Discrepancy between GUI and Commandline.mlx results

[weic liu]

Hi William and PIVlab users! 

I’m batch-processing a set of experiments under identical conditions using Commandline.mlx (with r{1,2}=r{2,2}=0.0451), but when I compare the output to the GUI results (with 1 px/frame = 0.0451 m/s) I see a clear order-of-magnitude difference (igure 1 and 2). 

I’ve tried adjusting the calibration index r{1,2} and r{2,2} to values like 1 or 0.5, yet the computed velocities never change. I’m unsure whether I’ve misunderstood how these parameters work or if there’s an issue in my modified script.

For reference, I’ve attached the .mlx file I’m using (it includes a few Chinese comments but should be self-explanatory) and .jpg files. Could you please advise on the correct role of these coefficients or point out anything I might be missing?  

Thanks,

LIU

[William]

Hi, the  numbers in r{1,2} and r{2,2} report the conversion factors to the validation algorithms. But they do not change the data. After validation, you need to multiply the result (u and v) with your calibration factors.