There are many situations where high ISO settings or high dynamic range situations introduce noise into your images. Applying noise reduction can reduce detail and sharpness and introduce artifacts. Fortunately, most photo-processing software provides tools to reduce noise levels, but they all behave differently. Some behave better for landscapes, while others work better for wildlife images. The following individuals have run a few experiments with some of the leading software to term how to choose from these options depending on the type of photography you do.
Matt Kloskowski
Using wildlife photographs, Matt compares Topaz, Lightroom’s older noise reduction process and the new Lightroom AI process. He finds that the new Lightroom AI program and Topaz are comparable and involve the same amount of time. However, Topaz does combine sharpening and upsizing into its process, which he suggests would be better to use if you wish to do all three. It should be noted that in these cases, DNGs are used to move the images back into Lightroom. He does suggest that the sharpening in Lightroom, compared to other plugins, is lacklustre. He also suggests that ON1 also produces similar results.
Nick Page
Nick makes a comparison between Lightroom, DXO Pure and Topaz. His conclusion is with Lightroom and DXO because the way they work needs to be done at the beginning of the processing workflow. He feels this is better done at the end of the process, which he feels Topax allows. He also finds the DXO Pure sharpening process lacks fine adjustments and, as a result, is so aggressive that it removes too much detail for his liking. But he does say Lightroom’s new AI noise reduction is excellent.
Alex Armitage
Alex compares the noise reduction in Lightroom AI noise reduction, Topaz, DXO and ON1. In this case, he uses a night shot with stars in the image. This leads him to reject Topaz based on the number of removed stars. In ON1, he sees a lot of fractals or artifacts in the dark areas, unlike what he discovers in Lightroom, in addition, he sees an unpleasant colour difference in ON1. In the case of DXO pure RAW, he got his best results. However, when he tried to use the new XD mode, he did find odd lines appearing like spider webs connecting some stars. He found using the deep prime mode and not the XD produced a better result. Finally, after looking at the results from Lightroom, he concludes the best results come from Lightroom’s AI noise reduction feature.
Noise Reduction Methods at 12800 ISO
In this example, I take an extremely noisy photograph, apply the standard noise reduction methods and compare them to Lightroom’s AI noise reduction tool.
