Re: [Jack-Devel] netjack1 / netjack2

Chris Caudle wrote:
> On Mon, October 3, 2011 5:17 pm, Tobias Hoffmann wrote:
>   
>> My current issues are much more mundane, so here they go:
>> 1. It is my understanding, that netjack (as slave) is one of simplest
>> audio-over-ip protocols ...especially for fpga implementations
>>     
> Wouldn't the FPGA end have to be the master, because it has the audio
> hardware?  And the computer would be the slave, because it has no digital
> to analog converters, just the network connection to the FPGA. 
Not necessarily (I believe). On the FPGA end I have every option (PLL, VCXO, etc.) to generate a "locked" word clock for the A/D or D/A converters from the incoming network packets. 
I think that for bigger networks (WAN, etc.) the timing might well be close to unsalvageable, but I expect very small LANs to be rather well-behaved.

OTOH my plan is to start as simple as possible, and when that works (more or less) to decide on further steps, like better word clock generator, or the ability to act as netjack master -- which I think requires more of the protocol to be handled in software (i.e. embedded CPU).

  Tobias

1317694848.10413_0.ltw:2,a <4E8A6D52.5080509 at thax dot hardliners dot org>