[Ekiga-list] Alsa device enumeration
There are some problems w the current code which enumerates the alsa devices. I have a sketch for a patch, this is really not that complicated these days, there are new library routines which support it. The problem is just what devices to show, and how. Basically, there is two kind of devices: the classic devices which we partly see today, and several "new" devices such as pulse, hdmi, and several surround*. They exist for both capture and playback. The list for my box is below. The basic ideas are stolen from Skype. The important thing is the left column, the right is just internal stuff (hey, use a fixed font to see this...). Any thoughts out there about this list? Cheers! --alec Capture devices: null null default default pulse pulse AD198x Analog - hw:0,0hw:0,0 AD198x Analog - plughw:0,0plughw:0,0 AD198x Digital - hw:0,1 hw:0,1 AD198x Digital - plughw:0,1 plughw:0,1 Playback devices: iec958 - NVidia,0 iec958 null null default default pulse pulse AD198x Analog - hw:0,0hw:0,0 AD198x Analog - plughw:0,0plughw:0,0 AD198x Digital - hw:0,1 hw:0,1 AD198x Digital - plughw:0,1 plughw:0,1 surround40 - NVidia,0 surround40 surround41 - NVidia,0 surround41 surround50 - NVidia,0 surround50 surround51 - NVidia,0 surround51 surround71 - NVidia,0 surround71 front - NVidia,0 front ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
[Ekiga-list] Alsa device enumeration
OK. That was really hard to read. Try ftp://mumin.dnsalias.net/pub/alsatest.log instead. --alec ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
Re: [Ekiga-list] Alsa device enumeration
Le mercredi 11 mars 2009 à 20:46 +0100, Alec Leamas a écrit : > There are some problems w the current code which enumerates the alsa > devices. I have a sketch for a patch, this is really not that > complicated these days, there are new library routines which support it. > > The problem is just what devices to show, and how. Basically, there is > two kind of devices: the classic devices which we partly see today, and > several "new" devices such as pulse, hdmi, and several surround*. They > exist for both capture and playback. The list for my box is below. The > basic ideas are stolen from Skype. The important thing is the left > column, the right is just internal stuff (hey, use a fixed font to see > this...). > > Any thoughts out there about this list? > I would clean it and remove entries with hw only. Because that's complex. What is Surround40 capture ? -- _ Damien Sandras (o- //\Ekiga Softphone : http://www.ekiga.org/ v_/_ Be IP : http://www.beip.be/ FOSDEM : http://www.fosdem.org/ SIP Phone : sip:dsand...@ekiga.net ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
Re: [Ekiga-list] Alsa device enumeration
Damien Sandras wrote: Le mercredi 11 mars 2009 à 20:46 +0100, Alec Leamas a écrit : There are some problems w the current code which enumerates the alsa devices. I have a sketch for a patch, this is really not that complicated these days, there are new library routines which support it. The problem is just what devices to show, and how. Basically, there is two kind of devices: the classic devices which we partly see today, and several "new" devices such as pulse, hdmi, and several surround*. They exist for both capture and playback. The list for my box is below. The basic ideas are stolen from Skype. The important thing is the left column, the right is just internal stuff (hey, use a fixed font to see this...). Any thoughts out there about this list? I would clean it and remove entries with hw only. Because that's complex. What is Surround40 capture ? Surround capture is in my eyes complete nonsense. It's reported, though. I totally agree that it should not be visible to users. The hw discussion is complex. Actually, there are documents on the alsa site which boils down to that using hw:/plughw: is deprecated, they recommend "front", "surround*" etc. But for capture, it just makes no sense. And, more important, if you have more than one card there is no way to e. g., direct sound to "device front at the usb headset". Which device "front" actually refers to is a part of fthe alsa configuration (normally the default device). So, to direct sound to a specific card you must either mess with the alsa configuration, or use the hw:/plughw:. This is my current understanding. I know there are people who knows more which is reading this... My basic impression is that the alsa device addressing schema is a bit hard to handle, especially with more than