Volume Peak display

Volume Peak display

[Manuel Jander]

Hello,

I'm implementing Hardware EQ peak visualization code in the Aureal
Vortex driver, and took as a reference the ICE1712 MultiTrack peak
kcontrol code.

The problems are:

- The peaks doesnt get updated by alsamixer nor alsamixergui.
I set the "volatile" flag, but is doesnt have any effect.
- Alsamixer shows only 4 of the peak control set (20 total), and other
mixers just show 2 of them.
- The controls are declared as read only, but i can change their value.
This shouldn't be allowed.

Any suggestions ?

Best Regards

Manuel Jander.


static int snd_vortex_peaks_info(snd_kcontrol_t *kcontrol,
snd_ctl_elem_info_t *uinfo) {
	uinfo->type = SNDRV_CTL_ELEM_TYPE_INTEGER;
	uinfo->count = 20;
	uinfo->value.integer.min = 0x0000;
	uinfo->value.integer.max = 0xffff;
	return 0;
}

static int snd_vortex_peaks_get(snd_kcontrol_t *kcontrol,
snd_ctl_elem_value_t *ucontrol) {
	vortex_t *vortex = snd_kcontrol_chip(kcontrol);
	int i, count;
	u16 peaks[20];
	
	vortex_Eqlzr_GetAllPeaks(vortex, peaks, &count);
	printk("vortex: Peak = %x\n", peaks[5]);
	for (i=0; i<20; i++)
		ucontrol->value.integer.value[i] = peaks[i];
	
	return 0;
}

static snd_kcontrol_new_t vortex_levels_kcontrol __devinitdata = {
	.iface = SNDRV_CTL_ELEM_IFACE_MIXER,
	.name = "EQ Peaks",
	.access = SNDRV_CTL_ELEM_ACCESS_READ | SNDRV_CTL_ELEM_ACCESS_VOLATILE,
	.info = snd_vortex_peaks_info,
	.get = snd_vortex_peaks_get,
};





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

Re: Volume Peak display

[Takashi Iwai]

At Sat, 20 Sep 2003 01:11:12 -0400,
Manuel Jander wrote:
> 
> Hello,
> 
> I'm implementing Hardware EQ peak visualization code in the Aureal
> Vortex driver, and took as a reference the ICE1712 MultiTrack peak
> kcontrol code.
> 
> The problems are:
> 
> - The peaks doesnt get updated by alsamixer nor alsamixergui.
> I set the "volatile" flag, but is doesnt have any effect.

this is inteintional.
(otherwise it will be too heavy for terminals over a slow network
 connection :)

for these purposes, we'll need anyway a special mixer app like
envy24control.

> - Alsamixer shows only 4 of the peak control set (20 total), and other
> mixers just show 2 of them.

maybe the result of abstraction by simple-mixer API.

> - The controls are declared as read only, but i can change their value.
> This shouldn't be allowed.

really?  then it must be fixed.
could you elaborate how this can happen?


Takashi


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

Re: Volume Peak display

[Takashi Iwai]

At Wed, 24 Sep 2003 10:33:43 -0400,
Manuel Jander wrote:
> 
> Hi Takashi,
> 
> I was thinking that wouldn't it be a good idea to have a "visualization"
> category for alsa controls ? Or at least some naming scheme that allows
> to identify controls that should just show some data, like VU meters, or
> the hardware EQ peak i'm implementing for the aureal vortex.

well, the volatile flag should be used for such a purpose.
if it's set, the control change won't be notified to the application.
the application which shows these controls must re-read the data
periodically (not passively like others).

the other flag name might be more obvious, though.

> 
> > > 
> > > - The peaks doesnt get updated by alsamixer nor alsamixergui.
> > > I set the "volatile" flag, but is doesnt have any effect.
> > 
> > this is inteintional.
> > (otherwise it will be too heavy for terminals over a slow network
> >  connection :)
> > 
> > for these purposes, we'll need anyway a special mixer app like
> > envy24control.
> 
> Ok. I got it. I'll start writing our "Vortex turbo  mixer" as soon as i
> can :-)
> 
> > > - Alsamixer shows only 4 of the peak control set (20 total), and other
> > > mixers just show 2 of them.
> > 
> > maybe the result of abstraction by simple-mixer API.
> > 
> > > - The controls are declared as read only, but i can change their value.
> > > This shouldn't be allowed.
> > 
> > really?  then it must be fixed.
> > could you elaborate how this can happen?
> 
> This could be a problem of the mixer app. I mean that i can modify the
> sliders on alsamixer or alsamixergui. The alsalib doesnt try to invoke a
> "control_put()" function (which is correct), since that would yield a
> segfault (i didnt implemented the control_put(), because that would be
> nonsense). 
>
> What i wanted to say, is that it shouldn't be possible in a mixer app to
> "Move" sliders from readonly controls. Their appearance should be
> different too. They should look like "progress bars" or something else.
> Or better than that, VU meters should look like VU meters :)

agreed.  at least, showing the elements with volatile in alsamixer
doens't make sense unless alsamixer shows vu meter (hey, it's not bad
:)


