Problems with HDSP Multiface+CardBus

Problems with HDSP Multiface+CardBus

[Jaakko Prättälä]

Encouraged by Jesse's recent success I tried again,
now with firmware rev 10. The module loads,
firmware is loaded but with
Hwdep ioctl error on card hw:0 : Input/output error.

Still, to my disappointment,
aplay -D hw:0,0 fails with
aplay: set_params:800: Access type not available
If the error checking for access type is commented out,
aplay fails on not being able to determine the sample format
(the next error check in aplay.c)
Why is this information not available?
Any ideas?

-- 
Jaakko Prättälä
Jaakko.Prattala@Helsinki.FI



-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[Jaroslav Kysela]

On Thu, 12 Jun 2003, Jaakko [iso-8859-15] Prättälä wrote:

> Encouraged by Jesse's recent success I tried again,
> now with firmware rev 10. The module loads,
> firmware is loaded but with
> Hwdep ioctl error on card hw:0 : Input/output error.
> 
> Still, to my disappointment,
> aplay -D hw:0,0 fails with

aplay -D plughw:0,0

						Jaroslav

-----
Jaroslav Kysela <perex@suse.cz>
Linux Kernel Sound Maintainer
ALSA Project, SuSE Labs



-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[Jesse Chappell]

Jaakko =?iso-8859-15?q?Pr=E4tt=E4l=E4?= wrote on Thu, 12-Jun-2003:

 > Encouraged by Jesse's recent success I tried again,
 > now with firmware rev 10. The module loads,
 > firmware is loaded but with
 > Hwdep ioctl error on card hw:0 : Input/output error.
 
If you're still getting the I/O error on the hdsploader then its
hard to know what state the thing is really in.

The current drivers only work for me if I boot to windows, (red
light goes out), reboot to linux (red light flashing until modules
load).  I see no errors of any kind if I do that.  Is this what you 
are doing?  I have rev 11, BTW.

jlc


-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[Jaakko Prättälä]

On Thursday 12 June 2003 18:14, Jesse Chappell wrote:
> If you're still getting the I/O error on the hdsploader then its
> hard to know what state the thing is really in.
>
> The current drivers only work for me if I boot to windows, (red
> light goes out), reboot to linux (red light flashing until modules
> load).  I see no errors of any kind if I do that.  Is this what you
> are doing?  I have rev 11, BTW.
>
> jlc

No. I use hdsploader. With rev 10 fw the led goes off, with 11 it
stays on (iirc, it's been a while since I tried 0x0b). After that I can
play stuff with aplay -D plughw:0,0, as suggested
by Jaroslav (thank you :-), but I get no sound. The files are read though,
judging by how aplay's running time correlates with file length.

Maybe this is just because I don't know how to use the
mixer... What I did was
 # /sbin/modprobe snd-hdsp line_outs_monitor=1; hdsploader
 # for i in $(seq 18); do amixer cset name=Chn,index=$i 32768; done;
 # amixer cset numid=5 26,26,32768
 # amixer cset numid=5 27,27,32768
 # aplay ...
Isn't this supposed to give me something through the
headphones?

As with Jochem's hdsp9652,
cat /proc/asound/card0/pcm0p/sub0/status
shows signs of life (state: RUNNING etc.)

-- 
Jaakko Prättälä
Jaakko.Prattala@Helsinki.FI



-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[Jesse Chappell]

Jaakko =?iso-8859-15?q?Pr=E4tt=E4l=E4?= wrote on Thu, 12-Jun-2003:

 > Maybe this is just because I don't know how to use the
 > mixer... What I did was
 >  # /sbin/modprobe snd-hdsp line_outs_monitor=1; hdsploader
 >  # for i in $(seq 18); do amixer cset name=Chn,index=$i 32768; done;
 >  # amixer cset numid=5 26,26,32768
 >  # amixer cset numid=5 27,27,32768
 >  # aplay ...
 > Isn't this supposed to give me something through the
 > headphones?
 
