[LAU] [Fwd: [Alsa-devel-old] RME madiface]

[Reuben Martin]

Back on Wednesday 30 July 2008, Paul Davis was like:
> forwarded to LAU where you're a bit more likely to find help (lower
> signal-to-noise ratio).
>
> --p

I have interest in anything hdspm related because I built a system for a 
friend using an AES32...

I forwarded this to Florian Faber and this was his reply:

Reuben,

> Thread on the LAU list. More people having trouble locking sync with
> the hdspm driver.

Thanks. I don't know the list. The source of the problem is that the 
MADIface runs with 125MHz internally and the old driver doesn't know 
about this, so the wordclock the MADIface uses is about 17% too  slow.

If people want to patch the old driver themselves, it's just one line:

Go into hdspm.c, to line 845:

        n = 110100480000000ULL;    /* Value checked for AES32 and MADI*/

And change it to

          n = 131072000000000ULL;  // 125 MHz


Flo