Dear Jichang,

based on the attached paper, I would say it does not really depend on the software but on the way of the calibration of SEM for such measurements. All the Gwyddion does are just simple calculations according the x,y and z data. If your z axis is well calibrated on SEM, which seems to be the most difficult and time consuming part, you can use any software for roughness calculations (even MS Excel) if you get the formula right. The idea of surface roughness estimation from SEM seems interesting and may work if you get it just right, but frankly it's just more convenient, more accurate and mostly much quicker to use the SPM method such as AFM or stylus profilometry.

Good luck! Tomas.

The calibration method they used only can be applied to ideal/high controlled/well known surfaces. If you have an unknown sample, by SEM you won't be able to reach the minimun or higher values of height and won't be able to calibrate your sample.

Tomáš Halenkovič Thanks for your answer.Best wishes！

Luiz Fernando de Sousa Lima Thanks for your answer. Best wishes!

From my experience, SEM estimation of roughness is not strictly accurate but may be OK if you're just comparing samples. SEM is however not a metrological instrument and you should turn to dedicated instruments if you want a "correct" measurement.

The issue with the SEM is calibration, as was indicated above ; plus in-SEM 3D reconstruction will work better if you work on rough samples.

The choice of a profilometer/2D topograph/AFM will depend on your sample and the scale of your topography.

Pierre Caulet Thanks for your answer.Best wishes!

The choice of a profilometer/2D topograph/AFM will give surface roughness of the sample.

SEM gives morphology of the surface

Hemalatha Kandi : SEM is capable of roughness computation, if you use pair/tri based image 3D reconstruction. It's just not metrologically correct, except if you manage to calibrate it against standards.