Building the Studio Conners Audio Pipeline

We started with a fairly specific question: how should Studio Conners create original audio for its vertical videos?

That quickly became a bigger production experiment.

We built and tested a local ComfyUI workflow for generating original music, then used it against a real Studio Conners video plan. The first track worked technically but not structurally. It developed at the wrong time and ended too early for the story we had planned.

That failure turned out to be useful.

Instead of treating music as something that gets dropped underneath a finished edit, we changed the workflow. We can plan the story and rough visual beats first, generate the soundtrack, listen to what it actually does, and then use its transitions, buildup, payoff, and ending to help determine the final cut.

In other words, the audio becomes a checkpoint.

The day also made something else clear: producing a polished video after spending an entire day on an Obsession is probably the wrong rhythm. So we began designing a different one — weekday work can feed a queue, while a dedicated production day handles sound, generated visuals, editing, masters, and distribution.

That led to the beginnings of a new Studio Conners tool we’re calling the Video Desk: one place to hold the script, assets, soundtrack, cut plan, generation instructions, edit directions, masters, and distribution record for every video.

What started as an audio experiment ended up defining much more of how Studio Conners will make video.