cs46xx experimental DSP drivers

cs46xx experimental DSP drivers

[Weijia Yang]

I am testing out the cs46xx experimental DSP drivers on the 9/22 cvs 
build on a TB Santa Cruz. The drivers work fine except for the fact that 
all sound is "flat"--almost no base at all. This is in comparison to the 
original cs46xx drivers that come with rc3. The sound quality difference 
is very noticeable. I have tried tweaking with all the mixer settings 
with no result. I have also tried the experimental DSP drivers on the 
rc3 build with the same results. Anyone know a fix? Thanks
-- 
Weijia Yang
School of Electrical and Computer Engineering
Georgia Institute of Technology
Atlanta, GA



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Benny Sjostrand]

> I am testing out the cs46xx experimental DSP drivers on the 9/22 cvs 
> build on a TB Santa Cruz. The drivers work fine except for the fact 
> that all sound is "flat"--almost no base at all. This is in comparison 
> to the original cs46xx drivers that come with rc3. The sound quality 
> difference is very noticeable. I have tried tweaking with all the 
> mixer settings with no result. I have also tried the experimental DSP 
> drivers on the rc3 build with the same results. Anyone know a fix? Thanks 


It's hard to say what the problem can be. The Santa Cruz card have been 
reported to work fine.
Maybe it's the soundcard amplifier possibly is turned off, have you
tried with "external_amp=1" ???

Can you do a "lspci -a" and a "lspci -a -n" send me the output, just to 
check
is your card possibly has a different vendor/ID.

/Benny





-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Weijia Yang]

Benny Sjostrand wrote:
>> I am testing out the cs46xx experimental DSP drivers on the 9/22 cvs 
>> build on a TB Santa Cruz. The drivers work fine except for the fact 
>> that all sound is "flat"--almost no base at all. This is in comparison 
>> to the original cs46xx drivers that come with rc3. The sound quality 
>> difference is very noticeable. I have tried tweaking with all the 
>> mixer settings with no result. I have also tried the experimental DSP 
>> drivers on the rc3 build with the same results. Anyone know a fix? Thanks 
> 
> 
> 
> It's hard to say what the problem can be. The Santa Cruz card have been 
> reported to work fine.
> Maybe it's the soundcard amplifier possibly is turned off, have you
> tried with "external_amp=1" ???
> 
> Can you do a "lspci -a" and a "lspci -a -n" send me the output, just to 
> check
> is your card possibly has a different vendor/ID.
> 
> /Benny
> 

Benny--
	Where exactly do I set this "external_amp" option? I see a 'External 
Amplifier Power Down' in amixer, but only option is muting, which turns 
sound off completely.  Also, several other people have also reported the 
"flat" sounds, as well as cpu usage spikes when initiating a sound 
stream. The posts can be seen at 
<http://forums.gentoo.org/viewtopic.php?t=10826&highlight=cs46xx>
and
<http://forums.gentoo.org/viewtopic.php?t=10810&highlight=cs46xx>

Here is the output you asked for:
lspci
00:00.0 Host bridge: Advanced Micro Devices [AMD] AMD-760 [IGD4-1P] 
System Contr
oller (rev 13)
00:01.0 PCI bridge: Advanced Micro Devices [AMD] AMD-760 [IGD4-1P] AGP 
Bridge
00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super South] 
(rev 40
)
00:07.1 IDE interface: VIA Technologies, Inc. VT82C586B PIPC Bus Master 
IDE (rev
  06)
00:07.2 USB Controller: VIA Technologies, Inc. USB (rev 16)
00:07.3 USB Controller: VIA Technologies, Inc. USB (rev 16)
00:07.4 SMBus: VIA Technologies, Inc. VT82C686 [Apollo Super ACPI] (rev 40)
00:0a.0 Multimedia audio controller: Cirrus Logic CS 4614/22/24 
[CrystalClear So
undFusion Audio Accelerator] (rev 01)
00:0d.0 Ethernet controller: 3Com Corporation 3c905C-TX/TX-M [Tornado] 
(rev 78)
00:10.0 Unknown mass storage controller: Promise Technology, Inc. 20265 
(rev 02)
01:05.0 VGA compatible controller: nVidia Corporation NV20 [GeForce3] 
(rev a3)