Takashi


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

Re: Volume Peak display

[Raymond]

Takashi Iwai wrote:
> At Wed, 24 Sep 2003 10:33:43 -0400,
> Manuel Jander wrote:
> 
>>Hi Takashi,
>>
>>I was thinking that wouldn't it be a good idea to have a "visualization"
>>category for alsa controls ? Or at least some naming scheme that allows
>>to identify controls that should just show some data, like VU meters, or
>>the hardware EQ peak i'm implementing for the aureal vortex.
> 

Are those VU meters measure in db or percentage ?

> 
> well, the volatile flag should be used for such a purpose.
> if it's set, the control change won't be notified to the application.
> the application which shows these controls must re-read the data
> periodically (not passively like others).
> 
> the other flag name might be more obvious, though.
> 
> 
>>>>- The peaks doesnt get updated by alsamixer nor alsamixergui.
>>>>I set the "volatile" flag, but is doesnt have any effect.
>>>
>>>this is inteintional.
>>>(otherwise it will be too heavy for terminals over a slow network
>>> connection :)
>>>
>>>for these purposes, we'll need anyway a special mixer app like
>>>envy24control.
>>
>>Ok. I got it. I'll start writing our "Vortex turbo  mixer" as soon as i
>>can :-)
>>

10-bands Graphic equalizer has been implemented in vortexcontrol.

>>
>>>>- Alsamixer shows only 4 of the peak control set (20 total), and other
>>>>mixers just show 2 of them.
>>>
>>>maybe the result of abstraction by simple-mixer API.
>>>
>>>
>>>>- The controls are declared as read only, but i can change their value.
>>>>This shouldn't be allowed.
>>>
>>>really?  then it must be fixed.
>>>could you elaborate how this can happen?
>>

>>This could be a problem of the mixer app. I mean that i can modify the
>>sliders on alsamixer or alsamixergui. The alsalib doesnt try to invoke a
>>"control_put()" function (which is correct), since that would yield a
>>segfault (i didnt implemented the control_put(), because that would be
>>nonsense). 

The message return by amixer seem to be misleading.


# amixer cget numid=12
numid=12,iface=MIXER,name='EQ Peaks'
   ; type=INTEGER,access=r----,values=20,min=0,max=32767,step=0
   : values=0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0

# amixer sget numid=12 1,1,1,1,1,1,1,1,1
amixer: Unable to find simple control 'numid=12',0



static snd_kcontrol_new_t vortex_levels_kcontrol __devinitdata = {
	.iface = SNDRV_CTL_ELEM_IFACE_MIXER,
	.name = "EQ Peaks",
	.access = SNDRV_CTL_ELEM_ACCESS_READ | SNDRV_CTL_ELEM_ACCESS_VOLATILE,
	.info = snd_vortex_peaks_info,
	.get = snd_vortex_peaks_get,
};

Refer to Alien's png

http://article.gmane.org/gmane.linux.alsa.devel/23112

There are 4 sliders marked as "EQ peaks" in alsamixer (1.0.9a) , the
first two are stereo and the last two are mono.

The 4 EQ peaks sliders in alsamixer can be adjusted and it is strange
that alsamixer do not get "segfault" when ".put" has not been
implemented in au88x0.


The following error only appear with compile with option --with-debug=full

Vortex: init.... <3>ALSA control.c:325: control 3:0:0:Playback PCM
advanced processing:0 is already present
done.

It seem that this HRTF kcontrols with 6 values cannot be created.

Does it mean that kcontrols ( not enumerated ) cannot contains more than
2 values ?

>>
>>What i wanted to say, is that it shouldn't be possible in a mixer app to
>>"Move" sliders from readonly controls. Their appearance should be
>>different too. They should look like "progress bars" or something else.
>>Or better than that, VU meters should look like VU meters :)
> 
> 
> agreed.  at least, showing the elements with volatile in alsamixer
> doens't make sense unless alsamixer shows vu meter (hey, it's not bad
> :)
> 
> 


Is it possible to standardise the API of multi-band equalizer ( hardware 
equalizer / graphic equalizer )

Beside Turtle Beach Montego II (au8830) , it seem that other cards such 
as Santa Cruz (cs46xx) , Montego DDL  also have hardware equalizer.