I didn't do the line_outs_monitor=1, but I did the amixer stuff
like that, and it worked.  But a word of warning if you do get sound going
eventually: 32768 for the LineOut(headphone) port is *way* too
hot for headphones plugged directly in.  I had to crank it down to 4000 
to get a sane maximum volume.  You don't want to damage your headphones
or your ears.....

jlc



-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[d]

Jesse Chappell wrote:
>  But a word of warning if you do get sound going
> eventually: 32768 for the LineOut(headphone) port is *way* too
> hot for headphones plugged directly in.  I had to crank it down to 4000 
> to get a sane maximum volume.  You don't want to damage your headphones
> or your ears.....

that's interresting. 32767 has always been fine for me (on headphones 
just like on line-outs), and all descriptions about the mixer-matrix say 
32767 means 0db. even sinewaves with 0db are loud but not distorted on 
my headphones...

btw: does anybody have a "complete" description of the possibilities for
amixer -c 1 cset numid=XYZ ???

thanks&regards
d13b




-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[David E. Storey]

[-- Attachment #1: Type: text/plain, Size: 493 bytes --]

On Thu, 2003-06-12 at 14:03, d wrote:
> that's interresting. 32767 has always been fine for me (on headphones 
> just like on line-outs), and all descriptions about the mixer-matrix say 
> 32767 means 0db. even sinewaves with 0db are loud but not distorted on 
> my headphones...

technically, it's unity gain at 32768 and +6db at 65535.

> btw: does anybody have a "complete" description of the possibilities for
> amixer -c 1 cset numid=XYZ ???

amixer -c 1 contents

=)

d!

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Problems with HDSP Multiface+CardBus

[Jesse Chappell]

 > On Thu, 2003-06-12 at 14:03, d wrote:
 > > that's interresting. 32767 has always been fine for me (on headphones 
 > > just like on line-outs), and all descriptions about the mixer-matrix say 
 > > 32767 means 0db. even sinewaves with 0db are loud but not distorted on 
 > > my headphones...
 > 
 > technically, it's unity gain at 32768 and +6db at 65535.
 
Different headphones have different sensitivities, and line level
can get pretty hot electrically.  But I was surprised at how far down
I had to set it though.  I wonder if I was getting the signal
from channels 1 & 2 through several paths summing on each other.
It is a bit confusing.

David S (or Thomas C), what are the chances the work you guys
have done (independently) on a UI hdsp mixer could be released?  I think 
I'll be needing some easier control.  I'll be tempted to write my own if
nothing shows up to work with soon ;)

jlc


-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[David E. Storey]

[-- Attachment #1: Type: text/plain, Size: 2088 bytes --]

> Different headphones have different sensitivities, and line level
> can get pretty hot electrically.  But I was surprised at how far down
> I had to set it though.  I wonder if I was getting the signal
> from channels 1 & 2 through several paths summing on each other.
> It is a bit confusing.

I believe that the Hammerfall series sums the signal source to a 36 bit
value and the last few least significant bits are truncated to bring it
back to 24 bits. I've been REALLY curious about software mixing and the
RME guys seem to know what they're talking about:

<http://www.rme-audio.de/english/techinfo/hdsp_tmhard.htm>

> David S (or Thomas C), what are the chances the work you guys
> have done (independently) on a UI hdsp mixer could be released?  I think 
> I'll be needing some easier control.  I'll be tempted to write my own if
> nothing shows up to work with soon ;)

I could release something now, BUT... it's still very much in
development and it's not technically correct. I'm waiting to hear back
from the TotalMix guys with some specifics on algorithms used to render
information back and forth from their UI and the hdsp. My communication
with them is somewhat latent as I'm going through a couple people, but I
do know that my questions have reached the right people. As a proof of
concept, it works. (I started on it a week ago and have been rabidly
working on it in my "spare" time since I got a fairly big gig coming up
with some people interested in going IEM by then.)

