Quantum Noise Analysis (QNA)

Overview

Quantum Noise Analysis (QNA) is a PyQt6 desktop application for quantitative noise analysis of scanning tunneling microscopy (STM) data. It is designed for frame-by-frame inspection of STM images and ROI-based comparison of local noise metrics across single images, STM frame series and multi-frame MPP datasets.

The application helps the user move from raw STM data to derived noise maps, per-frame metrics, potential-dependent plots, ROI-pair comparisons and STP export in one reproducible desktop workflow.

What QNA Does

QNA focuses on the quantitative analysis of local STM current fluctuations. For each loaded STM entry, the application can generate the derived noise map:

(I - I_SET)^2

where I is the image signal and I_SET is the user-defined tunneling-current setpoint. The software then combines this derived map with user-defined ROI geometry and potential values to compute and visualize noise-related quantities across frames.

Main capabilities:

load STM data from .mpp, .stp and .s94 files,
inspect raw STM frames and derived (I - I_SET)^2 maps,
define multiple named ROI pairs composed of Site ROI and Set ROI,
compute noise metrics for selected frames or for the full frame list,
analyze up to 25 ROI pairs in one workflow,
assign potential values from an external E(V) text file,
plot metrics versus frame number or potential,
review statistics, spot comparisons and frame comparisons,
export derived maps as .stp,
batch-save all derived maps,
convert listed .s94 files to .stp.

Supported Data

Input Formats

QNA currently supports:

Format	Use in QNA
`.mpp`	Omicron multi-frame data; used for frame-by-frame analysis.
`.stp`	Omicron single STM image.
`.s94`	SPECS STM image; can also be converted to `.stp`.

Current Loading Rules

Data can be loaded through File -> Open file(s)....
Directories can be loaded through File -> Open directory...; the current directory loader scans for .stp and .s94 files.
Only one .mpp file can be loaded at a time.
Each MPP frame appears as a separate list entry.
No example STM datasets are shipped with the repository; users should provide their own STM files.

Core Workflow

1. Load STM Data

Open a single STM file, multiple files or a directory of STM images.

Typical actions:

File -> Open file(s)...
File -> Open directory...

Loaded entries appear in the left panel. The list can be filtered by file type: All, STP, S94 or MPP.

2. Set Analysis Inputs

Set the tunneling-current setpoint used for noise-map computation:

I_SET (A)

For multi-frame potential-dependent workflows, load an E(V) text file:

Load E(V) file

The potential file is assigned row-by-row to the current list entries.

3. Define ROI Pairs

QNA uses paired regions of interest:

Site ROI - local region associated with the site or spot being evaluated.
Set ROI - reference area used for A_set when ROI mode is enabled.

The application supports multiple named ROI pairs. Only the active pair is editable, which reduces accidental modification during long analyses.

Useful ROI controls include:

add ROI pair,
duplicate ROI pair,
rename ROI pair,
remove ROI pair,
lock active pair,
switch ROI palette for visibility,
toggle ROI preview,
choose A_set: Full image or A_set: ROI.

4. Run Analysis

For a single selected entry:

File -> Analyze selected
Ctrl+R

For all entries and all ROI pairs:

File -> Full analysis (all frames)
Ctrl+Shift+R

Full analysis requires:

at least one ROI pair,
a Site ROI for every pair,
potential values assigned to all loaded entries.

5. Review Maps, Metrics and Plots

After analysis, QNA displays:

raw STM image,
derived (I - I_SET)^2 map,
ROI preview,
active-frame metrics,
property plot versus frame,
L0(E) / L(E) plot,
statistics across frames,
spot comparison table,
frame comparison plot or heatmap.

6. Export and Convert

Export actions are available from the File menu:

Save (I - I_SET)^2 map to .stp
Save all (I - I_SET)^2 maps to .stp
Convert S94 to STP (from list)

Current export support is centered on STP map export and S94-to-STP conversion. Native project/session saving and table export are not the primary implemented export path in the current codebase.

Computed Quantities

QNA computes and displays the following analysis quantities:

Quantity	Meaning in the QNA workflow
`L0`	Noise-related baseline quantity derived from the `(I - I_SET)^2` map and ROI/area normalization.
`L`	Corrected/faradic noise-related quantity derived from `L0`, offset and correction factor.
`K`	Correction factor based on the difference between mean current and `I_SET`.
`I_A`	Mean current over the selected reference area.
`D`	Area ratio between `A_site` and `A_set`.
`A_site`	Physical or pixel area of the Site ROI.
`A_set`	Full-image area or Set ROI area, depending on the selected mode.
`E`	Potential value assigned to the frame.

When physical scan dimensions are available in metadata, areas are reported in nm². Otherwise, QNA falls back to pixel units.