one card. And although it's not that common that users have more than one card in their box, USB and bluetooth headsets are important use cases involving more than one card from the alsa perpective. It's also important to sort out the different roles in device management. The ultimate goal in my eyes is to create a new UI focused on making it much easier to setup the sound system. The major task is really up to the UI code here. My basic view is that the driver should provide a reasonably complete list of devices with as much information as possible. This does not preclude hiding nonsense like surround capture, But the finer tricks should be done by the UI, and we should be careful not to hide to much in the driver. The UI has other options than to just take the list from the driver and place in a select box IMHO. The plughw/hw is indeed complex. But to my understanding, it's really the only way to fix the basic task "Send sound to and get capture from the USB headset/main box/bluetooth headset". Of course, we could handle this by reconfiguring alsa. But since devices comes and goes away, and then comes back with a new number (sigh...) this is not that easy. Maybe the complexity of hw/plughw could be handled by a new UI? --alec ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
Re: [Ekiga-list] Alsa device enumeration
Le jeudi 12 mars 2009 à 20:41 +0100, Alec Leamas a écrit : > Damien Sandras wrote: > > Le mercredi 11 mars 2009 à 20:46 +0100, Alec Leamas a écrit : > > > >> There are some problems w the current code which enumerates the alsa > >> devices. I have a sketch for a patch, this is really not that > >> complicated these days, there are new library routines which support it. > >> > >> The problem is just what devices to show, and how. Basically, there is > >> two kind of devices: the classic devices which we partly see today, and > >> several "new" devices such as pulse, hdmi, and several surround*. They > >> exist for both capture and playback. The list for my box is below. The > >> basic ideas are stolen from Skype. The important thing is the left > >> column, the right is just internal stuff (hey, use a fixed font to see > >> this...). > >> > >> Any thoughts out there about this list? > >> > >> > > > > I would clean it and remove entries with hw only. Because that's > > complex. What is Surround40 capture ? > > > Surround capture is in my eyes complete nonsense. It's reported, though. > I totally agree that it should not be visible to users. > > The hw discussion is complex. Actually, there are documents on the alsa > site which boils down to that using hw:/plughw: is deprecated, they > recommend "front", "surround*" etc. But for capture, it just makes no > sense. And, more important, if you have more than one card there is no > way to e. g., direct sound to "device front at the usb headset". Which > device "front" actually refers to is a part of fthe alsa configuration > (normally the default device). So, to direct sound to a specific card > you must either mess with the alsa configuration, or use the > hw:/plughw:. This is my current understanding. I know there are people > who knows more which is reading this... > > My basic impression is that the alsa device addressing schema is a bit > hard to handle, especially with more than one card. And although it's > not that common that users have more than one card in their box, USB and > bluetooth headsets are important use cases involving more than one card > from the alsa perpective. > > It's also important to sort out the different roles in device > management. The ultimate goal in my eyes is to create a new UI focused > on making it much easier to setup the sound system. The major task is > really up to the UI code here. My basic view is that the driver should > provide a reasonably complete list of devices with as much information > as possible. This does not preclude hiding nonsense like surround > capture, But the finer tricks should be done by the UI, and we should be > careful not to hide to much in the driver. The UI has other options than > to just take the list from the driver and place in a select box IMHO. > > The plughw/hw is indeed complex. But to my understanding, it's really > the only way to fix the basic task "Send sound to and get capture from > the USB headset/main box/bluetooth headset". Of course, we could handle > this by reconfiguring alsa. But since devices comes and goes away, and > then comes back with a new number (sigh...) this is not that easy. > Maybe the complexity of hw/plughw could be handled by a new UI? I'm not sure. How can we know surround is non sense? How can we know other similar non sense entries... To my eyes, Alsa is really purely broken :-/ -- _ Damien Sandras (o- //\Ekiga Softphone : http://www.ekiga.org/ v_/_ Be IP : http://www.beip.be/ FOSDEM : http://www.fosdem.org/ SIP Phone : sip:dsand...@ekiga.net ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
