What Does Cg Mean In Photography And Retouching?

Whenever I spot 'CG' in a shoot credit or a retoucher's notes, I picture two worlds colliding: real-life photography and computer-made magic. In simple terms, 'CG' stands for computer graphics or computer-generated imagery — that includes anything rendered in 3D software or painted digitally rather than captured straight from a camera. In photography and retouching it often means that parts (or all) of the image were modeled, textured, lit, and rendered in tools like Blender, Maya, Cinema 4D, or Substance, and then composited into the photo. Sometimes it's subtle — a CG prop, a sky replacement, or a hair strand — and other times it's the whole scene recreated for absolute control.

From a practical retouching point of view, working with CG means thinking like both a photographer and a lighting artist. I match camera focal length, film back, and lens distortion; I make sure the rendered light uses HDRI maps or physically based lights so reflections and shadows behave. Then I composite render passes (diffuse, specular, ambient occlusion, beauty, z-depth) in Photoshop or Nuke, tweak color grading and add matching grain, chromatic aberration, and motion blur so the CG reads like it belongs. If the CG is used for product shots, you get cleaner seams and perfect materials; for portraits, CG can fix backgrounds or add elements that would be impossible on set.

People sometimes worry that CG will feel fake, but the trick is honesty: matching imperfections. Real cameras have noise, slight flare, bokeh quirks, and subtle color casts — and adding those to renders helps them sit in the photo. I love how CG expands creative choices: you can sculpt impossible lighting, swap textures in a heartbeat, or preview a shot entirely in 3D before wasting training models or props. It keeps work flexible and, when done well, totally believable. Personally, I get a kick out of the hunt for the tiny details that sell the illusion.

Quick breakdown: 'CG' means computer graphics — stuff made or rendered in 3D or painted digitally instead of photo-captured. In everyday retouching that can be as small as a CG prop or as big as an entirely rendered background or product. The key challenge I think about is visual consistency. A render gives you perfect geometry and lighting control, but a camera image carries grain, lens imperfections, depth of field quirks, and color casts. To sell the combo I match camera settings, add matching film grain, mimic lens bokeh and distortion, and composite using additive passes (specular, diffuse, AO, etc.).

From my experience, good collaboration between the artist who made the render and the retoucher speeds things up: when I get properly labeled EXRs and information about the render camera and lights, I can integrate elements faster and more convincingly. On shoots where CG replaces dangerous or impractical setups, it’s a lifesaver. I still love the little victory of nudging a shadow or warming a highlight until a CG element finally 'breathes' like the photo around it — feels like cheating in the best way possible.

In practical terms, 'CG' in photography means any element generated on a computer rather than captured by the sensor. When I'm finishing a commercial image, that might mean a product was modeled and rendered so every angle looks perfect, or a background was digitally constructed to match a campaign aesthetic. The retoucher's role then becomes compositing and tonal matching — taking the rendered pieces and blending them with photographs so the viewer can't tell where reality ends and rendering begins.

Technically, this leads to a specific workflow I appreciate: obtain or set the camera parameters (focal length, aperture, sensor size), render with physically based materials and an HDRI for environment lighting, export layered EXR passes, and composite. I pay attention to linear-light workflows, LUTs, and gamma to avoid color shifts. Cryptomatte and AOVs save time when isolating materials for targeted color or specular adjustments. On the retouch side, techniques like frequency separation, dodge-and-burn, grain overlays, and edge treatment help meld pixels from different origins.

There are frequent pitfalls — scale mismatches, wrong shadow softness, or CG that's too clean compared to the photo. So I deliberately add 'flaws': slight dirt on reflections, micro-scratches on metal, or a subtle misfocus cue. In advertising, CG is liberating because it handles complex reflections, impossible camera rigs, and rapid iterations. Personally, I prefer projects where CG and photo skills are both respected; the best images feel like a single crafted moment rather than a collage.