---
title: Extended-mirror Fryett Q optimizer
---

# Extended-mirror Fryett Q optimizer

**Worker:** not active · **status:** `paused_success_q100k_resumable` ·
**checkpoint age:** 28.9 min · **accepted updates:** 12 ·
**candidate trials:** 14

This frequency-free 16 µm campaign is forked from the preserved 8 µm cavity. Its outer five complete cells imply a 256.589 nm period; their registered two-dimensional median cell is repeated through the added mirror. It uses an adaptive replay-validated L-BFGS-B trust region.
It is paused at this accepted landmark and remains exactly resumable for later binarization and smooth-boundary work.

| quantity | current |
|---|---:|
| design length | **16 µm** |
| inferred mirror period | 256.589 nm |
| parent 8 µm Q | 6321.57 |
| independent 4 ps seed audit Q | **42361.1** |
| 4 ps seed field / energy Q | 42361.1 / 42362.7 |
| mirror splice | x=3.85 µm in each half |
| trusted Q | **138284** |
| best trusted Q | 138284 |
| progress to Q=100,000 | 138.284% |
| field-fit Q | 138284 |
| energy-fit Q | 139046 |
| estimator ratio | 1.00551 (must be ≤1.25) |
| pole wavelength | 820.137 nm |
| source-center wavelength | 819.771 nm |
| wavelength constraint | **off** |
| source-center updates | 2 |
| normalized mode volume | 1.41469 (λ/n_target)³ |
| objective | 11.8398 = log harmonic Q |
| trust radius | 0.0189844 raw density |
| L-BFGS memory | 10 / 10 pairs |
| last candidate | accepted |
| fixed FDTD tape | 2.0 ps |
| binarization / projection / filter | **off / off / off** |
| physical beta | **not measured** |

## Q and optimizer history

![Q and trust history](../_static/generated/fryett_shifted_q_history.png)

The Q axis is linear. The plot stays zoomed to the achieved values so early
progress remains visible; the 100,000 goal is annotated until it enters range.
Only candidates whose field and energy lifetimes agree and whose smooth harmonic-Q objective increases can enter the accepted curve. Frequency is telemetry, not an acceptance constraint. Red trust markers are forward-only rejected trials.

## Ringdown evidence

![Measured response and pole fit](../_static/generated/fryett_shifted_q_response.png)

There is no prescribed unity-amplitude ideal trace in this campaign. The
orange curve is the fitted single pole in the actual source-off data, with the
same observed normalization as the measured blue field. Absolute waveform MSE
is retained as telemetry only and cannot trade Q away for amplitude or phase.

## Geometry and field

![Geometry evolution](../_static/generated/fryett_shifted_q_geometry.png)

![Tracked field](../_static/generated/fryett_shifted_q_fields.png)

{download}`Exact campaign methodology <../_static/downloads/fryett/PORT_DIPOLE_EXTENDED_MIRROR_Q_METHOD.md>`

{download}`Plateau evidence motivating the handoff <../_static/downloads/fryett/PORT_DIPOLE_PLATEAU_AUDIT.md>`