lspci -n
00:00.0 Class 0600: 1022:700e (rev 13)
00:01.0 Class 0604: 1022:700f
00:07.0 Class 0601: 1106:0686 (rev 40)
00:07.1 Class 0101: 1106:0571 (rev 06)
00:07.2 Class 0c03: 1106:3038 (rev 16)
00:07.3 Class 0c03: 1106:3038 (rev 16)
00:07.4 Class 0c05: 1106:3057 (rev 40)
00:0a.0 Class 0401: 1013:6003 (rev 01)
00:0d.0 Class 0200: 10b7:9200 (rev 78)
00:10.0 Class 0180: 105a:0d30 (rev 02)
01:05.0 Class 0300: 10de:0200 (rev a3)


-- 
Weijia Yang
School of Electrical and Computer Engineering
Georgia Institute of Technology
Atlanta, GA



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Benny Sjostrand]

> Benny--
>     Where exactly do I set this "external_amp" option? I see a 
> 'External Amplifier Power Down' in amixer, but only option is muting, 
> which turns sound off completely.  Also, several other people have 
> also reported the "flat" sounds, as well as cpu usage spikes when 
> initiating a sound stream. The posts can be seen at 
> <http://forums.gentoo.org/viewtopic.php?t=10826&highlight=cs46xx>
> and
> <http://forums.gentoo.org/viewtopic.php?t=10810&highlight=cs46xx>

Finally, so far I know the Santa Cruz is yet another card where we dont 
have  technical specifications. It's possible that there is any GPIO or 
whatever kind of logic that controls "something" that makes the sound 
better.
Not having a Santa Cruz card dont make things easier ... I had one I 
would start testing all the EGPIO's by the hard way seing what happens, 
(maybe
someone that have time and such a card can do this ?) In the current 
cs46xx driver there is some debug code to control the EGPIO's from the
ALSA mixer that can be activated by defining 
CONFIG_SND_CS46XX_DEBUG_GPIO in cs46xx_lib.c, that I've been using to 
test EGPIO functionality on
the Game Theater XP card.

It's possible that the current ALSA driver needs to be optimized. When a 
new PCM channel is opened first time the driver needs to do a lot a 
things. Sure that the you got a small and short  peak when audio 
playback starts. Can you can meusure if it's System time, User time ?

The "external_amp" parameter is specified when module is loaded, eg. on 
a line in /etc/module.conf:
option snd-cs46xx external_amp=1

/Benny



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Weijia Yang]

Benny Sjostrand wrote:>> Benny--
> 
> 
> Finally, so far I know the Santa Cruz is yet another card where we dont 
> have  technical specifications. It's possible that there is any GPIO or 
> whatever kind of logic that controls "something" that makes the sound 
> better.
> Not having a Santa Cruz card dont make things easier ... I had one I 
> would start testing all the EGPIO's by the hard way seing what happens, 
> (maybe
> someone that have time and such a card can do this ?) In the current 
> cs46xx driver there is some debug code to control the EGPIO's from the
> ALSA mixer that can be activated by defining 
> CONFIG_SND_CS46XX_DEBUG_GPIO in cs46xx_lib.c, that I've been using to 
> test EGPIO functionality on
> the Game Theater XP card.
> 
> It's possible that the current ALSA driver needs to be optimized. When a 
> new PCM channel is opened first time the driver needs to do a lot a 
> things. Sure that the you got a small and short  peak when audio 
> playback starts. Can you can meusure if it's System time, User time ?
> 
> The "external_amp" parameter is specified when module is loaded, eg. on 
> a line in /etc/module.conf:
> option snd-cs46xx external_amp=1
> 



