effective narration and navigation. Thus, take noise seriously, and you will be on your way to creating flawless work.