Thank you. I'll definitely contact to NI. It took me more than 4 hours to compile the sample Voice Recorder and at the end it did the same thing as I can acquire without using FPGA. I don't have much time to learn the entire FPGA process and I haven't been able to find any comprehensive reference for myRIO to learn how to use the DAQ from simple projects to more complicated ones. I think my only option is to get direct help from NI app engineer and I hope it is possible to perform what I want using myRIO in an efficient way.
I would recommend to aquire the data directly from the FPGA using DMA Channels to transfer these data to the Host and then use the standard sound output Vis to playback your sound. Be aware of the sample frequency, which have to match with the aquire rate.
If you will still have some questions, please let me know.