What is the difference between actin and loading control in normalization of western blot?

Regarding normalization of western blot analysis, I have found different strategies people follow but I couldn’t understand the principle and which one is the accurate,

1- some refer to loading control (which I assume Amido black or coomassie blue), then after quantification, they first normalize the values of the loading control and then divide each quantification value of the samples on the respective normalized loading control. this one strategy

2- Others, when they use ß-actin they divide the quantification value of the samples on the respective quantification value of ß-actin and then normalize the output

Would you please clarify the principle behind each strategy and which one should I follow, as I tried both and got different values!!! Kindly you will find the file attached

Thanks in advanceScreenshot 2020-12-05 at 19.36.04