Getting Started

The following section describes how to get started with your Time Tagger.

First, please install the most recent driver/software, including graphical user interfaces, libraries, and examples for C++, Python, .NET, C#, LabVIEW, Matlab, and Mathematica.

You are highly encouraged to read the sections below to get started with the graphical user interface and/or the Time Tagger programming libraries.

Additional information about the hardware, API, etc. can be found in the menu bar on the left and on our main website:

If you are using Linux, please take a look at the Linux section.

Get familiar with your Time Tagger

To learn more about the Time Tagger you are encouraged to consult the following resources.

  1. Run the Graphical User Interface, Time Tagger Lab or the Web Application, to play with your Time Tagger interactively. The examples below will allow you to experience basic data acquisition. Play with the Time Tagger settings to see their effects on typical measurements.

  2. Check out the Application Programming Interface chapter. It gives you a detailed overview of all Time Tagger features and data processing classes. Check out the following sections to get started using the Time Tagger software library in the programming language of your choice.

  3. Study the code examples in the .\examples\<language>\ folders of your Time Tagger installation.

Graphical User Interfaces

Time Tagger Lab


Time Tagger Lab is the Graphical User Interface (GUI) application for Windows operating systems. It is designed to perform standard measurements quickly and to get an interactive experience with your Time Tagger.

  1. Download and install the most recent Time Tagger software from our downloads site.

  2. Start Time Tagger Lab from the Windows start menu.

  3. The Time Tagger Lab welcome screen opens and presents all available devices. If you plug or unplug a Time Tagger, the display will update automatically.


Time Tagger Lab allows you to work with your Time Tagger interactively. We will now use the Time Tagger’s internal test signal to measure a cross correlation between two channels as an example.

  1. On the welcome screen, click on the panel representing your desired device. The Home screen of your Time Tagger shows up.

  2. On the top of the Home screen, you can toggle between the basic display you are currently looking at and the Detailed View. Activate the Detailed View.

  3. In the upper panel of the Detailed View, you can change settings for every input. Activate the internal test signal on inputs 1 and 2 by checking the boxes on the rightmost column.

  4. Go back to the basic display by deactivating Detailed view. The count rate displays (cps) show a test signal count rate of 800 to 900 kHz, and the trigger level setting is replaced by the label test signal.

With the activated test signal, we can start our first measurement.

  1. Click on Open creator (F2) in the Measurements panel on the left. The creator gives you an overview of all available measurements.

  2. Pick Bidirectional histogram - Correlation. Now you can read through a detailed description and change the initial settings of your measurement.

  3. Set Reference channel to 1 and Click channel to 2. Click Add measurement.

  4. The measurement graph is showing up. To start the measurement, click the Play button next to the graph.

  5. A Gaussian peak should be displayed. You can zoom in using the controls on the plot. If the resolution is not sufficient, try a smaller Bin width in the measurement settings.

  6. The detection jitter of a single channel is \frac{1}{\sqrt{2}} times the standard deviation of this two-channel measurement (the FWHM of the Gaussian peak is 2.35 times its standard deviation).

You have just verified the time resolution (detection jitter) of your Time Tagger.

Web Application

The Web Application is the traditional provided GUI for Windows and Linux for quick measurements. On Windows systems, we recommend switching to Time Tagger Lab today.

  1. Download and install the most recent Time Tagger software from our downloads site.

  2. Start the Time Tagger Application from the Windows start menu.

  3. The Web Application should show up in your browser.


The Web Application uses the TCP port 50120 as default port. If this collides with another application you can change the port with passing the argument TimeTaggerServer.exe -p 50120.

The Web Application allows you to work with your Time Tagger interactively. We will now use the Time Tagger’s internal test signal to measure a cross correlation between two channels as an example.

  1. Click Add TimeTagger, click Init (select resolution if available) on any of the available Time Taggers

  2. Click Create measurement, look for Bidirectional Histogram (Class: Correlation) and click Create next to it.

  3. Select Rising edge 1 for Channel 1 and Rising edge 2 for Channel 2.

  4. Set Binwidth to 10 ps and leave Number of data points at 1000, click Initialize.

The Time Tagger is now acquiring data, but it does not yet have a signal. We will now enable its internal test signal.

