Open for Voting
P: Better frequency tools to reduce the harsh "artificial look" from digital photos
One of the biggest challenges when editing digital photos is reducing the artificial, overly crisp appearance that’s often associated with the “digital look.” While current frequency-based tools in ACR (texture, clarity and dehaze) are useful, they fall short when trying to soften this digital harshness. Pushing these sliders to the left often produces undesirable artifacts and side effects.
Main limitations with current tools (texture, clarity and dehaze):
- Color bleeding and smearing: Reducing clarity or dehaze tends to spread colors beyond their natural boundaries, causing them to mix with neighboring tones in an unnatural way.
- Unwanted darkening: If there is a large area without highlights, what area will become overly dark when reducing dehaze (adding haze). Usually this ends up darkening faces or other relevant parts of the photo.
- Dark halos and artifacts: When reducing clarity, a noticeable dark halo often appears around dark objects placed against midtone backgrounds, producing an unflattering overprocessed look (the opposite of our goal).
- Color temperature is massively affected by Dehaze tool.
The proposed idea would be to have a set of similar tools, but:
- Brightness-only diffusion: Blur or soften only the brightness (luminance) channel, without affecting color saturation or hue (or maybe reducing saturation but never increasing it). This would avoid the "color smearing" and temperature shift seen in current tools. Plus: A slider to select how much of the color is diffused (apart of the brightness).
- Asymmetric diffusion direction: Allow brightness to spread from brighter to darker areas, but not the reverse. This is to prevent undesirable dark halos (clarity) and to avoid darkening already dark areas (dehaze).
- Ideally: Frequency-based curve control: Instead of fixed sliders for low, medium, and high frequency ranges (like texture, clarity, and dehaze), provide a customizable frequency-response curve, where:
- The horizontal axis represents spatial frequency (from low to high detail)
- The vertical axis controls how much diffusion is applied at each frequency
This would give users precise control over how softness is applied across the image, enabling more organic and subtle results, much like the way light behaves in analog film or lenses with natural falloff.
Why this matters:
The digital image pipeline has become incredibly sharp and clean, but often too much so. There is growing demand among photographers, retouchers, and visual artists for tools that reintroduce softness and imperfection in a controlled and elegant way. This proposal would fill that gap.
Would love to hear if others also feel limited by the current softening tools, and if this approach resonates with your workflow.
AdChoices