🚀 go-pugleaf

      I am trying to convert a voice recording from 44100Hz, 16-bit, Stereo
      to Mono, 8khz, 8 bit . The original has no problem with hiss etc but the
      converted files have huge problems with hiss and noise problems.




That's hardly surprising. You've converted a recording with potentially a 96dB signal to noise ratio to one with at best 48dB - somewhat worse than a cassette recorder could manage, and cut the frequency response back to 4kHz max (actually slightly less than that after the anti-alias filter).

What did you expect to happen if you did this? What you got was pretty much what you deserve to get, I'd say. If you want less hiss, then leave it as a 16-bit signal, and if you actually want to hear what's being said, then 11k sampling is as low as you should go. If you need to convert for a specific purpose, then you need to find out what the real format you need actually is - some of them are optimised to reduce these problems somewhat, in a way that .wav files aren't.