log scale y-axis for periodogram

The periodograms from an electronic signal I am analyzing have such high power periodicity in the low frequency bandwidths that, without zooming in on the data, these are almost impossible to read. (see the example below) Is it possible, when using the [(pxx,f) = periodogram(x,window,nfft)], to adjust the y-axis to a log-scale to improve the graph's readibility? The y-axis in the example represents the magnitude of the periodogram (i.e. abs(pxx)).

Answer 1

[Mischa Kim]

Philippe, it is. Use

pxx = periodogram(x,window,nfft)
plot(10*log10(pxx))

to plot the periodogram.

[Philippe]

Dear Mischa,

Thanks for your answer. However, in the presented periodogram I've plotted the frequency vector f (output from [(pxx,f) = periodogram(x,window,nfft)]) against the magnitude of pxx (through calculating abs(pxx)). This wasn't clearly put in the original question, I've now adjusted it.

As you said, plot(10*log10(pxx)) gives a log-scale y-axis, but - and I'm probably overlooking something - until now I've not succeeded in converting my magnitude axis to a log-scale..

[Mischa Kim]

I am not quite sure I understand. For plotting frequency vs. pxx you would use

plot(f,10*log(pxx))

instead, correct. Is that what you are referring to? Or are you still not able to get the log-scale on the y-axis?