Discrete color or luminance data (R, G, B) in a digital image — processable independently. RAW files carry more channels than compressed formats.
Every digital recording breaks down into its color components—these are the channels. You work with them every day without thinking much about it. Red, Green, Blue: three separate streams of information that together form the image you see on the monitor. In editing or color correction, you can process each channel in isolation without touching the others. This is the core power of the concept.
In a compressed H.264 file or ProRes, you typically get four channels: RGB plus Alpha (transparency). With RAW material—whether from a RED, ARRI, or Blackmagic—there are significantly more. For example, a BRAW file stores raw data with extended color information and metadata in separate channels. This allows you to precisely control color casts later without affecting the entire image. You adjust red while blue remains untouched—precise work instead of a broad stroke.
Practical on set and in editing: If a scene has a green cast (common with LED lighting), you address the green channel in isolation. In Fairlight audio or image processing in DaVinci Resolve, colorists routinely work channel-based. You create masks for a single channel, apply curves separately, and define quality layers. RAW material offers exponentially more room for maneuver than 8-bit compressed—it's not for nothing that production houses pay for RAW workflows.
An important side effect: The file size grows with the number of channels. A 12-bit RAW clip is many times larger than a log-encoded ProRes file. Therefore, a conscious decision is made on set: Do I need the flexibility of extra channels, or is compressed sufficient for this shoot? For green screen or critical color grading scenes: RAW, no question. For documentary shoots with tight storage budgets: Log-compressed and then handled carefully in editing. The channel architecture determines your entire post-production pipeline.