Plotting and Comparison Views

Property vs Frame

The Property vs frame plot shows selected metrics across the loaded frame list. Available metric choices include:

L0,
L,
K,
I_A,
D,
E,
A_site,
A_set.

The view is useful for checking frame-to-frame trends, outliers, unstable ROIs or potential-dependent changes.

L0(E) / L(E)

The L0(E) / L(E) view plots noise-related quantities as a function of potential. The user can choose the visible series:

L0 + L,
L0 only,
L only.

When SciPy is available and enough valid data points exist, QNA can compute optional sigmoid fits and display fit-related parameters such as E_inf and E_on.

Statistics Across Frames

The statistics view summarizes the active ROI pair across the loaded frame sequence, including metrics such as:

L0,
L,
D,
I_A,
K,
A_site,
A_set,
fit-derived values when available.

Spot Comparison

Spot comparison is intended for workflows with multiple ROI pairs. It helps compare active sites or candidate spots using fitted and summary quantities, including dL, E_on and L0/L statistics.

Frame Comparison

Frame comparison provides all-frame review across all ROI pairs. It is useful for checking whether trends are shared by multiple sites or localized to a specific ROI.

E(V) File Format

The potential file should be a plain text file with two whitespace-separated columns and no header:

1 0.100
2 0.125
3 0.150

Rules:

column 1: frame index,
column 2: potential value E(V),
row order is used to assign potentials to the current list entries,
the number of rows must match the number of loaded entries.

Typical Use Cases

Single-Image Noise Inspection

Use QNA to analyze a single .stp, .s94 or selected MPP frame.

Workflow:

Load one or more STM files.
Select the entry to inspect.
Set I_SET (A).
Define a Site ROI and optionally a Set ROI.
Choose A_set: Full image or A_set: ROI.
Run Analyze selected.
Inspect the derived (I - I_SET)^2 map and active metrics.
Export the derived map to .stp if needed.

This mode is useful for quick inspection of a single active site, checking whether a local region shows elevated noise and preparing derived maps for external STM/SPM tools.

Multi-Frame MPP Analysis

Use QNA for potential-dependent analysis across an MPP frame sequence.

Workflow:

Load one .mpp file.
Set I_SET (A).
Load the matching E(V) text file.
Define one or more ROI pairs.
Run Full analysis.
Review Property vs frame and L0(E) / L(E).
Compare spots using statistics, spot comparison and frame comparison views.
Export derived maps as .stp when needed.

This mode is intended for frame-by-frame STM noise analysis, comparing potential-dependent behavior between ROI pairs and screening multiple candidate active sites in one STM sequence.

Comparing Multiple Sites

Use multiple ROI pairs to compare sites or spots on the same STM frames. Pairs can be duplicated, renamed for example as spot 1, spot 2, reference, defect or terrace, and then compared using statistics and fit parameters.

STP Export and S94 Conversion

QNA can act as a bridge between SPECS/Omicron STM files and downstream tools that expect STP-compatible data. It supports .s94 to .stp conversion and saving (I - I_SET)^2 maps as .stp.

User Interface Overview

Left Panel

loaded entries and frames,
file-type filtering,
remove selected,
clear list,
keyboard navigation through frames.

Center Panel

I_SET input,
E(V) file loader,
potential input for current frame,
raw/derived image viewer,
colormap and level controls,
property and L0/L plots.

ROI Panel

named ROI pairs,
add / duplicate / rename / remove,
lock active pair,
ROI preview controls,
A_set mode switch.

Analysis Panels

active metrics,
statistics across frames,
spot comparison,
frame comparison.

Installation and Launch

Python 3.11 is recommended for the current repository state. After creating or activating an environment:

python -m pip install -r requirements.txt

Optional SciPy installation for sigmoid fitting:

python -m pip install scipy

Run the application:

python -m qna.main

Core dependencies from the repository include:

numpy,
PyQt6,
pyqtgraph.

Development Status

QNA is an active, focused research tool rather than a frozen commercial application. The current implementation already supports the central STM noise-analysis workflow: loading STM data, ROI-pair definition, derived noise maps, full-frame analysis, potential-dependent plotting, statistics/comparison views and STP export.

Current practical notes:

one .mpp file can be loaded at a time,
directory loading currently targets .stp and .s94, while .mpp is loaded through file selection,
sigmoid fitting is optional and depends on SciPy,
no example STM datasets are bundled with the repository,
project/session persistence is not the main implemented workflow at this stage,
export is focused on derived STP maps and S94-to-STP conversion.

Future extensions may include stronger result-table export, session persistence, richer batch reporting and additional QA utilities, depending on project needs.