Subject: RE: [linux-audio-user] HDSPmixer problem
From: Ivica Ico Bukvic (ico_AT_fuse.net)
Date: Wed Dec 15 2004 - 18:27:07 EET
For the record, I have the exact same problem and if I am not mistaken Tim
Blechman also does. Both Tim and myself had a problem with our PCMCIA
controllers (garbled audio) until some of the registers were changed which
enabled burst mode and some other things (different solutions for our two
controllers as they were manufactured by different companies). This fixed
the sound I/O problems and the card now works flawlessly in that respect,
but the VU meter has *exactly* the same behavior as described by Kevin
(FWIW, this does not happen in Win32 so hardware is apparently not the
issue, at least not in my case).
>From briefly studying the issue: (the following applies to the middle row of
meters, namely 1 and 2) it seems that the code for slowly degrading the VU
meter's values through time works accurately but perhaps slower than usual
(i.e. when there is no sound, the meter should degrade at a relatively fast
pace until it reaches -infinity), but apparently refreshes of the VU meter
values are not often enough and therefore the peaks are usually well below
the actual levels and they only occasionally show up at all. It's almost as
if the refresh of the meter values is slowed down.
Also, the input signal monitoring (top row) only shows info for the first 6
or 7 inputs, the rest are dead (there should be at least 8 of the analog
ones that work), while the output levels (bottom row) generally behave in
the same fashion as the middle row but with much lower values, meaning that
most of the time they show 0 activity.
All this leads me to believe that there is a bug somewhere in the mixer that
only affects certain hardware.
Also, just for the record my PCMCIA controller is ENE1410 on an AMD64 3000+
running 32-bit Linux on a heavily modified 2.6.7 kernel.
Best wishes,
Ivica Ico Bukvic, composer & multimedia sculptor
http://meowing.ccm.uc.edu/~ico/
> -----Original Message-----
> From: linux-audio-user-bounces_AT_music.columbia.edu [mailto:linux-audio-
> user-bounces_AT_music.columbia.edu] On Behalf Of Mark Knecht
> Sent: Wednesday, December 15, 2004 10:34 AM
> To: kevine_AT_esm.rochester.edu; A list for linux audio users
> Subject: Re: [linux-audio-user] HDSPmixer problem
>
> Hi,
> Family member with the HDSP 9652 using the same kernel but using
> Planet Alsa support. Possibly your RPMs might have something to do
> with this?
>
> To me this sounds like a firmware revision issue. Check with other
> MultiFace users as to which revision they are using.
>
> On Wed, 15 Dec 2004 03:54:06 -0500, Kevin Ernste <kevinernste_AT_gmail.com>
> wrote:
> > Hello -
> >
> > I just installed a new HDSP multiface with PCI on a PlanetCCRMA/FC3
> > box, kernel 2.6.8.1-1.520.2vS7.ll, ALSA 1.07 (homemade rpms) and am
> > not seeing any output levels on HDSPmixer (bottom row) despite
> > actually hearing the signal I expect (sounds great).
> >
> > Modules/firmware load beautifully, I can route signal with HDSPmixer
> > as expected, adjust levels, etc, just no visual feedback from the
> > outputs row either with vu levels or changing peak values.
>
> Seems a bit strange. Default operation on my copy is that the
> input/top row routes nowhere, so for input audio I have to select
> where you want it to go and bring the level up before I see anything
> in the bottom row. Note that if you choose a route but do not bring
> the level up on the top row then the next time you look it's not
> selected.
>
> Default operation for the internal audio is vertical routing.
> 1-internal goes to 1-output, 2-internal goes to 2-output, etc. Default
> levels are set correctly for this. This way if you choose a default
> app it routes to internal 1/2 and you see output 1/2 with the same
> value. Hook your output D/A there and you get sound. (Supposedly!) ;-)
> >
> > Also on mixer channels that do show vu's (input and internal, top and
> > middle rows respectively) the yellow peak indicators appear only
> > randomly every 2-3 seconds or so, sort or popping around without any
> > clear relation to the signal peak. Something's not right here.
>
> I see the yellow peak about 1/4" above the top of the green VU moving
> as the audio does.
>
> >
> > /sbin/lspci:
> >
> > 02:0d.0 Multimedia audio controller: Xilinx Corporation RME Hammerfall
> > DSP (rev 0b)
> >
> > cat /proc/pci:
> >
> > Bus 2, device 13, function 0:
> > Class 0401: PCI device 10ee:3fc5 (rev 11).
> > IRQ 5.
> > Master Capable. Latency=255.
> > Non-prefetchable 32 bit memory at 0xec800000 [0xec80ffff].
> >
> > "proc/interrupts says the card is on it's own interrupt, IRQ 5.
>
> Is it set non-threaded?
>
> >
> > Kevin
> >
>
> - Mark
>
> --
> No virus found in this incoming message.
> Checked by AVG Anti-Virus.
> Version: 7.0.296 / Virus Database: 265.5.4 - Release Date: 12/15/2004
>
-- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.296 / Virus Database: 265.5.4 - Release Date: 12/15/2004
This archive was generated by hypermail 2b28 : Wed Dec 15 2004 - 18:32:09 EET