Re: [Ekiga-list] Alsa device enumeration
Hi, On Thu, 12 Mar 2009, Damien Sandras wrote: I'm not sure. How can we know surround is non sense? How can we know other similar non sense entries... To my eyes, Alsa is really purely broken :-/ Very broken. Alsa stands for advanced linux sound architecture - which is misleading. Alsa rhymes with, and sounds very much like, the word "ulcer". = Sound cards - my colleagues and I did a count of the typical number. On a PC prior to 2000, (or so) the motherboard had a deficient onboard sound system, and you had to install a proper sound card. On the more recent PCs, (and laptops) the onboard sound is fine. However, the user will plugin in USB headsets (which are represented by alsa) as a sound card. Thus, we have to consider a PC as containing 2 sound cards: 1 that is always there 1 that comes and goes while the application is running. = The labelling is deficient. I have a machine here that has two sound cards. 1 on the motherboard, 3x3.5mm sockets, mic, speaker, line in (yes, 3 sockets.. only one is output) 1 ES1371 (ensoniq) 4x3.5mm sockets, mic, speker, linein, lineout Ok, now remember this description of 7x3.5mm sockets. How does alsa report this as ? output of aplay -l List of PLAYBACK Hardware Devices card 0: AudioPCI [Ensoniq AudioPCI], device 0: ES1371/1 [ES1371 DAC2/ADC] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: AudioPCI [Ensoniq AudioPCI], device 1: ES1371/2 [ES1371 DAC1] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: AudioPCI_1 [Ensoniq AudioPCI], device 0: ES1371/1 [ES1371 DAC2/ADC] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: AudioPCI_1 [Ensoniq AudioPCI], device 1: ES1371/2 [ES1371 DAC1] Subdevices: 1/1 Subdevice #0: subdevice #0 output of aplay -L front:CARD=AudioPCI,DEV=0 Ensoniq AudioPCI, ES1371 DAC2/ADC Front speakers surround40:CARD=AudioPCI,DEV=0 Ensoniq AudioPCI, ES1371 DAC2/ADC 4.0 Surround output to Front and Rear speakers iec958:CARD=AudioPCI,DEV=0 Ensoniq AudioPCI, ES1371 DAC2/ADC IEC958 (S/PDIF) Digital Audio Output null Discard all samples (playback) or generate zero samples (capture) front:CARD=AudioPCI_1,DEV=0 Ensoniq AudioPCI, ES1371 DAC2/ADC Front speakers surround40:CARD=AudioPCI_1,DEV=0 Ensoniq AudioPCI, ES1371 DAC2/ADC 4.0 Surround output to Front and Rear speakers iec958:CARD=AudioPCI_1,DEV=0 Ensoniq AudioPCI, ES1371 DAC2/ADC IEC958 (S/PDIF) Digital Audio Output = Here is the problem: we need to turn the above two lists into something meaningful for the user. both cards on this machine are reported as having two play subdevices. Yet only one is available. Which card is which? Which card is the motherboard, and which card is in the PCI socket? One of the tasks of the alsa code has to be to work out what the physical location of each card is, and report it accordingly. Now, this is just one example of the brokenness of alsa. Every application that uses alsa has to enumerate sound cards, work out their physical location, work out when new cards (usb headsets etc) are plugged in... Isn't that what libraries are for - so that common code goes in the library, and applications share the output of the common code??? Derek. -- Derek Smithies Ph.D. IndraNet Technologies Ltd. Email: de...@indranet.co.nz ph +64 3 365 6485 Web: http://www.indranet-technologies.com/ ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
Re: [Ekiga-list] Alsa device enumeration
Damien Sandras wrote: [ cut] I'm not sure. How can we know surround is non sense? How can we know other similar non sense entries... To my eyes, Alsa is really purely broken :-/ I didn't mean that surround is nonsense, no way. I just mean that surround capture/record/input makes no sense. 6 microphones, or what? One important aspect is that this "new" list contains new entries which are definitely meaningful including the pulse device. I have the same impression as you of the alsa naming scheme, But we have to live with t (?) I'll make Yet Another List. ___ ekiga-list mailing list ekiga-list@gnome.org http://mail.gnome.org/mailman/listinfo/ekiga-list
