A remote light meter for photography that can be used, via Bluetooth LE, from an Android/Linux phone.
Python C Java JavaScript HTML C++ Other
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
android_client
buspirate_client
pics
puremodules
python_client
rfduino_server
.gitignore
CONTRIBUTING
DISCLAIMER
LICENSE
README.md
bootstrap_config.json

README.md

Blue Light Meter

This project develops a remote light meter for photography that can be used, via Bluetooth LE, from an Android/Linux phone. Maybe some day in the future a lightmeter GATT profile could be developed and camera may use it. Miniaturized BTLE enabled light meters could be easily carried around or concealed in clothes while shooting.

Hardware

The initial hardware is very simple. A TSL2561 Sparkfun breakout is connected to a RFDuino module (the 4 lines SCA, SCL, +3.3V and GND). The biggest part is the RFDuino coin cell battery module. The battery itself would probably be the biggest part in a custom designed hardware, considered the dimension of the nRF51822 and the light sensor. I compared the result to low-cost light meters and the agreement is OK, but more testing is needed. Also I'm experimenting with different sensors that also deliver an information about the color temperature of the light. An enclosure made with a 3D printer is in the plans, the most problematic part is the diffuser to protect the light sensor.

The prototype: BlueLightMeter and BlueLightMeter2.

Software

The functionality is pretty standard as for all light meters. You choose which one you want to calculate from the aperture value, the exposure time and the ISO. You set the value for the other 2. You can select normal or flash mode. In the first case the calculation will be done on the average measurement from the last 3 seconds, otherwise the maximum will be used.

The rest of the controls is specific to the TSL2561 sensor. You can see the medium and maximum measured Luxs or EV in the last 3 seconds. Then you can select if the parameters for the sensor are set manually or automatically by the app. In the former case you can select the integrating time from the built-ins ore use a custom one. Also you can enable hi-gain mode. If you choose automatic mode the app itself will try to choose the right parameters (having a significant reading without risking a saturation). You can choose between using all available modes, only faster ones (for quicker response) or try to use low gain (for better SNR). There are also debug information available, status of connection and, after it is established, the value read from the sensor. For the Android version also the value read from the internal light sensor (in Lux) is available to check the BlueLightMeter readings. This value is also used if BLM is not found, so yes, this is also yet another built-in light meter application.

RFDuino server

The application uses the simple main loop approach idling while broadcasting its identification. When connected it starts reading the light sensor, sending the data and receiving configuration updates. It was tested with RFDuino BSP 2.3.1. Before compiling (simply invoking make) you have to set the environment variables RFDUINO to the location of the BSP (usually $HOME/.arduino15/packages/RFduino/hardware/RFduino/2.3.1/ if you followed standard installation) and RFDUINO_PORT to the USB port used (/dev/ttyUSB0 or following, depending on which devices you have connected). The target make download installs the compiled application to RFDuino.

BusPirate client

This is a simple BusPirate library in C that I use for testing sensors. It has an implementation of the TSL2561 reading algorithms.

Python client

The Python client uses dbus interface to bluetoothd for accessing the RFDuino. Please note that the experimental interface for accessing GATT services is need, so assure that bluetoothd is started with the -E option (eventually by editing the configuration file, for systemd it is bluetooth.service). The clients needs python-dbus and pygtk-2.0 Python modules.

The client is structured in 3 separate classes for simpler reuse. BLM is the low-level communication with RFDuino that can be used as a template for generically accessing Bluetooth LE device under Linux/Python. BLMThread implements the sensor algorithms and GUI is the user interface.

Android client

For a rapid development, only the communication part is in Java, the sensor algorithms and the UI are in JavaScript and use a responsive Bootstrap user interface (running in a WebView). The BlueLightMeterDriver class implements the low-level communication with RFDuino. The data gathered are exported, through the data-model class BlueLightMeterData, as BLM to the WebView. BlueLightMeterData also exposes methods for setting the sensor parameters.

IMPORTANT: I cannot include code code copyrighted by a third party. You need to add in android_client/assets:

  • jquery-2.1.4.min.js. If you install a later version you have to change blm.html as well.

  • A customized Twitter Bootstrap for handling a responsive layout for screen rotation. Yo can download the customization but a copy of the customize.json is available in the BlueLightMeter repository. Feel free to contact me if you have any problems.

Contacts

Feel free to drop me a line Christian Pellegrin.