If we leave the connection quality and bandwidth between a caller and callee out of the question: Skype and Viber are closed proprietary networks, which use predefined codecs, sampling rates and bitrates, whereas SIP is an open standard. With SIP the call quality depends on which codecs the clients support and how the codec negotiation has been configured in the server and clients. Even though SIP supports anything from lofi to hifi, you will most likely get a "normal telephone call quality" sound from a randomly chosen SIP provider.