It's not really designed to be TotalMix for Linux per se, but rather a
digital IEM mixer. As such, I had slightly different requirements
although the basic ideas are the same. I just got mute and solo working
monday night which I'm really happy about. They're slightly more complex
than I had originally thought since one has to take into account channel
state, but I think what I have is fairly efficient. I also don't have
the playback channels mapping yet and I don't have any real particular
need to for what I'm doing other than for completeness.

d!

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Problems with HDSP Multiface+CardBus

[Jesse Chappell]

 > It's not really designed to be TotalMix for Linux per se, but rather a
 > digital IEM mixer. As such, I had slightly different requirements
 > although the basic ideas are the same. I just got mute and solo working
 > monday night which I'm really happy about. They're slightly more complex
 > than I had originally thought since one has to take into account channel
 > state, but I think what I have is fairly efficient. I also don't have
 > the playback channels mapping yet and I don't have any real particular
 > need to for what I'm doing other than for completeness.
 
I would be using it primarily for headphone mix(es) during tracking,
a similar goal.  However, I will need playback channels in there
too :)  I would be happy to contribute to your codebase.... no
need for wide release if you would rather wait (can discuss offlist).

What toolkit and language are you using?

jlc


-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5

Re: Problems with HDSP Multiface+CardBus

[Paul Davis]

>Maybe this is just because I don't know how to use the
>mixer... What I did was
> # /sbin/modprobe snd-hdsp line_outs_monitor=1; hdsploader
> # for i in $(seq 18); do amixer cset name=Chn,index=$i 32768; done;
> # amixer cset numid=5 26,26,32768
> # amixer cset numid=5 27,27,32768
> # aplay ...
>Isn't this supposed to give me something through the
>headphones?

use alsamixer.

as with all ALSA drivers, after loading, all channels are muted. if
there is no state to restore in asound.state or whatever its called,
they will stay muted. 

to make it "simple" to get the volumes up, the driver presents a
"simple mixer" interface that alsamixer and other similar programs can
use. it controls the gain applied to routing between a playback stream
(from the CPU) and an output (to a physical channel). you don't need
to know anything about the matrix mixer to use it, but it obviously
can't represent the full power of the matrix mixer either.

if you pass line_outs_monitor=1, then you should hear all inputs and
all outputs in the headphone outs. you don't need to set the mixer to
get this to work. i also enclose a very helpful post from a few months
back that we should really put into the wiki (below).

>As with Jochem's hdsp9652,
>cat /proc/asound/card0/pcm0p/sub0/status
>shows signs of life (state: RUNNING etc.)

please don't confuse your situation with the hdsp9652. there is a bug
in the code for the hdsp9652 that prevents any sound regardless of the
mixer settings.

-- post from roger williams --------------------------------------------

Some of the numbering is discontinuous (because the Multiface doesn't
have the Digiface's ADAT2 range of channels), so the full Multiface
I/O list is:

Inputs to HDSP Mixer
========================================================================
Multiface analogue inputs 1-8      = amixer source channels 0-7
Multiface ADAT inputs 1-8          = amixer source channels 16-23
Multiface SPDIF input              = amixer source channels 24-25
alsa_pcm:playback_1-26             = amixer source channels 26-51

Outputs from HDSP Mixer
========================================================================
Multiface analogue outputs 1-8     = amixer destination channels 0-7
Multiface ADAT outputs 1-8         = amixer destination channels 16-23
Multiface SPDIF output             = amixer destination channels 24-25
Multiface line (headphone) output  = amixer destination channels 26-27

Mapping between Multiface inputs and alsa_pcm:capture channels
========================================================================
Multiface analogue inputs 1-8      = capture_1-8
Multiface ADAT inputs 1-8          = capture_9-16
Multiface SPDIF input              = capture_17-18


-------------------------------------------------------
This SF.NET email is sponsored by: eBay
Great deals on office technology -- on eBay now! Click here:
http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5