This is my question too.
I'm surprised this question is not asked more often.
With a linear freq scale you can really only see what's happening in the highest frequency regions.
Cute if you want to see the interference of the sampling frequency on the audio, but as a full range spectogram it's actually pretty useless.
But this question has been asked for more than 9 years now, but unfortunately without resulting in a logarithmic frequency scale...
Is it that hard to code?
Or is it the inevitable resolution loss in the low frequency regions that scared the attempts off?
BR, J.