Re: [alsa-devel] About 'SNDRV_PCM_CLASS_MULTI' for devices with multi-channels
On 02/19/2014 03:57 PM, Takashi Sakamoto wrote:
Hi all,
I have two questions about the usage of this macro.
Should I use this macro for new drivers which I'm working for?' http://mailman.alsa-project.org/pipermail/alsa-devel/2014-January/071820.htm...
My drivers support some firewire devices, which is desinged for studio-use and have many channels. Its combination of channels are not designed for surround sound like 5.1ch.
Just judging from its name, 'SNDRV_PCM_CLASS_MULTI', I should use this. But there are no drivers which use this macro in current tree, against my expectation that some drivers (ICE1712-pro, RME9652, 6fire usb and so on) should use this macro.
Are there any reasons?
I've never heard of it before. I had a quick look in alsa-lib and couldn't find any difference in handling between CLASS_GENERIC and CLASS_MULTI.
It's probably something historic that nobody uses any more. Maybe people who have been along longer knows what it is.
Hi David, (My message was not delivered by ALSA's mailman...)
Again all,
I have two questions about the usage of this macro.
Should I use this macro for new drivers which I'm working for?'
http://mailman.alsa-project.org/pipermail/alsa-devel/2014-January/071820.htm...
My drivers support some firewire devices, which is desinged for studio-use and have many channels. Its combination of channels are not designed for surround sound like 5.1ch.
Just judging from its name, 'SNDRV_PCM_CLASS_MULTI', I should use this. But there are no drivers which use this macro in current tree, against my expectation that some drivers (ICE1712-pro, RME9652, 6fire usb and so on) should use this macro.
Are there any reasons?
In current tree, there are no drivers which uses this macro.
As long as I investigated in ftp server, this macro is firstly added in ALSA 0.9.0beta1 (2001/02/21). So this macro has been ignored over 10 years...
Regards
Takashi Sakamoto o-takashi@sakamocchi.jp
At Thu, 20 Feb 2014 11:28:48 +0900, Takashi Sakamoto wrote:
Hi David, (My message was not delivered by ALSA's mailman...)
Because you posted from an address that isn't subscribed. It takes time until manually approved.
Again all,
I have two questions about the usage of this macro.
Should I use this macro for new drivers which I'm working for?'
http://mailman.alsa-project.org/pipermail/alsa-devel/2014-January/071820.htm...
My drivers support some firewire devices, which is desinged for studio-use and have many channels. Its combination of channels are not designed for surround sound like 5.1ch.
Just judging from its name, 'SNDRV_PCM_CLASS_MULTI', I should use this. But there are no drivers which use this macro in current tree, against my expectation that some drivers (ICE1712-pro, RME9652, 6fire usb and so on) should use this macro.
Are there any reasons?
In current tree, there are no drivers which uses this macro.
As long as I investigated in ftp server, this macro is firstly added in ALSA 0.9.0beta1 (2001/02/21). So this macro has been ignored over 10 years...
Right. And the likely reason is that such a differentiation doesn't help much. You can still use your device for driving 5.1 surround :)
So, I rather would like to keep them RIP. Maybe worth to remove it now.
Takashi
Right. And the likely reason is that such a differentiation doesn't help much.
I thought usage of this macro can tell applications that the PCM channels are not for surround sound.
But for this purpose, card configuration is better. Applications can get to know by seeking 'surround' PCM devices such like 'surround51'.
You can still use your device for driving 5.1 surround :)
It's not my intention. I'm considering about the way to prevent applications from using the devices for surround sound.
I don't prepare for card configuration. So devices which my drivers support have no 'surround' PCM devices.
So, I rather would like to keep them RIP. Maybe worth to remove it now.
Yes.
In the same reason, 'SND_PCM_SUBCLASS_MULTI_MIX' should be removed. As a result, 'SND_PCM_SUBCLASS_XXX', 'struct snd_pcm_info.dev_subclass' are also needless...
Thanks
Takashi Sakamoto o-takashi@sakamocchi.jp
participants (4)
-
David Henningsson -
Takashi Iwai -
Takashi Sakamoto -
Takashi Sakamoto