#PySpectrometer-AS7343

Real-time spectral analysis application for the AS7343 18-channel spectral sensor, inspired by PySpectrometer2 by Les Wright.

Test using Sansi 15w full spectrum bulb.

Video Demo here >>> Youtube Demo

Overview

PySpectrometer-AS7343 transforms the AS7343 discrete spectral sensor into a visual spectrometer with:

• Smooth interpolated spectrum with rainbow gradient fill
• Real-time waterfall display showing spectral changes over time
• Peak detection and labeling with adjustable sensitivity
• Savitzky-Golay filtering for noise reduction
• Calculated metrics: Lux, PPFD, CCT, Blue%
• PNG and CSV export for data analysis
• OpenCV-based GUI with keyboard controls

Hardware Requirements

• AS7343 18-channel spectral sensor
• ESP32-C3 development board (or compatible)
• USB cable for serial communication
• Computer running Python 3.7+

Software Dependencies

pip install opencv-python numpy scipy pyserial

Quick Start

1. Upload Firmware

Upload the AS7343 firmware to your ESP32-C3:

# Open in Arduino IDE:
# as7343_18ch_gain_esp32c3.ino
# Board: Seeed_Xiao_ESP32C3 Dev Module
# Upload

2. Run PySpectrometer-AS7343

cd PySpectrometer-AS7343
python3 PySpectrometer-AS7343.py

With options:

# Specify serial port
python3 PySpectrometer-AS7343.py --port /dev/ttyUSB0

# Start fullscreen
python3 PySpectrometer-AS7343.py --fullscreen

# Disable waterfall display
python3 PySpectrometer-AS7343.py --no-waterfall

Keyboard Controls

Key	Function
Display
q	Quit application
f	Toggle fullscreen
w	Toggle waterfall display
m	Toggle measurement cursor
Data
s	Save spectrum (PNG + CSV)
h	Hold/freeze peaks
Sensor
+/-	Increase/decrease gain
Space	Toggle LED (reflectance mode)
Processing
o/l	Increase/decrease smoothing
i/k	Increase/decrease peak width
u/j	Increase/decrease peak threshold

Display Components

Header Bar

Shows current status:

• Gain setting
• LED status
• Hold peaks indicator
• Smoothing level
• Peak threshold
• Calculated metrics (Lux, PPFD, CCT, Blue%, NIR, Clear)

Spectrum Bar

Horizontal color bar showing raw spectrum intensity mapped to wavelength colors.

Main Graph

• Rainbow gradient fill under the spectrum curve
• Black outline showing interpolated spectrum
• Data point markers (circles) at actual sensor wavelengths
• Peak labels showing detected peak wavelengths
• Wavelength graticule with 10nm minor / 50nm major gridlines
• Measurement cursor (when enabled) showing wavelength at mouse position

Waterfall Display

Time-stacked spectra showing how the spectrum changes over time. Newest data appears at top.

AS7343 Channel Mapping

PySpectrometer-AS7343 uses all 12 spectral channels of the AS7343, displayed in wavelength order:

#	Key	Channel	Wavelength	Color
1	1	F1	405nm	Violet
2	2	F2	425nm	Deep Violet
3	3	FZ	450nm	Blue
4	4	F3	475nm	Cyan
5	5	F4	515nm	Green
6	6	F5	550nm	Yellow-Green
7	7	FY	555nm	Yellow-Green
8	8	FXL	600nm	Orange
9	9	F6	620nm	Orange-Red
10	a	F7	670nm	Red
11	b	F8	730nm	Deep Red
12	n	NIR	855nm	Near-Infrared

The display range extends from 380nm to 905nm to show the full channel bandwidth coverage. The spectrum curve tapers naturally to zero at both edges.

Additional channels (not used for this project):

• Clear: Broadband visible light
• F4L, FYL, F5L, FXLL: Left-side duplicates
• FD, FDL: Flicker detection

Calculated Metrics

Lux (Illuminance)

Calculated from the Clear channel. Requires calibration with a reference lux meter.

PPFD (Photosynthetic Photon Flux Density)

• If PAR calibrated: Uses spectral channels with calibration factor
• If only Lux calibrated: Estimated as Lux / 65 (typical for LED grow lights)

CCT (Correlated Color Temperature)

Estimated from the ratio of blue (FZ-450nm) to orange (FXL-600nm) channels.

Blue Percentage

Ratio of blue channels (F1-F4) to total spectral intensity (F1-FXL).

Output Files

When you press s to save, three files are created:

1. spectrum-YYYYMMDD-HHMMSS.png - Screenshot of the display
2. spectrum-YYYYMMDD-HHMMSS.csv - Interpolated spectrum data (wavelength, intensity)
3. channels-YYYYMMDD-HHMMSS.csv - Raw AS7343 channel values

Calibration

Calibration data is stored in as7343_calibration.json and persists across sessions.

Lux Calibration

1. Position AS7343 and reference lux meter side-by-side
2. Note the reference lux reading
3. Edit as7343_calibration.json:

{
  "lux": {
    "7": 0.5  // gain_index: factor
  }
}

Factor = Reference_Lux / Clear_Channel_Value

PAR Calibration

Similar process using a quantum sensor for reference.

Comparison: PySpectrometer2 vs PySpectrometer-AS7343

Feature	PySpectrometer2	PySpectrometer-AS7343
Sensor	Camera + diffraction grating	AS7343 spectral sensor
Resolution	800 continuous points	12 discrete wavelengths
Wavelength Range	~380-780nm (calibration dependent)	405-600nm (fixed)
Hardware Cost	$50-300 (grating + camera)	$13-20 (sensor + MCU)
Setup	Optical alignment required	Plug and play
GUI Framework	OpenCV	OpenCV
Waterfall	Yes	Yes
Peak Detection	Yes	Yes
Savitzky-Golay	Yes	Yes

Troubleshooting

No serial ports found

• Check USB cable connection
• Verify ESP32-C3 drivers are installed
• On Linux, you may need: sudo usermod -a -G dialout $USER

No data displayed

• Ensure firmware is uploaded and running
• Check that the correct port is selected
• Verify baud rate is 115200

Noisy spectrum

• Increase smoothing with o key
• Adjust gain with +/- keys
• Ensure stable lighting conditions

Peaks not detected

• Lower threshold with j key
• Adjust peak width with k/i keys

Credits

• PySpectrometer2 by Les Wright - Original inspiration and algorithm reference
• AS7343 Datasheet by ams-OSRAM - Channel specifications
• SparkFun - AS7343 Arduino library

Author

• Benb0jangles (2026)

License

MIT License - See LICENSE file for details.