Re: [Jack-Devel] Ethernet-based audio interface using (net)jack idea

On Wed, Aug 17, 2011 at 12:41:00PM +0200, Jeroen Van den Keybus wrote:
> What jitter performance would actually be required by pro-audio ?

Putting hard numbers on this is difficult because jitter specs given
by many converter and clock manufacturers are effectively meaningless
due to insufficient information about how the figure was derived.

There's an AES paper about this by Bruno Putzeys which should be required
reading for anyone designing converter clocks IMO. A copy is available
at the Grimm site:
http://www.grimmaudio.com/whitepapers
It's the "clock jitter spec.pdf".

What would be acceptable for consumer audio wouldn't necessarily be
good enough for pro audio. For example, many consumer grade interfaces
use the raw clock output from an S/PDIF receiver chip like the CS8414
as the converter clock source when syncing to an S/PDIF input. This is
audibly inferior to a using decent crystal oscillator. For a pro work
additional jitter reduction would be required to reduce the jitter to
an acceptable level in this example. 

The holy grail for designers is a converter that sounds uniformly
excellent whatever the clock sync source. Even some well known pro level
converters fail in this regard.

John

1313604284.2186_0.ltw:2,a <20110817180749.GA2550 at localhost0 dot localdomain>