Monthly lab clinic — bring a frame, watch the evidence trail

Stage this page moves: → 2 (frames + metadata received), and it re-engages stage-1 labs that went quiet. Buyer pages: reproducibility bundle, workflow fit, methods supplement, teaching module (point instructors here; it reuses the clinic's walkthrough).

Once a month, a free ~45-minute online clinic. A participant brings one compatible frame with its metadata and watches it go from frame + settings to a result ID, evidence record and methods supplement, live. Small: up to 6 participants, 2–3 frames shown.

Format


When	same week each month (e.g. second Thursday), 45 minutes, a time that works for Europe and the Americas (16:00 UTC)
Where	a video call link from the company account; no recording unless every participant agrees in writing beforehand
Who	researchers from stage-1 labs, talk attendees, referrals, inbound
Cost	free; nothing to buy
Host	you; the operator or an engineer optional for `needs_review` setups

Sign-up handling

Invitation (plain text) to every stage-1 lab and every talk attendee, with a reply-to-register line. No public sign-up form; reply-to-register keeps it small and lets you check fit first.
Before confirming a seat, run the fit check (page 24). unsupported → reply with the reasons, no seat. needs_review → a seat as an observer, frame discussed, not run live.
Ask each confirmed participant for one frame + wavelength, pitch, distance (and a blank frame if they have one) at least 2 working days before. Run it once privately to make sure it completes.

Artifact: a sign-up sheet (date, name, org, fit verdict, frame received y/n, consent y/n).

Consent for showing frames

Nothing is shown on screen without written consent from the frame owner, by email, before the session:

For the <date> clinic: may we show the frame you sent (<filename>) and its
reconstruction on screen to the other participants (up to 6 researchers from
other labs)? We will not record, share or keep it for any other purpose, and
we delete it after the session unless you ask us to keep it on your account.
Reply "yes, you may show it" to confirm. If not, we will walk through our
public sample instead and send you your own result privately.

No reply = no consent = public sample on screen, their result sent privately.

Evidence-trail walkthrough script (per frame, ~10 minutes)

"Here is the frame sent, with wavelength, pitch and distance."
Submit it (show the two-step public call or the signed-in upload).
Open the result ZIP. "This is the result ID — a content address. Same frame plus same settings gives the same ID; change one setting and it changes."
Open the evidence record: software version, SHA-256 of the input, SHA-256 of the config. "Anyone with the frame and these settings can recompute and compare."
Open the methods supplement. "This is the paragraph a student can adapt for a methods section; it cites the result ID."
Say what it does not show: "This proves how the result was made. It does not tell us how close it is to the true surface — that needs a reference measurement on your side, which is what a design partnership does."
Questions from the owner first, then the room.

Follow-up (within 24 hours)

Each frame owner: their result ZIP privately, the reproducibility bundle page, and the stage-3 ask: "do you have one independent measurement of the same target?"
Each observer: their fit verdict and an ask for one frame.
Delete shown frames unless the owner asked us to keep them; log the deletion.

Artifact: the clinic row in the log (date, participants, frames shown, consents filed, follow-ups sent, deletions logged).