Tech cue sheet automation for live theatre

Define

Problem Statement

A sketch comedy show produced in 10-hours (conceive → write → rehearse → perform → receive adulation) required complex technical cuesheets:

• Multiple background videos (often custom edited)
• Sound effects and music cues
• Lighting cues coordinated with QLab (cueing software)
• Physical props and scene changes

Baseline process:

• Tech operator manually compiled cuesheets during the day
• Assets finalized minutes before show, sometimes missing
• Manual data entry into QLab cueing system
• Zero time for dress rehearsal with complete tech
• Tech failures during performance due to last-minute cuesheet changes or simple lack of rehearsal

Critical incident:

• Complex video sketch (Five minutes to midnight) required processed video: Berlin walkthrough → reversed → sped up → cut with soundtrack → timed to scene beats
• All in all 1/5 of the ten hours was spent getting the assets show ready
• Tech operator insisted on building entire cuesheet manually per established process
• Show had video-heavy opening, rapper doing battle with sound effects, multiple video cues
• Tech not finalized until break between shows (8pm → 9:30pm program)
• Result: Actor running against static background, soundtrack timing off, audience confusion

Project Goal: Reduce cuesheet creation from hours (with frazzled operator) to <2 minutes, enabling full tech rehearsal and last-minute creative changes.

Scope

• In scope: Cuesheet generation from media assets, QLab automation, stop cues
• Out of scope: Media editing pipeline, lighting design, sound mixing

---

Measure

Baseline Data

Metric	Baseline Value
Manual cuesheet creation time	2–4 hours
Operator state after build	Frazzled
Time available for tech dress	0
Last-minute asset changes	Impossible
Cuesheet errors during show	Regular

---

Improve

Solution Design

Insight: All cuesheet creation consisted of:

1. Scanning organized folder structure
2. Creating QLab groups per sketch with lights up and blackout cues
3. Adding video/image files with stop cues
4. Adding sound effect cues
5. Adding default images (cafe, park, restaurant) for missing assets

Automation approach:

Built script that:

• Reads organized folder structure (sketch name → ordered assets)
• Communicates with QLab via JSX
• Creates groups per sketch with auto-stop cues
• Embeds default placeholder images for missing video
• Outputs ready-to-load QLab project

Initial implementation: AppleScript proved insufficient for complex data types

Intended refinement: OSC (Open Sound Control) protocol for more flexible QLab integration

Results

Deployment: Single 10-hour sprint production

Metric	Before	After	Change
Cuesheet generation time	2–4 hrs	<1 min	-99%
Operator readiness	Frazzled	Calm	Improved
Tech dress rehearsal	~25% of sketches	All sketches	+300%
Last-minute asset changes	None	Easy	Enabled
Performance tech issues	Regular	Resolved	Eliminated

---

Lessons Learned

1. Repetitive process automation was the constraint, not the quality of tech work. The operator’s skill wasn’t the issue; the manual data entry was killing all available time.

2. Wrong tool choice JSX didn’t kill the project, but proved less flexible than OSC would have been, e.g., users could add media via local webservice

3. Default placeholder assets enabled graceful degradation. When final video assets weren’t ready, the show could still proceed with placeholder visuals instead of failing entirely.

4. Integration into existing workflow is critical. The script output had to save directly as a QLab project that could be opened immediately - no additional conversion steps.