SuperRack SoundGrid and InTrigger: crackles and missed drum triggers

InTrigger: separate detection, playback and the drum signal path

Use a copy of the session and a short virtual-soundcheck loop to compare the crackle, hit detection and layer output. If you reduce the input level or the number of layers, also record which hits or dynamics stop being reproduced. Test changes separately on each drum before applying them to the whole kit.

These are proposed tests. They do not establish the cause of a problem or document measurements made by GLMedia. Keep a separate record for each setup and listening destination.

Per-core load and errors: save the current state before changing the buffer

Waves Common Questions, DSP performance meters explains that each rack is processed on one core. Overall average load and peak load describe different things; buffer settings, reloading and sample rate can redistribute the load. Record the passage and configuration used for each screenshot so you can compare them.

The 156-page SoundGrid manual, pp. 58–60 shows average and peak load per core, plus Routing View for connections. Page 37 distinguishes processing overload from high network usage: increasing Server Network Buffer can improve stability at the cost of added delay. A buffer adjustment alone does not establish the cause of a crackle.

Proposed test: use the same track and listening points, and capture the meters before the passage, during the crackle and at the end. Record the counter’s source and destination and the test duration. Save the initial state before clearing counters. Counts from different time periods are not directly comparable.

First, repeat the pattern without changing the session. Then compare a copy with fewer layers or less processing. In a separate test, change only the buffer. Also record missed hits and return delay. If one configuration stops the crackle, keep the file and result; the rest of the set still needs checking.

To distinguish processing capacity from backup capacity, the SoundGrid manual, p. 56 allows racks to be assigned to server groups and provides Test Redundancy. Waves Server Redundancy guidance recommends matching server models and warns of a mute during switchover, while the manual says “seamlessly.” Do not assume an inaudible switchover. The server Quick Start Guide, p. 4 describes Titan-R’s two power supplies. Backup power is different from a second server taking over DSP processing.

If you propose another active server for the triggers, draw which racks it will process and identify the backup for each group. Do not reassign a backup server on the assumption that it will still fill both roles. Schedule the switchover test for an agreed break and record what is heard at the destination.

InTrigger and Live: use the controls available in your component

The 49-page InTrigger manual, revision unspecified, pp. 4/14/49 describes Learn and Detection Graph in the standard component. Live uses Threshold and four layers; it does not include those functions or individual-hit editing. Confirm which component is loaded.

Save a screenshot and the actual plugin build before rehearsal to identify the component. The table summarizes the manual’s scope. Identify the components in your Quantum7 setup before following these steps.

The product page describes Live as “zero latency.” The manual, p. 49 distinguishes 0 ms Plugin Latency from 1.5 ms Detection Latency. These are plugin figures; measure the path to the monitors separately.

Missed hit or unexpected sample: trace input, detection and layer output

In standard InTrigger, pp. 9/14–16, a hit outside Target Scope will not trigger even if it exceeds Threshold. Retrigger sets the minimum interval; changing Focus or Separation, or enabling filters, resets detection. These checks address more than input level alone. For Live, follow the limits in the previous section.

Choose a passage with soft, loud and closely spaced hits, plus another with bleed. Mark the expected hits in a reference and repeat without changes. Record whether the input receives a signal, whether a hit is detected and whether the layer produces output. If detection occurs but the sample is missing, keep that distinction in the record; do not describe it as a hit the detector never saw.

In another copy, change only the control relevant to the stage you are checking. If that operation resets detection in the standard component, let the material play before comparing the new selection. Record when the change was made and which material was used. Running Learn on different passages can produce configurations that are not directly comparable.

To check layers, start with just one in the test copy. Then test the intended playback mode and combination, recording which layer you expect for each hit. Do not lower the entire input until the dynamics disappear without documenting what changed. Keep the original signal and return available during the test, listening to each separately before combining them.

Save a revision with the reference passage, close and reopen that copy outside the event, and compare the same hits. If the result differs, save both states for support without assigning a cause yet.

Test worksheet and handover for drums, FOH and monitors

Complete the InTrigger–SoundGrid worksheet with the component, builds, samples, layers, rack, group and server. Leave results blank until you run the tests.

Before rehearsal, agree on who watches input and detection and who records load and return audio. If front-of-house (FOH) and monitor engineers are involved, identify the destination for each comparison. A change that stops crackles but misses hits or delays the return needs further assessment. Keep both observations.

Common preparation errors include looking for Learn in Live, using overall average load as the only load check, combining counters from different sessions, or counting a redundant power supply as a second processor. Use the worksheet to hand over a reproducible symptom and the state the next operator must be able to restore.

Related guides: Performer vocal latency and effects returns, DiGiCo–SuperRack sessions and show continuity. Agree on the test and backup path with the operator of the event sound system.