Thanks for your service, it is really good, I use it almost every day!

But sometimes I have a trouble recognizing sound signals (not sure if anybody mentioned it).

Probably not everybody has perfect ear to distinguish sound tones if there is a significant time interval between them. Could you please use more complicated sound samples so that it is easier to distinguish them? Or you could just combine the existing two (instead of "high" and "low" you could play "low-high" and "high-low"). Should be easy to implement.

Thanks again!