/Benny

I have tested the modules option (the parameter is actually 
snd_external_amp) but no difference. I am attempting to use the debug 
code, but after defining I get the following errors during compile:

gcc -D__KERNEL__ -DMODULE=1 
-I/home/weijia/Downloads/alsa-cvs/alsa-driver/include 
-I/lib/modules/2.4.19/build/include -O2 -mpreferred-stack-boundary=2 
-march=i686 -malign-functions=4 -DLINUX -Wall -Wstrict-prototypes 
-fomit-frame-pointer -pipe -DALSA_BUILD -DCONFIG_SND_CS46XX_NEW_DSP=1 
-DKBUILD_BASENAME=cs46xx_lib   -c -o cs46xx_lib.o cs46xx_lib.c
In file included from cs46xx_lib.c:2:
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c: In function 
`snd_cs46xx_iec958_put':
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1615: warning: `res' might be 
used uninitialized in this function
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c: In function 
`snd_cs46xx_egpio_select_get':
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1767: structure has no member 
named `current_gpio'
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c: In function 
`snd_cs46xx_egpio_select_put':
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1776: structure has no member 
named `current_gpio'
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1777: structure has no member 
named `current_gpio'
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c: In function 
`snd_cs46xx_egpio_get':
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1791: structure has no member 
named `current_gpio'
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c: In function 
`snd_cs46xx_egpio_put':
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1806: structure has no member 
named `current_gpio'
./../alsa-kernel/pci/cs46xx/cs46xx_lib.c:1808: structure has no member 
named `current_gpio'
make[3]: *** [cs46xx_lib.o] Error 1
make[3]: Leaving directory 
`/home/weijia/Downloads/alsa-cvs/alsa-driver/pci/cs46xx'
make[2]: *** [_modsubdir_cs46xx] Error 2
make[2]: Leaving directory `/home/weijia/Downloads/alsa-cvs/alsa-driver/pci'
make[1]: *** [compile] Error 1
make[1]: Leaving directory `/home/weijia/Downloads/alsa-cvs/alsa-driver'
make: *** [build] Error 2


any ideas? Thanks

-- 
Weijia Yang
School of Electrical and Computer Engineering
Georgia Institute of Technology
Atlanta, GA



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Peter Heatwole]

On 2002.09.24 17:10 Weijia Yang wrote:
> I am attempting to use the debug code, but after defining I get the
> following errors during compile:
[snip]


    You probably just stuck a define in cs46xx_lib.c.  Append the 
following into alsa-driver/pci/cs46xx/Makefile, to the end of the
EXTRA_FLAGS line:

-DCONFIG_SND_CS46XX_DEBUG_GPIO=1


HTH,
-- Peter Heatwole
"Murphy was just a well known pessimist."



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Weijia Yang]

> 
>    You probably just stuck a define in cs46xx_lib.c.  Append the 
> following into alsa-driver/pci/cs46xx/Makefile, to the end of the
> EXTRA_FLAGS line:
> 
> -DCONFIG_SND_CS46XX_DEBUG_GPIO=1
> 
> 
> HTH,
> -- Peter Heatwole
> "Murphy was just a well known pessimist."
> 
> 

Thanks for the tip--everything compiles fine now.  Now, how exactly do I 
go about testing these EGPIO settings? I see several EGPIO options in 
alsamixer--do I just adjust these and find something that works? Thanks


-- 
Weijia Yang
School of Electrical and Computer Engineering
Georgia Institute of Technology
Atlanta, GA



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Peter Heatwole]

On 2002.09.24 18:45 Weijia Yang wrote:
> Thanks for the tip--everything compiles fine now.  Now, how exactly do I 
> go about testing these EGPIO settings? I see several EGPIO options in 
> alsamixer--do I just adjust these and find something that works? Thanks

    I suppose so. I just leave an MP3 playing in XMMS, and restart it after
making a change (though this isn't required for all changes; I'm just
cautious).
    I _think_ you modify each pin using the "EGPIO On/Off" to turn each pin
on and off, and select the different pins using "EGPIO select".
    When I have more time I will try to learn more about this, and may be
able to help more.

-- Peter Heatwole
"Murphy was just a well known pessimist."



