Job: Downsample Particles

Downsample particles to save space.


Downsample particles and save them with smaller box sizes and optionally apply filters.


  • Particles (blob required; alignments2D and alignments3D optional)

Common Parameters

  • Crop / pad to box size (pix): Crop in real space to this size before downsampling. If this number is larger than the input size, each particle will be padded with its mean pixel value

  • Desired approx. pixel size (A): If set, the output particles will be downsampled to have approximately the specified pixel size. Only this parameter or Fourier crop to box size (pix) should be set, not both.

  • Fourier crop to box size (pix): If set, the output particles will be downsampled to have the specified number of pixels in each dimension. Only this parameter or Desired approx. pixel size (A) should be set, not both.

  • Lowpass resolution (A): Resolution of corner frequency in Angstroms

  • Recenter using aligned shifts: Activate this parameter to recenter the particles based on their shifts.

    • Particles will only be recentered if this is true, and alignments3D or alignments2D are connected.

    • Note that alignments2D will be used for recentering only if alignments3D are unavailable.

    • Job: Volume Alignment Tools may be used to change the geometric center of a particle stack prior to downsampling and recentering, for the purpose of subparticle cropping.

  • Save results in 16-bit floating point: Saves the output particles in float16 format (CryoSPARC v4.4+).


  • Downsampled Particles

Common Next Steps

  • Running 2D Classification, 3D Classification, 3D Variability Analysis or Heterogeneous Refinement on downsampled particles. To refine a structure using the full particle resolution, you can reference the original (non-downsampled) particle data using an input-group override as described in this tutorial.

Last updated