Hi,

let me describe again what I did.

I had an deterministic source for a signal (an fm sweep)
repping away every .4 seconds or so.

I plugged this source into the #1 input of my  midiman delta1010,
and

after
        installing the OSS drivers,
        running ossxmixer and setting the source for OUT#1 to IN#1,

I observed the In#1 and out#1 of the delta1010 on my scope.

This was done with an unpatched  mandrake 8.1 linux box ( kernel-2.4.8 ),
and a 400Mhz Pentium.

the I/O latency (respective delay between in and out) as observed on
the scope was about 1.4 msecs, which corresponds ~ 64 samples at with
fs=48Khz.

As I review this mailing  list I've come to understand that I should expect
latencies O(10's of milliseconds) with an unpatched kernel.
Instead I'm seeing a rock solid 1.4msec.

If these latencies are a fact of life, then the only way I can explain
this is that the ossxmixer program is setting up a register (software)
enabled hardware loopback mode where the ADC's output are fed directly
to input of the DACs.  Most codecs have this feature as a first
bootstrap capability to test the wiring of a new circuit.

How else do  I explain my observation ?

Thanks,
Ira

Ekhaus                          [EMAIL PROTECTED]
http://www.synapcorp.com
Synap Corporation               phone:781 641 3295
71 Glenburn Rd                  fax  :781 623 5533
Arlington, MA 02476

Reply via email to