5.1 → Stereo downmix · ITU-R BS.775

All processing happens locally in your browser. Nothing is uploaded. Inputs are read in chunks (never fully loaded) and the output is written chunk-by-chunk into a download — the JS heap stays flat, with the bytes held in the browser's disk-backed blob store. Output is an interleaved stereo WAV, written as RF64 when it would exceed the 4 GB classic-WAV limit.

1 · Channels (drop a mono WAV on each slot)

Or drop up to 6 files here at once — they fill empty slots in order (L, R, C, LFE, Ls, Rs).

2 · Downmix options

Center (C) level into L & R

Surround (Ls/Rs) level

Include LFE

dB (often omitted per ITU)

Output format

Attenuate to prevent clipping Scales output down only if the sum peaks above 0 dBFS.

Force RF64 container Otherwise classic WAV is used below 4 GB.

Measure EBU R128 Loudness + true peak. Slow on very long files — untick to skip.

3 · Result spectrum (engineering view)

4 · EBU R128 loudness (ITU-R BS.1770-4)

About the downmix coefficients

Following ITU-R BS.775 the stereo downmix is:

Lo	= L + g_c·C + g_s·Ls (+ g_lfe·LFE)
Ro	= R + g_c·C + g_s·Rs (+ g_lfe·LFE)

The recommended levels are g_c = g_s = −3 dB (0.7071). LFE is normally discarded in the downmix. A slot left empty is treated as silence. All source files must share the same sample rate; shorter files are zero-padded to the longest.

Supported input: PCM 16/24/32-bit, 32-bit IEEE float, mono or multichannel (channel 1 is used), RIFF/WAVE and RF64.