I'm working on calculating the time-frequency content of EEG signals in MATLAB (STFT).
The power values of my signals vary, which in turn affects the range of the color bar in my plots. (Consequently, the same color (e.g., yellow) doesn't represent the same power value across different plots, complicating comparative analysis.).
Here's the MATLAB code snippet I'm using :
pspectrum(EEG(k,:),eeg_time,'spectrogram','TimeResolution',1.5);
To improve comparison between plots, I was considering normalizing the power within each window to a range of (-1,1). However, I'm unsure how to implement this normalization in MATLAB, as I couldn't find an option for it in the documentation.
Additionally, I'm thinking about whether it would be more logical to normalize the power of each window individually rather than normalizing based on the entire time range of power values.
Would this approach make sense?
Could anyone advise on how to normalize the power values of time-frequency plots in MATLAB or share their thoughts on the normalization scope (individual windows vs. the entire time range)?
P.S.: Attempting to unify the color bar ranges across all plots proved ineffective. Plots with significantly lower values became indiscernible, suggesting this method may not suit my needs.