  1. On the top left, click on the settings wheel next to Time Tagger.

  2. On the far right, check Test signal for channels 1 and 2, click Ok.

  3. A Gaussian peak should be displayed. You can zoom in using the controls on the plot.

  4. The detection jitter of a single channel is \frac{1}{\sqrt{2}} times the standard deviation of this two-channel measurement (the FWHM of the Gaussian peak is 2.35 times its standard deviation).

You have just verified the time resolution (detection jitter) of your Time Tagger.

Programming languages


  1. Make sure that your Time Tagger device is connected to your computer and the Time Tagger Web Application (especially the server window) is closed.

  2. Make sure the Time Tagger software and a Python distribution (on Windows OS we recommend Anaconda) are installed.

  3. Open a command shell and navigate to the .\examples\python\1-Quickstart folder in your Time Tagger installation directory

  4. Start an ipython shell with plotting support by entering ipython --pylab

  5. Run the script by entering run hello_world

The hello_world executes a simple yet useful measurement that demonstrates many essential features of the Time Tagger programming interface:

  1. Connect your Time Tagger

  2. Start the built-in test signal (~0.8 MHz square wave) and apply it to channels 1 and 2

  3. Control the trigger level of your inputs - although it is not necessary here

  4. Initialize a standard measurement (Correlation) in order to find the delay of the test signal between channel 1 and 2

  5. How to control the delay of different inputs programmatically.

You are encouraged to open and read the file in an editor to understand what it is doing. With this basic knowledge, you can explore the other examples in the 1-Quickstart folder:



Classes & Methods

Basic software control (folder 1-basic_software_control)


Create a measurement

createTimeTagger(), Counter.getData(),


Count rate trace



Start & stop measurements

Countrate, start(), stop(), startFor()


Synchronize measurements


Use different histograms

Correlation, Histogram,

StartStop, HistogramLogBins


Virtual Channels

DelayedChannel, Coincidence, GatedChannel


Logging errors



(External) software clock



Controlling the hardware (folder 2-controlling-the-hardware)


Get hardware information

scanTimeTagger(), TimeTagger.getSerial(),

TimeTagger.getModel(), TimeTagger.getSensorData(),



The input trigger level

TimeTagger.setTriggerLevel(), TimeTagger.getDACRange()


Filter tags on hardware




Control input delays

TimeTaggerBase.setInputDelay(), Histogram2D






HighRes mode

createTimeTagger(), TimeDifferences

Dump and re-analyze time-tags (folder 3-dump-and-reanalyze-time-tags)


Dump tags by FileWriter



The Time Tagger Virtual

createTimeTaggerVirtual(), TimeTaggerVirtual

Working with raw time-tags (folder 4-working-with-raw-time-tags)


The FileReader

FileReader, TimeTagStreamBuffer


Streaming raw time-tags



Custom Measurements


More details about the software interface are covered by the API documentation in the subsequent section

LabVIEW (via .NET)

In LabVIEW, you can access and program your Time Tagger through .NET interoperability.


A set of examples is provided in .\examples\LabVIEW\ for LabVIEW 2014 and higher (32 and 64 bit).

Matlab (wrapper for .NET)

Wrapper classes are provided for Matlab so that native Matlab variables can be used.

The Time Tagger toolbox is automatically installed during the setup. If TimeTagger is not available in your Matlab environment try to reinstall the toolbox from .\driver\Matlab\TimeTaggerMatlab.mltbx.

The following changes in respect to the .NET library have been made:

  • static functions are available through the TimeTagger class

  • all classes except for the TimeTagger, TimeTaggerNetwork, and TimeTaggerVirtual classes will have a TT prefix (e.g. TTCountrate) to prevent conflict with any variables/classes in your Matlab environment

An example of how to use the Time Tagger with Matlab can be found in .\examples\Matlab\.

