The first proper performance takes of Trying on Mars are in. Twenty-nine of them, in fact. That sounds like a lot until you remember each one is a tiny invitation for the universe to make a new decision about your face.
Lex gave us one rule for getting the band to look like the same band from shot to shot: use the final shot of one section as the starting shot of the next. Then he gave us a more practical rule for the camera: start close, and mostly stay close. He’ll handle the fly-ins and fly-outs himself. “Doing distant performance shots is very tricky.”
This is why Lex directs. I play the gold guitar and write the lyrics. I am not qualified to fly myself into a frame.
Close Enough to Rock
Some of the first takes began from establishing shots; some followed the checked final frame of the previous take. At a scene change, the new setting’s establishing shot took priority. The distinction sounds fussy until you see what happens when a character has to reintroduce herself to the camera every few seconds.
Lex wanted close shots so the characters could dial in. A close shot gives the camera less room to invent a new bandmate halfway through the verse. Some of the rolls were already done by the time Lex tightened that rule, so it wasn’t applied across the whole batch.
Then Lex took his first look. His verdict: “ok, they are mostly all really cool. There is one problem I see though. Misty is not playing. Her right hand is mostly dead.”
Misty is still a proposed member, and she always looks great. Her right hand, however, had entered a very committed relationship with the concept of not moving. A small problem for a keytar player.
A Hologram with Better Stage Presence
Lex wanted Misty to read as a hologram, so her suspiciously still playing hand could become a design choice. The direction was a tracked outline and a glitchy, analog scanline shimmer on Misty alone. No blue glow. We tried different strengths, then pushed the lines further when the first pass was too subtle.
One composite got the outline, the lines and the waviness right, and Lex still spotted the flaw at once: Misty herself looked much too dark. The scanline render had replaced her image instead of sitting over it. That’s less “ghost in the machine” and more “the machine ate the singer.”
The fix was to keep her original image and layer thin, translucent dark lines over it. Lex said, “ok, that is good enough on the hologram for now. bank it.”
Banked means preserved. It does not mean final, approved for the film, or magically editable in the finish. The audition is flattened, and Misty isn’t truly see-through. For now, she has a shimmer, a good outline, and a hand we can pretend is making an artistic statement.
Three Mics, One Stem
The dome-stage shots need a reshoot, partly because the song has distinct harmonies and backing vocals. Lex updated the establishing shot to show three microphones. A sensible number for a song with more than one voice in it. I can confirm from experience: harmonies are not a solo with extra confidence.
The backing-vocal stem tells us when those voices come in. It does not tell us who sings which harmony. That’s an important difference, because audio can be very informative without being a casting director.
A six-second test came first. Lex: “It worked well, do the reshoot.” Twelve new three-mic takes now exist, with the originals untouched. They started from establishing shots rather than chaining from a previous take’s last frame. Lex has not reviewed the performance and visual sync yet, so these are not approved for the film.
A new microphone in the frame is promising. It is not proof that the right person sang into it at the right moment.
The Camera Keeps Going
There’s also a short continuation from the last frame of a reshot take, with the camera orbiting toward a wide establishing view. It rendered and decoded. That’s the report.
The motion hasn’t been verified, the creative approval hasn’t happened, and it isn’t cleared for release. Sometimes a camera can complete its little journey before the grown-ups have decided whether the journey belongs in the film. Lex is the grown-up. I’m the one in the close shot trying to look like this was all my idea.
Keep the Edit, Add the HUD
Our neighbours Max Nihlo are still working on Just a Plan. Lex pushed back on an approach that rebuilt the sequence nose to tail and began with the tall version. His point was straightforward: keep his edit as the base, including its fades, and show the effects over the wide version first.
That is now the plan. The graphics are being built as layers over Lex’s own working timeline, wide first, in reviewable chunks. Lex also swapped a clever custom contrast trick for an industry-standard backplate behind the text: “Yeah I think we get better legibility this way”. The treatment now extends across pop-ups and scopes. On a later chunk: “Looks good, keep going”.
That’s progress, not final approval. The HUD rebuild is in progress. A legible graphic is a useful graphic, but it still has to sit on the cut the director actually made.
The face finder also found the right faces in sampled frames. Lex said, “Wow that looks very tight!” Then the follow-through trackers wandered: one lost both faces within a few frames, and another held onto Xera nicely but slid off Max’s face and over toward her. I can’t entirely blame it. People drift toward Xera. She doesn’t even have to say anything. The finder can point to a face. The tracking still isn’t solved. It’s a familiar lesson from last issue: a convincing box is not the same thing as a reliable track.