How does Neat Image work?
Neat Image uses sophisticated math to tell noise from image details and to selectively reduce only noise while preserving the actual details. Its reduction algorithms are developed specifically for digital photography applications by a highly-qualified professional image processing research group. These algorithms surpass the quality of all classic noise reduction methods and even that of the wavelet-based methods. Although the wavelet-based methods were developed only 15–20 years ago and are still considered relatively modern, Neat Image uses an even newer and more efficient approach to noise reduction. This approach enables drawing a more clear distinction between noise and details in noisy images. This helps Neat Image reduce more noise and better preserve true details in digital photos and scans.
Neat Image builds and uses device noise profiles to adapt noise reduction to imaging device. A device noise profile is a reusable analysis of noise properties of an image acquisition device (a digital camera, scanner, etc.) working in a certain mode. Using a noise profile for an imaging device in effect makes Neat Image noise reduction custom-tailored to this imaging device. Device noise profile is a novel concept originally introduced by Neat Image team and supported by Neat Image user community who create noise profiles for many digital cameras and scanners.
If noise profiles for a specific camera are not yet added to our profile library, Neat Image will still be able to clean photos from that camera using the built-in Auto Profile function, which provides the easiest and quickest way to automatically build a noise profile in just one click.
Neat Image is also very fast because it is thoroughly optimized for parallel processing on multi-core CPUs and GPUs (NVIDIA CUDA and AMD OpenCL). In particular, Neat Image can use all available cores in multi-core processors as well as all processors in multiprocessor systems. Using a graphical accelerator can increase speed even more, especially when the GPU is used in conjunction with CPU or multiple GPUs are utilized. Depending on filtration parameters and available hardware, Neat Image is capable of filtering 30-70 megapixels a second or more on a modern computer.