[alsa-devel] envy24control can't set sample rate on Delta 1010

[Takashi Iwai]

At Mon, 21 Apr 2008 21:24:30 +0100,
John Rigg wrote:
> 
> On Mon, Apr 21, 2008 at 05:58:56PM +0100, John Rigg wrote:
> > On Mon, Apr 21, 2008 at 05:54:14PM +0200, Takashi Iwai wrote:
> > > At Mon, 21 Apr 2008 01:42:06 +0100,
> > > John Rigg wrote:
> > > > 
> > > > envy24control in alsa-tools-1.0.16 cannot set sample rate
> > > > on the M-Audio Delta 1010. The Actual Rate window appears to
> > > > be functioning correctly, as is the "S/PDIF in" button, but
> > > > none of the internal clock sample rate buttons do anything.
> > > > Setting the sample rate as a jackd parameter works as normal.
> > > > 
> > > > I tried it with 2.6.20 and 2.6.25 kernels. Behaviour was
> > > > the same on both. I've made a bug report (0003915).
> > > 
> > > And which version of alsa-tools works regarding this?
> > > If we can narrow down the last working version, it'd be relatively
> > > easy to find the culprit.
> > 
> > Sorry for lack of info (I was testing this at 1am).
> > So far I've tested alsa-tools-1.0.13 to 1.0.16 and
> > all those exhibit this behaviour. I'm using 1.0.13
> > on my production system, but I hadn't noticed it because
> > I don't usually use envy24control to change sample rates.
> > It was brought to my attention by several users who have emailed
> > me about it. 
> > 
> > It's quite an awkward problem because once the
> > clock source gets set to S/PDIF it can no longer be reset to
> > internal clock without using alsactl or re-initialising the card(s)
> > by cold rebooting with alsa init.d scripts disabled.
> > 
> > The problem obviously appeared before 1.0.13. It might
> > be a few days before I get time to test earlier versions,
> > but I'll try to find the last working version.
> 
> After further tests, this doesn't actually look so serious.
> When jackd is running, the clock source can be
> changed to S/PDIF in or Word Clock in, but can not be changed back
> to internal clock again until jackd is stopped. This might be a minor
> inconvenience to some, although it could be argued that clock sources
> should not be changed while jackd is running. Once jackd is stopped the 
> internal clock can be selected correctly with the sample rate buttons.
> I was mistaken about having to re-initialise the card to restore normal
> behaviour (that's what happens when testing at 1am). Sorry about the noise.

OK, then this looks like normal behavior.
The clock of ice1712 cannot be changed during a PCM stream is opened.

Thanks for checking.

Takashi