How to distinguish between Speech and Tone in RTP packet?

My requirement is to find out whether the RTP packet(received on my IP phone from a VOIP gateway) is containing Speech or some Tone. There is no field in RTP packet header to indicate this. I found RFC 4733 which mention 2 formats for RTP payload: Event payload and Tone payload. But how will I know that RTP payload is present in which format as there is no indication for this in RTP header. How can I decode the payload data until the payload format is not known to 开发者_Python百科me?