X-ray speckle tracking based methods can provide results with best reported angular accuracy up to 2 nrad. However, duo to the multi-frame requirement for phase retrieval and the possible instability of the x-ray beam, mechanical and background vibration, the actual accuracy will inevitably be degraded by these time-dependent fluctuations. Therefore, not only spatial position, but also temporal features of the speckle patterns need to be considered in order to maintain the superiority of the speckle-based methods. In this paper, we propose a parallel acquisition method with advantages of real time and high accuracy, which has potential applicability to dynamic samples imaging as well as on-line beam monitoring. Through simulations, we demonstrate that the proposed method can reduce the phase error caused by the fluctuations to 1% at most compared with current speckle tracking methods. Meanwhile, it can keep the accuracy deterioration within 0.03 nrad, making the high theoretical accuracy a reality. Also, we find that waveforms of the incident beam have a little impact on the phase retrieved and will not influence the actual accuracy, which relaxes the requirements for speckle-based experiments.

基于 X 射线散斑跟踪的方法可以提供具有高达 2 nrad 的最佳报告角精度的结果。然而，由于相位检索的多帧要求以及 X 射线束、机械和背景振动的可能不稳定性，这些时间相关的波动不可避免地会降低实际精度。因此，不仅需要考虑空间位置，还需要考虑散斑图案的时间特征，以保持基于散斑的方法的优越性。在本文中，我们提出了一种具有实时和高精度优点的并行采集方法，该方法在动态样品成像和在线光束监测方面具有潜在的适用性。通过模拟，我们证明，与当前的散斑跟踪方法相比，所提出的方法最多可以将波动引起的相位误差降低到 1%。同时，它可以将精度劣化控制在0.03 nrad以内，使高理论精度成为现实。此外，我们发现入射光束的波形对检索到的相位影响很小，不会影响实际精度，这放宽了基于散斑实验的要求。