Capacity Guide

Understanding how much data you can hide in an image depends on three factors: the image resolution (dimensions), the number of color channels, and the embedding density (bits per element).

Embedding Density explained

Embedding density is measured in bits per element (bpe).

• In LSB, you might use a higher density because the changes are simple.
• In HILL or S-UNIWARD, we typically cap the density at 0.05 bpe (5%) to ensure the modifications remain undetectable by modern steganalysis.

Common Capacity Tables (RGB Images)

The table below shows the maximum usable payload size for standard 24-bit RGB images (3 channels) at various safety levels.

Dimensions	Total Pixels	0.01 bpe (Ultra Stealth)	0.03 bpe (Balanced)	0.05 bpe (Safety Cap)
640x480	307,200	1.1 KiB	3.3 KiB	5.6 KiB
1280x720 (HD)	921,600	3.3 KiB	10.1 KiB	16.8 KiB
1920x1080 (FHD)	2,073,600	7.5 KiB	22.7 KiB	37.9 KiB
2560x1440 (QHD)	3,686,400	13.4 KiB	40.4 KiB	67.4 KiB
3840x2160 (4K)	8,294,400	30.3 KiB	91.1 KiB	151.8 KiB

The Math Behind the Limits

For a standard spatial image (PNG), the capacity is calculated as follows:

1. Calculate total elements: Width * Height * Channels (e.g., 3 for RGB).
2. Apply density: Multiply by the target bpe (e.g., 0.05).
3. Subtract overhead: HideMyByte subtracts ~36 bytes for encryption headers (Salt, IV) and aligns data to 16-byte blocks for security.

Important Constraints

• JPEG Capacity: Unlike PNG, JPEG capacity is content-dependent. It depends on the number of non-zero coefficients after compression. Busy, textured images can hold significantly more data than smooth, flat images.
• Payload Compression: We recommend compressing your payload (e.g., as a .zip or .gz) before embedding to maximize the usable space.
• Stealth Trade-off: As you push past 0.05 bpe, the statistical "noise" introduced by the embedding becomes increasingly detectable by automated steganalysis tools.