Re: [Jack-Devel] About ALSA drivers for Firewire sound devices

On Thu, Feb 06, 2014 at 01:19:46AM +0900, Takashi Sakamoto wrote:
>>> >With non-clock-synced devices these days, it is better to use JACK's own
>>> >capability for handling multiple devices. But since all the firewire
>>> >devices on a chain share the same clock, this seems right.
> >
>> While the firewire bus provides a synchronous bus clock (primarily used for
>> timing data transfers), very few firewire audio devices can make use of this
>> to derive an audio clock.  In a vast majority of cases, to achieve clock
>> sync two firewire interfaces on the same firewire bus still need to be
>> synced either via wordclock (best) or optical ports (the fallback in the
>> absence of wordclock I/O).
>
> My drivers don't support this type of synchronization: some devices in  
> the same IEEE1394 bus, drivers decide a sync master from the devices  
> then the others are sync slave, the drivers receive timestamp in packet  
> from master and transmit the timestamp in the packet to the others.
>
> And my drivers don't support this type of synchronization: some devices  
> in the same IEEE1394 bus, drivers connects between them to transfer  
> synchronization packets each other.
>
> The reasons are:
> 1.I have never seen public specifications which describe this type of  
> synchronizations.
> 2.I have never seen devices which support this type of synchronizations.

Some examples:

http://www.prismsound.com/music_recording/products_subs/orpheus/orpheus_home.php

http://www.presonus.com/products/FireStudio-Project

http://uk.focusrite.com/firewire-audio-interfaces/saffire-pro-40


John

1391627075.3752_0.ltw:2,a <20140205190250.GA18818 at localhost dot localdomain>