About

Raw byte data represents pixel information in its most fundamental form: sequential octets encoding color channels per the selected bit depth. Misinterpreting byte order, row padding, or channel count produces corrupted output. A 640×480 image at 24-bit depth requires exactly 921,600 bytes (3 bytes per pixel). The BMP format adds row padding to align each scanline to 4-byte boundaries. Omitting this padding shifts every subsequent row. This tool parses decimal, hexadecimal, base64, and JSON array inputs, maps bytes to RGBA pixel data at the correct stride, and exports standards-compliant BMP files with proper BITMAPINFOHEADER structures or lossless PNG via Canvas.

The converter auto-detects optimal image dimensions from byte count when width is unspecified. It factors the total pixel count and selects the pair of factors closest to a 1:1 aspect ratio. For incomplete final rows the tool zero-pads remaining bytes. Note: this tool assumes linear byte order (no embedded headers or compression). If your source data contains file headers, strip them before conversion.

Formulas

The total byte count required for an uncompressed bitmap is determined by color depth and dimensions:

B = W × H × D8

Where B = total bytes, W = width in pixels, H = height in pixels, D = bit depth (1, 8, 24, or 32).

BMP row padding ensures each scanline aligns to a 4-byte boundary:

P = (4 − (W × D8) mod 4) mod 4

Where P = padding bytes appended to each row. For 32-bit images, P is always 0 since 4 bytes per pixel inherently aligns.

Auto-dimension detection finds integer factor pair (w, h) of total pixel count N that minimizes:

|w − h| → min, where w × h ≥ N

BMP file size calculation:

FileSize = 14 + 40 + ColorTable + (W × D8 + P) × H

Where ColorTable = 0 for 24/32-bit, 1024 bytes (256 × 4) for 8-bit, and 8 bytes (2 × 4) for 1-bit.

Reference Data

Color Depth	Bytes/Pixel	Channels	Total Colors	Use Case	Row Padding Rule
1-bit Monochrome	0.125	1 (packed)	2	Barcodes, QR masks	Pad to 4-byte boundary
8-bit Grayscale	1	1	256	Medical imaging, heightmaps	Pad to 4-byte boundary
16-bit RGB555	2	3 (packed)	32,768	Legacy displays, embedded	Pad to 4-byte boundary
24-bit RGB	3	3 (R, G, B)	16,777,216	Photos, standard graphics	Pad to 4-byte boundary
32-bit RGBA	4	4 (R, G, B, A)	4,294,967,296	Transparency, compositing	No padding needed (already aligned)
BMP File Structure
BITMAPFILEHEADER	14 bytes		Signature (0x42 0x4D = "BM"), file size, data offset
BITMAPINFOHEADER	40 bytes		Width, height, bit depth, compression, image size
Color Table	Variable		Required for ≤ 8-bit. 4 bytes per entry (B, G, R, 0x00)
Pixel Data	Variable		Bottom-up row order, BGR channel order
Common Input Formats
Decimal CSV	`255, 128, 0, 255, 0, 128`
Hex (0x prefix)	`0xFF, 0x80, 0x00, 0xFF, 0x00, 0x80`
Hex (plain)	`FF 80 00 FF 00 80`
JSON Array	`[255, 128, 0, 255, 0, 128]`
Base64	`/4AAgACA` (binary encoded)
C-style Array	`{0xFF, 0x80, 0x00}`
Dimension Auto-Detection
100 pixels	10 × 10		Perfect square
1920 pixels	48 × 40		Closest factors to square
307,200 pixels	640 × 480		VGA resolution
2,073,600 pixels	1920 × 1080		Full HD

Frequently Asked Questions

This occurs when the specified width does not match the actual scanline width of your source data. Each row must consume exactly W × bytes-per-pixel bytes. If the width is off by even 1 pixel, every subsequent row shifts, creating a diagonal shearing artifact. Verify your width matches the original image dimensions, or use the auto-detect feature which factors the total pixel count to find matching dimensions.

Input bytes are interpreted in standard order: R, G, B (and A for 32-bit). This matches the memory layout used by Canvas ImageData and most programming languages. Note that the BMP file format internally stores pixels in BGR order - this tool handles the channel swap automatically during BMP export. If your source data is already BGR (e.g., from a raw BMP dump), swap the R and B channels in your source or the output colors will have inverted red/blue.

Each byte encodes 8 pixels. Bit 7 (MSB) is the leftmost pixel, bit 0 (LSB) is the rightmost. A 1 bit maps to white (255, 255, 255), a 0 bit maps to black (0, 0, 0). For a 100-pixel-wide image, each row consumes 13 bytes (with 4 unused bits in the last byte). BMP padding still applies to the byte-level row width.

The tool caps text input at 16 MB of raw text. For 24-bit RGB, this translates to approximately 5.3 million pixels (roughly 2300 × 2300). Canvas has its own limits - most browsers cap at 16,384 × 16,384 pixels. For very large arrays, use file upload instead of pasting text, as the FileReader API handles binary data more efficiently than string parsing.

BMP is uncompressed. Every pixel is stored verbatim, plus row padding bytes. A 100 × 100 image at 24-bit depth produces a BMP of 30,054 bytes (headers) + padded pixel data. PNG applies DEFLATE compression and delta filtering, often achieving 50 - 90% size reduction for images with uniform regions. Use PNG for distribution and BMP only when the specification demands uncompressed raster data.

Yes. Export the memory region as a binary file or hex dump. For frame buffers, you need to know the exact resolution and color depth. Common embedded formats include RGB565 (16-bit, not directly supported - convert to 24-bit first by expanding 5/6/5 bits to 8/8/8). Set the width to match the display's horizontal resolution. If the dump includes non-pixel header data, strip those bytes before input.