Understanding Mixing of sounds in ALSA

Question

I can't seem to bend ALSA completely to my will, even after I spent a lot of time configuring it. I just do not seem to understand it well enough.

Here is my /etc/asound.conf:

pcm.dmixed {
        type asym
        playback.pcm {
                # See plugin:dmix at http://www.alsa-project.org/alsa-doc/alsa-lib/pcm_plugins.html
                type dmix

                # Don't block other users, e.g. the Timidity midi-player daemon
                # http://www.alsa-project.org/alsa-doc/alsa-lib/pcm_plugins.html
                ipc_key_add_uid true

                ipc_key 5678293
                ipc_perm 0660
                ipc_gid audio

                slave {
                        # 2 for stereo, 6 for surround51, 8 for surround71
                        channels 6
                        pcm {
                                format S32_LE
                                #format S16_LE

                                #rate 44100
                                rate 96000

                                # http://www.alsa-project.org/alsa-doc/alsa-lib/pcm_plugins.html
                                nonblock true

                                type hw
                                card 0
                                device 0
                                subdevice 0
                        }

                        period_size 1024

                        # If too large, use CONFIG_SND_HDA_PREALLOC_SIZE=2048
                        buffer_size 16384
                }
        }
        capture.pcm "hw:0"
}

pcm.!default{
        type plug
        slave.pcm "upmix20_51"
}

pcm.!surround20 {
        type plug
        slave.pcm "upmix20_51"
}

pcm.!surround40 {
        type plug
        slave.pcm "dmixed"
        route_policy duplicate
}

pcm.!surround51 {
        type plug
        slave.pcm "dmixed"
}

pcm.upmix20_51 {
        slave.pcm "dmixed"
        slave.channels 6
        type route

        # Front and rear
        ttable.0.0 0.9
        ttable.1.1 0.9
        ttable.2.2 1
        ttable.3.3 1

        # Center and LFE
        ttable.4.4 1
        ttable.5.5 1

        # Front left/right to center
        ttable.0.4 0.5
        ttable.1.4 0.5

        # Front left/right to rear
        ttable.0.2 0.5
        ttable.1.3 0.5
}

What it currently achieves is:

• Dmixing (I do not understand what this actually does, I think it ensures that the device will play sound no matter how many sources are outputting to alsa) with dmixed

• Forces sample rate to 96Khz

• Forces format to S32_LE (I think this works, although my media player doesn't say it works, I think it works because if I put S24_LE the sound will break.)

• Upmix 2.0 sounds to 5.1 with "upmix20_51" which is set as a slave to the "!default" pcm which I think is the normal output for 2.0 files.

• Upmix 4.0 to 5.1 via channel duplication (route_policy duplicate)

• Plays 5.1 sound normally granted it uses the "!surround51" pcm as it should.

What I would like to know is:

• What exactly is dmixing for?
• Is the pcm.!surround20 redundant? (Should I just remove it?)
• Can I check somehow if sample rate conversion is currently done with hardware or software?
• Will forcing the Frequency to 96khz and format to 32-bit have any benefits for quality?
• Will 5.1 sources ever use the "!default" pcm? And if it can happen, will the upmixing that is tied to it be problematic for that 5.1 source? Is there a better way to upmix 2.0 sources only?
• How do I upmix Mono sound to Stereo sound and then upmix that Stereo sound to 5.1 sound
• How do I downmix 7.1 sources (i.e. divide the side speakers between the front and rear speakers) to 5.1 in case I get one of those rare sources
• Is there any better/higher quality method to do upmixing than what I am currently doing with alsa? Any method that allows for greater control over how the sound is mixed?