-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

Re: Re: Volume Peak display

[Takashi Iwai]

At Tue, 21 Jun 2005 17:08:41 +0800,
Raymond wrote:
> 
> >>
> >>>>- Alsamixer shows only 4 of the peak control set (20 total), and other
> >>>>mixers just show 2 of them.
> >>>
> >>>maybe the result of abstraction by simple-mixer API.
> >>>
> >>>
> >>>>- The controls are declared as read only, but i can change their value.
> >>>>This shouldn't be allowed.
> >>>
> >>>really?  then it must be fixed.
> >>>could you elaborate how this can happen?
> >>
> 
> >>This could be a problem of the mixer app. I mean that i can modify the
> >>sliders on alsamixer or alsamixergui. The alsalib doesnt try to invoke a
> >>"control_put()" function (which is correct), since that would yield a
> >>segfault (i didnt implemented the control_put(), because that would be
> >>nonsense). 
> 
> The message return by amixer seem to be misleading.
> 
> 
> # amixer cget numid=12
> numid=12,iface=MIXER,name='EQ Peaks'
>    ; type=INTEGER,access=r----,values=20,min=0,max=32767,step=0
>    : values=0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
> 
> # amixer sget numid=12 1,1,1,1,1,1,1,1,1
> amixer: Unable to find simple control 'numid=12',0

sget is for the simple controls.  There is no numid for simple
controls.


> static snd_kcontrol_new_t vortex_levels_kcontrol __devinitdata = {
> 	.iface = SNDRV_CTL_ELEM_IFACE_MIXER,
> 	.name = "EQ Peaks",
> 	.access = SNDRV_CTL_ELEM_ACCESS_READ | SNDRV_CTL_ELEM_ACCESS_VOLATILE,
> 	.info = snd_vortex_peaks_info,
> 	.get = snd_vortex_peaks_get,
> };
> 
> Refer to Alien's png
> 
> http://article.gmane.org/gmane.linux.alsa.devel/23112
> 
> There are 4 sliders marked as "EQ peaks" in alsamixer (1.0.9a) , the
> first two are stereo and the last two are mono.
> 
> The 4 EQ peaks sliders in alsamixer can be adjusted and it is strange
> that alsamixer do not get "segfault" when ".put" has not been
> implemented in au88x0.

It shouldn't segfault but be simply ignored.

> The following error only appear with compile with option --with-debug=full
> 
> Vortex: init.... <3>ALSA control.c:325: control 3:0:0:Playback PCM
> advanced processing:0 is already present
> done.
> 
> It seem that this HRTF kcontrols with 6 values cannot be created.
> 
> Does it mean that kcontrols ( not enumerated ) cannot contains more than
> 2 values ?

They can.  The above split of channels is just a matter of mixer
abstraction layer and alsamixer.  In the control API level, everyhing
is fine.


Takashi


-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

Re: Volume Peak display

[Raymond]

Takashi Iwai wrote:
> At Tue, 21 Jun 2005 17:08:41 +0800,
> Raymond wrote:
> 
 >>static snd_kcontrol_new_t vortex_levels_kcontrol __devinitdata = {
 >>	.iface = SNDRV_CTL_ELEM_IFACE_MIXER,
 >>	.name = "EQ Peaks",
 >>	.access = SNDRV_CTL_ELEM_ACCESS_READ | SNDRV_CTL_ELEM_ACCESS_VOLATILE,
 >>	.info = snd_vortex_peaks_info,
 >>	.get = snd_vortex_peaks_get,
 >>};
 >>
 >>Refer to Alien's png
 >>
 >>http://article.gmane.org/gmane.linux.alsa.devel/23112
 >>
 >>There are 4 sliders marked as "EQ peaks" in alsamixer (1.0.9a) , the
 >>first two are stereo and the last two are mono.
 >>
 >>The 4 EQ peaks sliders in alsamixer can be adjusted and it is strange
 >>that alsamixer do not get "segfault" when ".put" has not been
 >>implemented in au88x0.
 >
 >
 > It shouldn't segfault but be simply ignored.
 >


# amixer cget numid=12 numid=12,iface=MIXER,name='EQ Peaks'
   ; type=INTEGER,access=r----,values=20,min=0,max=32767,step=0
   : values=1,1,0,1,0,0,0,1,0,0,1,1,0,0,0,1,0,1,0,0

# amixer cset numid=12 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19
amixer: Control default element write error: Operation not permitted

> 
>>The following error only appear with compile with option --with-debug=full
>>
>>Vortex: init.... <3>ALSA control.c:325: control 3:0:0:Playback PCM
>>advanced processing:0 is already present
>>done.
>>
>>It seem that this HRTF kcontrols with 6 values cannot be created.
>>
>>Does it mean that kcontrols ( not enumerated ) cannot contains more than
>>2 values ?
> 
> 
> They can.  The above split of channels is just a matter of mixer
> abstraction layer and alsamixer.  In the control API level, everyhing
> is fine.
> 

The first HRTF kcontrol fail in snd_ctl_add() (there are 16 HRTF 
controls and each contain 6 values).

