The Tools Behind a Written Gallery Workflow Review

When a broadcaster asks us for a Live Production Workflow Review, they often expect us to arrive with software. We arrive with a notebook, a clipboard, a laptop with a spreadsheet template and a reasonable amount of patience. The review is a piece of written work, and the tools that serve it best are the ones that make a gallery's habits visible without disturbing them. Most reliability problems are not caused by a faulty piece of kit. They come from a procedure that only one person remembers, a signal path nobody has drawn since the last refit, or a fallback that has never been rehearsed. So our method is built to expose those gaps in a calm, repeatable way, and to put the findings on paper in terms a crew can act on the next morning.

This is the third part of our notes on the review. Earlier pieces cover the vocabulary and the reasoning; here we describe what we physically use during a visit, and what we leave in the van. Nothing here is exotic, which is rather the point. A crew that has to keep running a live programme every week needs documents it can read and maintain itself, not a dependency on us.

The first tool is a hand-drawn signal path. Before we look at any documentation the client has supplied, we walk the gallery from the cameras back to the vision mixer, from the mixer to the outputs, and from the talkback panels to the studio floor, and we sketch what we find as we go. Cables are followed, patch bays are opened and labels are read aloud to whoever is nearest. The sketch is then compared with the diagram the studio holds, if there is one. The differences are usually the most useful finding of the first day. A spare route that was added for a one-off show and never removed, a monitor feed that depends on a single converter, an intercom channel with no obvious owner: each of these goes into a numbered list with a photograph and a note of who confirmed it.

We redraw the whole thing cleanly afterwards, keeping one page per signal family so that audio, video, timing and communications can each be read without the others getting in the way.

• Follow every route physically before trusting a diagram
• Note single points of failure in a separate numbered list
• Redraw one page per signal family for the final report

The second tool is the run-of-show itself. We ask for the last few that the crew actually used, marked up in pen if possible, rather than the clean templates kept in a shared folder. The scribbled corrections tell us where the plan and the reality part company: a package that always runs long, a cue that is regularly called late, a handover between presenters that needs three attempts. We then sit in the gallery during a rehearsal or a live programme, silent and out of the way, and time what happens against the page. We log each cue, each camera change and each moment where someone leans across to ask a question. Questions asked during a broadcast are a reliable sign that something is missing from the paperwork.

Daniel Okafor, who writes most of our run-of-show and fallback documents, insists on one rule that we keep: every instruction must be something a person can do with their hands in under ten seconds, otherwise it does not belong on the live page.

Fallback plans are the area where paper and practice most often disagree. A plan may say that the backup vision mixer takes over within moments, yet nobody has pressed the button since commissioning. We agree a quiet window, usually outside transmission, and test each fallback in the order the plan lists them, with a stopwatch and a second pair of eyes. We test the unglamorous ones as well: switching to the spare talkback beltpack, changing the programme feed to the alternate output, restoring the studio clock after a reference failure, moving a presenter's mic to the standby channel. Each test is recorded as passed, passed with a note, or failed, together with how long it took and who carried it out. Failures are not embarrassing; they are the reason for the exercise. A fallback that fails in a quiet window costs nothing, whereas the same failure during a live programme costs a great deal.

Where a test cannot be done safely, we say so in the report and describe what evidence we relied on instead.

• Test in the order the written plan lists
• Time every step and name who performed it
• Record results as passed, passed with a note, or failed

The report is the product, so we spend as long on it as on the visit. It opens with a short summary a studio manager can read in five minutes, followed by the findings ranked by how much they threaten on-air reliability, and then by how easy they are to fix. Each finding states what we saw, why it matters and what we suggest, in that order and in plain sentences. We attach the redrawn signal paths, the annotated run-of-show, the fallback test log and a revised set of procedure sheets ready for printing. Where we recommend equipment changes, we describe the function needed rather than steering towards a particular product, because the decision usually involves budgets and supplier relationships that sit with the client.

A good report is also short enough to be read twice. We would rather give a gallery twelve clear pages it uses than forty it files away.

There are tools we do not bring. We do not install monitoring software to record the crew's every action, because a gallery that feels watched behaves differently and the findings become less honest. We do not hand over a slide deck of generic best practice, since the useful material is specific to the room and the people in it. We also avoid reviewing a studio in a single sitting.

Two shorter visits, a few days apart, show far more than one long one, because the crew's routine reasserts itself once the novelty of an observer has worn off. The second visit is when the real habits appear, and when we can check whether a small change made after the first visit has held. The discipline throughout is to keep the method plain enough that the crew could repeat it themselves, which in our experience is the best measure of whether the review was worth doing.