Wolfram Mathematica (via .NET)

Time Tagger functionality is provided to Mathematica via .NET interoperability interface. Please take a look at the examples in .\examples\Mathematica\.


We provide a .NET class library (32, 64 bit and CIL) for the TimeTagger which can be used to access the TimeTagger from many high-level languages.

The following are important to note:

  • Namespace: SwabianInstruments.TimeTagger

  • the corresponding library .\driver\xxx\SwabianInstruments.TimeTagger.dll is registered in the Global Assembly Cache (GAC)

  • static functions (e.g. to create an instance of a TimeTagger) are accessible via SwabianInstruments.TimeTagger.TT


A sample Visual Studio C# project provided in the .\examples\csharp\Quickstart directory covers the basics of how to use the Time Tagger .NET API. An example of creating ‘custom measurements’ is also included.

Please copy the project folder to a directory within the user environment such that files can be written within the directory.

An ‘Example Suite’ is provided in the .\examples\csharp\ExampleSuite directory. ‘Example Suite’ is an interactive application that demonstrates various measurements that can be performed with the TimeTagger. Reference source code to setup and plot (with OxyPlot) each measurement is also provided within the application. Additionally, the application contains examples for creating and using ‘Virtual channels’, ‘Filtering’ and ‘Accessing the raw time tags’.


Running the Example Suite requires ‘.NET Core 3.1 Desktop Runtime (v3.1.10)’.


The provided Visual Studio C++ project can be found in .\examples\cpp\. Using the C++ interface is the most performant way to interact with the Time Tagger as it supports writing custom measurement classes with no overhead. But it is more elaborate compared to the other high-level languages. Please visit .\documentation\Time Tagger C++ API Manual.pdf for more details on the C++ API.


  • the C++ headers are stored in the .\driver\include\ folder

  • the final assembly must link .\driver\xYZ\TimeTagger.lib

  • the library .\driver\xYZ\TimeTagger.dll is linked with the shared v142 or newer Visual Studio runtime (/MD)

  • use TimeTaggerD.lib and TimeTaggerD.dll for the Visual Studio debug runtime (/MDd)

  • use libTimeTagger.dll and libTimeTagger.a for the MinGW C++ ABI for the MINGW32 and UCRT64 environment.

Debug and release builds: Performance issues with debug builds

For this topic, it is required to be a bit more precise on what is called a Debug build. Visual Studio default compiler flags for Debug are /MDd /O0 /Zi and for Release /MD /O2 /DNDEBUG.

  • /O0 vs /O2

    This is the general optimization level. /O0 means no optimization, so every instruction is surrounded by load_from_memory and store_to_memory. This is a huge waste of CPU resources, but it guarantees that every local variable can be inspected at every time. We suggest to use the default optimization level /O2 by default and to overwrite it only for required methods using pragmas: MSVC optimize pragma.

  • /MDd vs /MD

    This flag selects which (incompatible) runtime C++ ABI you want to use for the whole project. The debug runtime MDd has additional fields and checks for e.g. better range checking. It is good practice to run your code with them at least once before each commit. As they are incompatible with each other, ALL linked C++ libraries MUST use the same runtime. This flag is the only difference between our debug TimeTaggerD.dll and non-debug TimeTagger.dll library. We suggest to use MD even for debug builds, else the debug runtime will degenerate the overall performance also in the internal Time Tagger code.

  • /Zi

    This flag tells the compiler to generate a *.pdb symbol file. This is used by the debugger to map e.g. the assembly to the C++ code and the stack to the local variable. As this has no performance overhead, always use this flag for all internal builds.


    The macro NDEBUG is the C way to disable assert(). They are usually good at catching same errors without much overhead. We also like to recommend to switch your compiler to CLANG and use its undefined behavior sanitizer, the address sanitizerm or the thread sanitizer. Microsoft announced it a bit ago that MSVC now has the address sanitizer, which finds out-of-bounds accesses, use-after-free and similar issues: AddressSanitizer (ASan) for Windows with MSVC.

So in the end, it is not about just using Debug or Release. It is about which set of flags is used. Please select each of them carefully based on your target.