Published by PitchChanger · Updated · How these guides are made
How to use this workflow
Write each peak as amplitude
Convert a peak P in dBFS with 10^(P/20). Work per channel.
Add the amplitudes
Their sum is an upper bound on the absolute sample peak. Convert back with 20 × log10(sum).
Mix and measure in a DAW
Apply gain where the mixer supports it, then inspect the rendered output. Keep margin for reconstruction peaks and later encoding.
Four aligned copies at −12 dBFS
Each copy has peak amplitude 0.251189. Four simultaneous positive peaks sum to 1.004755, or about +0.0412 dBFS. A source that looked quiet alone can cross full scale when copied and aligned.
For equal copies, reducing each by 20 × log10(N) preserves the original worst-case peak. Two copies need about 6.0206 dB reduction each; four need 12.0412 dB. These are authored arithmetic bounds, not measured stem exports.
Swipe or scroll sideways to read all columns.
| Equal aligned copies | Maximum peak rise | Each source at −12 dBFS yields |
|---|---|---|
| 1 | 0 dB | −12 dBFS |
| 2 | 6.0206 dB | −5.9794 dBFS |
| 4 | 12.0412 dB | +0.0412 dBFS |
Do not mistake the bound for a loudness prediction
Opposite-polarity copies can cancel; unrelated peaks may occur at different times. A simple sum bound cannot predict integrated loudness, perceived balance or true peak. Do not sum negative dB numbers or assume the +3 dB power rule describes identical in-phase audio.
Stem Splitter provides synchronized playback, mute, solo and named mix exports. It does not expose a DAW-style independent gain fader for every track. Export stems to an external editor for gain-controlled recombination and inspect the actual result there.