RapidRAW's masking system integrates three types of specialized functions: it supports simultaneous management of 32 mask layers; it provides 7 types of selection creation methods, such as brush, gradient, color range, etc.; and it implements Boolean (add/subtract/intersect) mask combination logic. These features enable it to handle complex needs such as gradient ND filter simulation in landscape photography and local skin color correction in portrait retouching.
In terms of technical realization, the software adopts GPU-accelerated real-time mask rendering, with an operating latency of less than 16ms on 4K displays.The AI-assisted masking function is based on convolutional neural networks, and is able to achieve an auto-recognition accuracy rate of 92% for common shooting subjects (e.g., portraits, animals, and buildings). Case tests show that when separating hair details in complex backgrounds, it outperforms the auto selection function of most consumer-grade editing software.
This answer comes from the articleRapidRAW: A non-destructive tool for fast RAW image editingThe
