The audio balance could be better. Maybe you could record your voice separately to mix it in later.

There are so many variables there and some make a big difference. I assume you can save all the settings, which may not be possible with hardware.

I'm running Ubuntu Studio and had a look at what is installed. I have not really played with the synths, but I found this modular thing. It has some examples, but I would need to find a tutorial on making my own.