Once the software starts, the New Project dialog appears.


The default Quick start tab allows the user to select three important settings:

  1. The project type and camera used for acquisition if any camera is available
  2. The top-level data directory
  3. The computational device



perClass Mira allows the users to either start from existing scans already stored in files or to acquire new scans from an attached spectral camera. There is a broad range of project types and many common spectral camera types available to start form. You may view all available options in the Scans in files and Live acquisition tabs.


From these tabs items can be marked as favorite by right clicking them. Items marked as favorite will be shown in Quick start tab lists.


Live acquisition in perClass Mira always uses the perClass project type. This assures that users of all camera types can take advantage of identical and complete work-flows for data correction.


The Top level data directory defines where the scans of a new project are located, and where all scans will be saved to. It is stored in the project file, but it can be changed at any time from the Image list context menu. perClass Mira does not write into the data directory unless the user explicitly asks to, for example when exporting the results or recording a new scan.


The Computational device combo box allows the user to define what computational device will be used for data processing. Each perClass Mira installation comes with two executables:

  • perClass_Mira.exe which is CPU-only and works on any PC.
  • perClass_Mira_gpu.exe which offers multiple backends including CPU, CUDA (for NVIDIA GPUs), and OpenCL (for CPU/GPU). If no correct GPU drivers are found, then this executable may not be able to start.


TIP: In case you experience crashes or very slow operation when using a GPU, please update your GPU drivers to the latest available, this resolves issues for most users.


WARNING: If any crashes happen when using the OpenCL backend with an Intel device, then please use the standard CPU backend.