Is the error cause by numid or duplicate name ?


int snd_ctl_add(snd_card_t * card, snd_kcontrol_t * kcontrol)
{
	snd_ctl_elem_id_t id;
	unsigned int idx;

	snd_runtime_check(card != NULL && kcontrol != NULL, return -EINVAL);
	snd_assert(kcontrol->info != NULL, return -EINVAL);
	id = kcontrol->id;
	down_write(&card->controls_rwsem);
	if (snd_ctl_find_id(card, &id)) {
		up_write(&card->controls_rwsem);
		snd_ctl_free_one(kcontrol);
		snd_printd(KERN_ERR "control %i:%i:%i:%s:%i is already present\n",
					id.iface,
					id.device,
					id.subdevice,
					id.name,
					id.index);
		return -EBUSY;
	}









-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

Re: Re: Volume Peak display

[Takashi Iwai]

At Wed, 22 Jun 2005 22:21:13 +0800,
Raymond wrote:
> 
> Takashi Iwai wrote:
> > At Tue, 21 Jun 2005 17:08:41 +0800,
> > Raymond wrote:
> > 
>  >>static snd_kcontrol_new_t vortex_levels_kcontrol __devinitdata = {
>  >>	.iface = SNDRV_CTL_ELEM_IFACE_MIXER,
>  >>	.name = "EQ Peaks",
>  >>	.access = SNDRV_CTL_ELEM_ACCESS_READ | SNDRV_CTL_ELEM_ACCESS_VOLATILE,
>  >>	.info = snd_vortex_peaks_info,
>  >>	.get = snd_vortex_peaks_get,
>  >>};
>  >>
>  >>Refer to Alien's png
>  >>
>  >>http://article.gmane.org/gmane.linux.alsa.devel/23112
>  >>
>  >>There are 4 sliders marked as "EQ peaks" in alsamixer (1.0.9a) , the
>  >>first two are stereo and the last two are mono.
>  >>
>  >>The 4 EQ peaks sliders in alsamixer can be adjusted and it is strange
>  >>that alsamixer do not get "segfault" when ".put" has not been
>  >>implemented in au88x0.
>  >
>  >
>  > It shouldn't segfault but be simply ignored.
>  >
> 
> 
> # amixer cget numid=12 numid=12,iface=MIXER,name='EQ Peaks'
>    ; type=INTEGER,access=r----,values=20,min=0,max=32767,step=0
>    : values=1,1,0,1,0,0,0,1,0,0,1,1,0,0,0,1,0,1,0,0
> 
> # amixer cset numid=12 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19
> amixer: Control default element write error: Operation not permitted

So it's even better than ignoring :)

> > 
> >>The following error only appear with compile with option --with-debug=full
> >>
> >>Vortex: init.... <3>ALSA control.c:325: control 3:0:0:Playback PCM
> >>advanced processing:0 is already present
> >>done.
> >>
> >>It seem that this HRTF kcontrols with 6 values cannot be created.
> >>
> >>Does it mean that kcontrols ( not enumerated ) cannot contains more than
> >>2 values ?
> > 
> > 
> > They can.  The above split of channels is just a matter of mixer
> > abstraction layer and alsamixer.  In the control API level, everyhing
> > is fine.
> > 
> 
> The first HRTF kcontrol fail in snd_ctl_add() (there are 16 HRTF 
> controls and each contain 6 values).
> 
> Is the error cause by numid or duplicate name ?

Yes.  (More exactly, this is not numid but the name.  The numid is
assigned first after the addition to the card.)
If you want to use the same name string, then give different "index"
values to distinct with each other.

Instead of creating each control with a diffeent index value, you can
create a single control with multiple instances by passing a positive
kcontrol_new_t.count value.  This will create only a single control
but it's handled as if multiple control instances from outside.  The
get/put/info callbacks should check the given instance offset via
snd_ctl_get_ioff() (defined in control.h).  This is used in emu10k1,
trident, etc.


Takashi


-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click