Six Common Gallery Wiring and Integration Mistakes, and How to Avoid Them

The most frequent mistake is buying and installing equipment piece by piece, each item chosen on its own merits. A vision mixer, a monitor wall, an intercom system and a playout server can each be excellent and still fail to work together on a live day. Clock sources differ, signal formats disagree, and the person who has to make them cooperate is usually a gallery operator ten minutes before transmission.

I approach a gallery as one system from the first drawing. Before anything is ordered, we write down every source, every destination and every format that will travel between them. That single sheet shows where a converter is needed, where a signal is split too many times, and which device quietly assumes a setting that another does not share.

It also gives the whole crew one reference. When the engineer, the director and the person who will inherit the room all read the same page, arguments about what was meant happen in the planning stage, where they cost nothing.

• Draw every signal path before ordering
• Agree one format and frame rate for the whole chain
• Keep a single source of truth for the room

A gallery that looks fine on a quiet afternoon can drift out of step under load. Audio arrives slightly after the picture, a remote guest appears a beat late, or a graphics machine shows a frame that no other source matches. These faults are small, and presenters notice them long before viewers can explain what is wrong.

The cure is a proper reference signal feeding every device that can accept one, and a written note of the delay each processing stage adds. I measure the latency of monitors and converters rather than trusting the figure on the data sheet, because the real number depends on the settings in use.

Where a delay cannot be removed, we compensate for it deliberately in the audio path and record it. Undocumented compensation is almost as bad as none, since the next person to change a setting will undo it without knowing why it was there.

Intercom is treated as an accessory, bought last from whatever budget remains. On a live programme it is the nervous system of the gallery, and a poor setup shows immediately: directors shouting over each other, floor staff who cannot hear a cue, and talkback that bleeds into a presenter's earpiece.

I plan the intercom alongside the vision and audio routing, deciding who needs to speak to whom and who only needs to listen. Separate channels for the director, the floor and the engineering team stop a technical query from interrupting a cue to talent.

Headset choice matters as well. Comfort over a two-hour show, a clear boom microphone and a sensible mute button remove a surprising number of errors. We test every position at the volume it will really run at, with the gallery fully populated, not in an empty room.

• Separate channels by role, not by habit
• Test headsets for a full-length show
• Check earpiece bleed before rehearsal

A tidy rack is not vanity. Cables run in a hurry, unlabelled and bundled with power, produce the intermittent faults that take hours to find. Interference from a mains lead lying alongside an audio line is a classic, and it tends to appear only when a particular piece of equipment is switched on.

We separate power from signal, leave service loops at both ends, and label each cable at each end with a name that matches the drawing. It adds time on installation day and saves far more of it over the following years.

Spare capacity deserves a mention too. A patch bay with every point in use leaves no room for the guest camera or the extra microphone a production will ask for in month three. Planning for a modest amount of growth costs little now and avoids a rewire later.

Individual tests pass, so the room is declared ready. Then the first real programme runs every source, every graphic, a remote contact and a playout roll in a single sequence, and the weak joint appears. Testing parts is not the same as testing the whole chain under realistic pressure.

I insist on at least one complete run that mirrors a real show: the same rundown, the same number of people, the same pace. Problems found in a rehearsal are an inconvenience. The same problems found on air are a story the crew retells for years.

During that run we also practise failure. Pull a source, drop a network link, lose a microphone, and watch how the team responds. The aim is not to alarm anyone but to make sure the fallback path exists, is labelled, and that someone knows where it is.

• Rehearse with the real rundown
• Practise one planned failure
• Note every fix in the room log

A well-built gallery can still become a liability if nobody understands it after the installers leave. Staff change, memories fade, and a configuration known only to one engineer disappears when that engineer moves on. I have walked into rooms where nobody could say why a particular switch was set the way it was.

Every integration we finish ends with a handover pack: the signal drawings, device settings, IP addresses, reference sources and a plain list of what to check before each show. We walk the operators through it on the floor, not in a meeting room.

A short written routine for start-up and shut-down is the single most useful page. It lets a new operator bring the room up safely, and it gives the whole team the same starting point when something does go wrong.