These improvements are available to everyone and are up to four times faster for Apple Silicon Mac users and 1.5 times faster on the best Windows architectures. The deep learning approach of DxO PhotoLab 5 and DeepPRIME, in particular, has been significantly optimized in terms of reactivity as well as processing and export times. With deep learning, DeepPRIME takes a holistic approach that combines the two steps into one. Traditionally, these two operations have been carried out separately, each introducing flaws that adversely affect the quality of the other. It drastically improves digital noise reduction while ensuring more efficient demosaicing. More on DeepPRIMEĪccording to DXO, DeepPRIME has been trained by deep learning methods using millions of images analyzed in DxO laboratories over the past 20 years, DxO DeepPRIME technology uses artificial intelligence to develop raw files. Your raw file stays raw and you can do more editing in Lightroom, Photoshop, or your editor of choice. It's the first thing I do with a raw file, and you get world-class optical correction, noise reduction, and sharpening. I've talked about it in my review of DXO PureRAW, offered as a standalone application, but now, that technology is incorporated into PhotoLab 5.
This new version also introduces a new DeepPRIME denoising tool that is up to four times faster than previous iterations as well as support for Fujifilm X-Trans sensors (Beta).
We've got a new and improved version of DxO PhotoLab available today, and version 5 will be of interest to advanced photographers and serious amateurs.