Find out why a parallel drum bus sounds hollow

The kick has weight, the compressed drum return sounds exciting, and the two together sound like somebody changed the tuning. Keep the drummer’s timing intact while you investigate. You can learn a great deal from one quiet test click, two identical signal paths and a short piece of the song.

Align two drum paths A snare drum sits below two transient tracks that merge into a single return.
Give the dry drum and the parallel return their own print tracks. A single combined bounce hides the timing difference.
Trace the parallel paths to their sum A mono test click feeds a dry path and a compression aux. Two print taps sit immediately before their common summing bus. Start with matched gains and identical processing. Invert one path only for that control, listen for cancellation, then restore polarity. With the crush processor engaged, mute the print tracks during capture. Then audition the recorded prints separately and compare their transients at sample resolution.Mono test clickSame sample on each pathDry pathUnity gainCrush auxBypass for controlPrint tap APre-sumPrint tap BPre-sumClean sum busControl: invert one pathRestore polarityReinstate crush and compare prints
Start with matched gains and identical processing. Invert one path only for that control, listen for cancellation, then restore polarity. With the crush processor engaged, mute the print tracks during capture. Then audition the recorded prints separately and compare their transients at sample resolution.
Enlarge diagram
Trace the parallel paths to their sum A mono test click feeds a dry path and a compression aux. Two print taps sit immediately before their common summing bus. Start with matched gains and identical processing. Invert one path only for that control, listen for cancellation, then restore polarity. With the crush processor engaged, mute the print tracks during capture. Then audition the recorded prints separately and compare their transients at sample resolution.Mono test clickSame sample on each pathDry pathUnity gainCrush auxBypass for controlPrint tap APre-sumPrint tap BPre-sumClean sum busControl: invert one pathRestore polarityReinstate crush and compare prints

Build an identical-path control

Save a diagnostic copy of the mix. Lower the monitor level, leave some output headroom and choose eight bars with a kick, a snare and a few rests. Place a short, quiet mono click before the drum excerpt. Allow a bar of silence before the click so that all processing is running by the time it arrives. Write down the sample rate. For the example below it is 48 kHz.

Feed the same mono source to two paths. Send the source output to a dry auxiliary channel and use a unity-gain send to feed a second auxiliary channel called Crush. Route both channel outputs to a clean stereo bus. Check that the source has no third route to the main output. Remove effects from both paths for the first test, match channel formats, center the pans and set both aux faders to unity. Keep automation off in this diagnostic copy. Listen to each path individually. Both should sound like the same unprocessed drum part.

Invert the polarity of one path on both output channels. Polarity inversion changes the sign of every sample. If both paths deliver identical samples at identical gains and times to the summing point, adding them produces silence. If the meter still shows the whole loop at a lower level, check the send gain, pan law, stereo balance and hidden dry route. One-sided cancellation calls for checking the left and right routing separately. Restore normal polarity before listening to the song.

Put the suspect processor back

Insert the compressor on Crush, set its mix fully wet and start with automatic makeup, saturation and modeled noise disabled wherever possible. Set the ratio to unity or set the threshold high enough that the test click and excerpt trigger no reduction. Keep the intended lookahead or oversampling settings active. A bypass switch can change how a host handles a plug-in’s delay. The goal of this pass is to keep the plug-in in the path with as little intentional waveform change as its design permits.

Enable plug-in delay compensation for the relevant tracks, buses and outputs. In Logic the All setting covers auxiliary and output channel strips. Test ordinary playback with low-latency monitoring off and tracks disarmed. In Live, reduced-latency monitoring can make an armed path and a return arrive at different times even when normal playback aligns. Compare recording-mode monitoring separately if the drummer first noticed the problem while playing. Logic’s plug-in latency settings · Live’s reduced-latency monitoring guide

Print both paths immediately before the common bus from the same timeline position, preferably in the same pass. Keep the two print tracks muted during recording. In an empty session, zoom to the first click at sample resolution. Suppose the dry click begins at sample 96,000 and the return at 96,096. That difference is 96 samples, or 2 milliseconds at 48 kHz. Check a later hit. A constant displacement on neutral paths is evidence of a timing offset. Check compensation mode, intentional track delay, external loop settings and the latency the plug-in reports before introducing an offset of your own. Ableton’s explanation of plug-in delay compensation

Interpret the difference you hear

In the test session, move the late print earlier by the measured amount and repeat the polarity test. If alignment substantially reduces the residual, you have a reproducible observation to take back to the live mix. Keep this correction on the diagnostic print. A permanent nudge of the drum performance could be wrong the next time you change a plug-in or recording mode.

A residual with matching onset times calls for another test. Bypass a colored EQ, disable modulation and compare the control again. EQ can change phase differently at different frequencies. Saturation adds harmonics. Once you restore the intended compression, gain changes over the hit are the point of the parallel path. Two differently shaped waveforms can begin together and leave a strong residual when subtracted. Moving their largest peaks together risks shifting the original attacks.

Finally, restore the song, normal polarity and the intended compressor settings. Bring up Crush from silence under the dry drums and trim the combined level for a fair comparison. Listen in mono, then stereo. Does the quiet snare detail become clearer while the kick still lands once? Print the loudest chorus and repeat the check outside the session. Save the measured offset, the setting that corrected it and the mix revision alongside the bounce.

Related reading

Back to blog

Leave a comment