When to use Parallel Stereo Recordings vs agent-controlled

VICIdial gives you two ways to capture a stereo track: let the Agent control it, or run a Parallel Stereo Recording that captures every second of every call regardless of what the agent does. They can also run together. Picking the right combination depends on why you need the recording in the first place.

What each approach guarantees

Agent-controlled Stereo recording — through ONDEMAND, ALLCALLS, or ALLFORCE — ties the stereo file to agent actions. ALLFORCE comes closest to guaranteed capture, but even then the recording starts when the call arrives at the agent screen, not at the carrier level.

Parallel Stereo Recordings removes agent involvement entirely. The recording starts the moment the call enters the Agent session and stops when it leaves — no button, no option, no gap.

Decision guide

flowchart TD
  A[Why do you need stereo?] --> B{Compliance or audit trail}
  A --> C{QA sampling or coaching}
  A --> D{Agent self-review or flagging}
  B --> E[Use Parallel FULL-RECORDING]
  C --> F[Agent-controlled ALLFORCE or Parallel]
  D --> G[Agent-controlled ONDEMAND]
  E --> H[Both can run together for redundancy]

Compliance and regulated industries

If your call center operates under TCPA, financial services rules, or other regulations that require a complete audio record, Parallel Stereo Recordings with FULL-RECORDING is the right choice. There is no way for an agent — accidentally or intentionally — to create a gap in the recording.

**Heads up:** Running Parallel Stereo Recordings alongside agent-controlled ALLFORCE means two stereo files per call. Budget your disk space and Recording retention policy for the additional footprint before enabling both.

QA and voice analytics programs

QA teams that review a sampled percentage of calls often find agent-controlled ALLFORCE sufficient. The agent cannot stop the stereo track, and the files land with consistent names that a QA platform can ingest automatically.

If you feed 100% of calls into a voice analytics engine — for example, to score every interaction automatically — Parallel Stereo Recordings with FULL-RECORDING ensures no call slips through. The consistent full-call duration also makes it easier to calculate Talk time and Wrap-up time from the audio itself.

Agent coaching and self-review

When agents are the primary consumers of recordings — listening back to their own calls as part of a coaching program — ONDEMAND agent-controlled recording makes more sense. Agents flag calls they want to review, and only those calls generate a stereo file. Storage stays lean and agents stay engaged in the process.

Running both at the same time

Nothing stops you from running both together. A common pattern for financial services teams is: Parallel Stereo FULL-RECORDING for the compliance archive, plus agent-controlled ONDEMAND for agents to flag calls worth coaching. Two files per flagged call, one file for every other call. The CO, CM, and FR filename prefixes on the parallel files keep them clearly separated from agent-initiated stereo files.

For the full recording architecture overview, see VICIdial call recording explained. For how ALLCALLS and ONDEMAND compare in standard recording, see when to use ONDEMAND vs ALLCALLS recording.

Want a hosted VICIdial system with stereo recording configured and ready? See VICIfast pricing.