Client-Side Photo RGB Histogram, Exposure EV & WebP Compressor (2026)

An RGB histogram averages or overlays the raw 0–255 values of the Red, Green, and Blue subpixels equally. A Rec.709 Luminance histogram weights each channel according to human eye cone sensitivity ($Y = 0.2126R + 0.7152G + 0.0722B$), because human vision is over 3x more sensitive to green light than red and nearly 10x more sensitive to green than blue.

When photographing saturated red roses, neon signs, or sunset skies, the Red channel can hit 255 (losing all texture detail) while Green and Blue remain low. Because Green dominates the Luminance formula ($71.5\%$), the overall Luminance histogram stays well below 255—which is why inspecting individual R, G, and B channels is essential.

At base ISO 100, Exposure Value is defined as $EV_{100} = \log_2(N^2 / t)$, where $N$ is the aperture f-number (e.g., $f/2.8$) and $t$ is exposure time in seconds (e.g., $1/250$). When shooting at a higher ISO $S$, the scene luminance EV adjusted to ISO 100 is $EV_{100} = \log_2(N^2 / t) - \log_2(S / 100)$. Bright daylight ('Sunny 16') equals $EV_{100} \approx 15$.

Google's WebP format uses VP8 predictive intra-frame block coding and entropy encoding, typically achieving 25% to 35% smaller file sizes than mozjpeg/standard JPEG at SSIM-equivalent quality (around quality 80–82), plus native 8-bit alpha channel transparency.

No. Both the pixel-level `getImageData()` histogram tally and the `canvas.toBlob('image/webp')` encoder execute 100% locally inside your browser.