I'm reading the well known wavelets tutorial by Robi Polikar here. In part 3, about figure 3.7 and 3.8, it says "lower scales (higher frequencies) have better scale resolution (narrower in scale, which means that it is less ambiguous what the exact value of the scale) which correspond to poorer frequency resolution". I don't understand why it is. From my opinion, scales are just reciprocal to the frequencies, but if scales are less ambiguous then so are frequencies. So I think good scale resolution also implies good frequency resolution. However, I know the conclusion is wrong. So how can I understand this statement.
Thanks
update:
I guess I still cannot convince myself by checking the following plot from the tutorial. The tutorial says it should have good frequency resolution and poor time resolution at low frequencies (high scale) and good time resolution and poor frequency resolution at high frequencies (low scale). From the plot, I can accept that high-frequency part has good time solution because it seems to have clear separation between different frequencies near scale 1. But which way should I understand that the low frequency has good frequency resolution? It seems to me that the low frequency part covers a quite large area and I cannot distinguish which frequencies it is actually