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Benny Sjostrand]

>    I suppose so. I just leave an MP3 playing in XMMS, and restart it 
> after
> making a change (though this isn't required for all changes; I'm just
> cautious).
>    I _think_ you modify each pin using the "EGPIO On/Off" to turn each 
> pin
> on and off, and select the different pins using "EGPIO select".
>    When I have more time I will try to learn more about this, and may be
> able to help more.
>
And "EGPIO In/Out" controls to use pin as input or output,
"EGPIO CMOS/Open drain", control if the pin use active CMOS
levels, or open drain. If think "Open drain" is like open collector, the
pin is floating when it's active, it needs a pull-up resistor or 
something, well,
just copied the text from the datasheet.

Playing with the EGPIO is very experimental and the consequences
can be _any_, but hope that you already knows that -;)

/Benny




-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Peter Heatwole]

On 2002.09.24 14:16 Benny Sjostrand wrote:
> Finally, so far I know the Santa Cruz is yet another card where we dont 
> have  technical specifications. It's possible that there is any GPIO or 
> whatever kind of logic that controls "something" that makes the sound 
> better.
> Not having a Santa Cruz card dont make things easier ... I had one I 
> would start testing all the EGPIO's by the hard way seing what happens, 
> (maybe
> someone that have time and such a card can do this ?) In the current 
> cs46xx driver there is some debug code to control the EGPIO's from the
> ALSA mixer that can be activated by defining 
> CONFIG_SND_CS46XX_DEBUG_GPIO in cs46xx_lib.c, that I've been using to 
> test EGPIO functionality on
> the Game Theater XP card.


    I've tried contacting Turtle Beach regarding this, and they wouldn't
release any information. I tried contacting VideoLogic, but they said they
could release any of the info on the Sonic Fury, since it's owned by
Turtle Beach.
    I had noticed earlier that you mentioned some code to control the EGPIO
pins, but when you didn't talk about it further, I incorrectly assumed it
was some code for personal use. Good to hear otherwise. =o)   I don't have
alot of time to spare, and I'm not very skilled in this area, but I'll
definitely give this a shot. Any advice that might make my attempt easier
or more profitable?


-- Peter Heatwole
"Murphy was just a well known pessimist."



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Peter Heatwole]

On 2002.09.24 14:16 Benny Sjostrand wrote:


    Booyah!   Just got digital out working properly on my Santa Cruz, and 
it
sounds great. I just needed to unmute "EGPIO In" and "IEC 958 Output".

Thanks again Benny,
-- Peter Heatwole
"Murphy was just a well known pessimist."



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: cs46xx experimental DSP drivers

[Peter Heatwole]

On 2002.09.24 18:08 Peter Heatwole wrote:
> 
> 
>    Booyah!   Just got digital out working properly on my Santa Cruz, and 
> it
> sounds great. I just needed to unmute "EGPIO In" and "IEC 958 Output".


    My apologies, I was rushed for a meeting and had to fire that e-mail 
off
in a hurry. (Not to mention I was elated to have digital output)
    To be more precise, here is the information I've collected.

    I've double checked, and from a fresh module insertion, I need to 
unmute
"EGPIO Input/Output" (therby setting bit one in BA0_EGPIODR), and unmute
"IEC 958 Output". Then coaxial digital out functions properly on my TB 
Santa
Cruz.
    Benny: what needs to be done so that the proper BA0_EGPIODR bits are 
set
to input or output when the user uses the "IEC 958 Input" and "IEC 958
Output" sliders? In other words, how could using the "IEC 958 Input" and
"IEC 958 Output" sliders properly set the BA0_EGPIODR bits? Also, can the
Game Theater have both optical in/out, and coaxial in/out simultaneously?

    On a side note, I noticed that the chip retains certain register 
values.
When I re-tested my instructions, I noticed that "EGPIO Input/Output" would
retain it's value after reloading the snd-cs46xx module. Just a heads up in
case you hadn't noticed this. It confused me for a minute, since I couldn't
understand why simply unmuting "IEC 958 Output" worked now, whereas earlier
it did not.

I hope all that makes sense. =o)

-- Peter Heatwole
"Murphy was just a well known pessimist."


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf