Have you ever taken a photo and noticed little dusty pixels all over your photo? Or maybe you’ve seen inconsistent colours that you don’t remember seeing when you took the photo… It’s annoying, right? In photography, we call this “noise”, and in this blog post, I explain how a professional photographer would enhance the quality of a photo by both removing and preventing noise and achieving results like the photos below; first is of a DJ performing at a low-light indoor venue below and the second is of St Paul's Cathedral during a low-light evening.
How to prevent noise appearing in your photos? Use a flash photography to your advantage.
This photo was taken using flash photography and usually noise won't appear when a flash is used correctly. "How to use flash photography correctly" blog is live on my website - click here to learn more about flash photography. This required very little editing thanks to my flash. Please note all my photos are saved as RAW files. I explain the importance of this in one my blogs - click here to learn more.
Before:
You can see the noise and grain because I pushed the camera settings as much as I could. Ideally you don't want to go above 1,000 ISO on a DSLR. Newer models can tolerate much more e.g. twice that amount. Nowadays thanks to AI technology in our editing software(s), we can correct excess ISO in most cases. I suggest pushing your camera to its limits when shooting recreationally to test the threshold so that there aren't any surprises during a paid shoot.
After:
Where flash isn't possible, ambient light is needed and ISO. This photo was taken on an old DLSR body with a lens that had a minimum f4 aperture. This means that ISO had to be cranked up, the file was saved a RAW file and then we were able to proceed to de-noising and colour grading to achieve this result in Adobe Lightroom and Topaz.
How do you de-noise a photo? There is a de-noise process that I explain in this section because there is more than one type of noise.
But first, let’s understand how noise appears in our photos in the first place:
Noise in photography refers to random variations in brightness or colour that can appear in an image. It degrades the overall quality of our photos and it is caused by the limitations of the camera's sensor, especially in low-light conditions or when using high ISO settings.
There are two main types of noise:
1) Luminance Noise or Grain: this appears as random variations in brightness or intensity, similar to the grain seen in traditional film photography. It often becomes more noticeable in the darker areas of an image.
I also write an article about grain and how grain’s relevance is reemerging in unique ways in modern photography.
2) Chromatic Noise (Colour Noise): this appears as random colour variations, usually in the form of small specks or splotches of unnatural colours.
We de-noise using sophisticated softwares with algorithms that aim to detect and remove these unwanted artefacts whilst preserving image details and sharpness as much as possible. I personally recommend using Topaz Labs which is better than Adobe Lightroom for de-noising for various reasons. One of the key reasons is that once you buy the software you own it forever and they usually do very affordable packages for all of their applications including their sharpening one. The other reason is that they’ve been in the de-noising game for much longer and their software specialises in this. You’ll notice the difference if you like your photos super sharp and noise-free.
I’ve left a link which makes it easier for you to find: https://topazlabs.com/ref/2266/
Comment and let me know your thoughts at the bottom of this blog or send me an email.
Here's an overview of the typical process for how noise is reduced in a photo during editing:
Noise Detection: de-noising algorithms start by identifying the noisy areas in the image using various statistical methods, such as analysing local pixel variations or comparing the image with a noise profile model.
Noise Profiling: In some cases, the software will create a noise profile for the specific camera and ISO settings used to capture the image. This profile helps the software better understand the characteristics of noise produced by the camera and provides more accurate noise reduction.
Noise Reduction Algorithms: There are several noise reduction algorithms used by editing software. One common method is “De-noising Filters," such as the bilateral filter, median filter, or wavelet-based methods. These filters analyse the neighbourhood of each pixel and smooth out noise while trying to retain important image details.
Smart-Blending: Some advanced noise reduction algorithms use smart-blending techniques to selectively reduce noise in different areas of the image. For example, they may apply stronger noise reduction to darker regions where noise is more noticeable and use milder noise reduction in areas with more detail.
Trade-off between Noise and Detail: Noise reduction involves a trade-off between noise reduction and preserving image details. Aggressive noise reduction can lead to loss of fine details and a "plastic" or more artificial look like you see on some AI-generated images. Therefore, finding the right balance is essential, and some photo editing software allows users to adjust noise reduction strength manually.
Colour Channel Separation: When dealing with chromatic noise, algorithms may perform noise reduction separately on each colour channel (Red, Green, Blue) to maintain better colour accuracy while reducing noise.
Post-processing Refinements: After applying noise reduction, additional post-processing steps may be taken to enhance image quality further. These steps can include sharpening the image or applying local contrast adjustments to regain lost details.
It's important to note that while noise reduction can significantly improve image quality, it may not completely eliminate noise, especially in extreme cases with very high ISO settings or extremely low-light conditions. Photographers often use noise reduction judiciously, considering the artistic intent and the overall quality of the final image. Here is an example of a photo taken on my old DSLR without flash and the noise was too high to recover:
If you’re looking for a camera that can hold its own in low-light conditions and minimise the risk of photos being ruined due to excess noise, then the Sony alpha models offer a fantastic solution. I've owned and used Canon and Fuji cameras but my photography experience has never been smoother since I bought my Sony Alpha A7iv with a variety of prime and zoom lenses. Sometimes during street photography walks, paid shoots e.g. at events where flash isn’t appropriate, I find that the Sony really saves me from trouble. I can de-noise pretty much all of my photos when shooting in raw mode. I’m always careful to underexpose slightly so I use a lower ISO and then it makes it easier to retain detail in the final image. Here is a link so it’s easier to find: https://www.sony.co.uk/interchangeable-lens-cameras/products/ilce-7m4?locale=en_GB
Comments