Re: [ecasound] buffer logic

[Patrick S. <psh...@bo...>]

On Wed, September 25, 2013 12:43 pm, Kai Vehmanen wrote:
> Hi,
>
> On Tue, 24 Sep 2013, Patrick Shirkey wrote:
>
>> 1: Does it push the data immediately to the output when it gets
>> something
>> from the input with latency defined by the period size of the buffer
>> 2: I/O is callback based with an internal logic for transmitting data
>> based on internal timing un/related to the period size of the buffer?
>
> option 1
>
> Unless you add some effect/chainop in the ecasound chains that processes
> the data as it passes by, there is no added latency.
>
> You can see the logic with quite a bit of code documentation in
> ecasound/libecasound/audioio_jack_manager.cpp:eca_jack_process_engine_iteration()
>
>> i.e does ecasound have an internal buffer that can store buffered data
>> if
>> there is an underrun or does it work in a type of passthrough mode. What
>> comes in goes out, even if there is an underrun?
>
> Ecasound is deterministic in this sense and there is no internal
> buffering. In the very, very unlikely case that ecasound is late in
> providing data to JACK, jackd will kick ecasound out as it's not behaving.
>

Thanks. That's what I was expecting but just needed to confirm so I can go
return the issue to the PA devs.


--
Patrick Shirkey
Boost Hardware Ltd