path: root/Documentation/sound/oss/README.awe
diff options
Diffstat (limited to 'Documentation/sound/oss/README.awe')
1 files changed, 218 insertions, 0 deletions
diff --git a/Documentation/sound/oss/README.awe b/Documentation/sound/oss/README.awe
new file mode 100644
index 00000000000..80054cd8fcd
--- /dev/null
+++ b/Documentation/sound/oss/README.awe
@@ -0,0 +1,218 @@
+ AWE32 Sound Driver for Linux / FreeBSD
+ version 0.4.3; Nov. 1, 1998
+ Takashi Iwai <>
+This is a sound driver extension for SoundBlaster AWE32 and other
+compatible cards (AWE32-PnP, SB32, SB32-PnP, AWE64 & etc) to enable
+the wave synth operations. The driver is provided for Linux 1.2.x
+and 2.[012].x kernels, as well as FreeBSD, on Intel x86 and DEC
+Alpha systems.
+This driver was written by Takashi Iwai <>,
+and provided "as is". The original source (awedrv-0.4.3.tar.gz) and
+binary packages are available on the following URL:
+Note that since the author is apart from this web site, the update is
+not frequent now.
+To enable this driver on linux-2.[01].x kernels, you need turn on
+"AWE32 synth" options in sound menu when configure your linux kernel
+and modules. The precise installation procedure is described in the
+AWE64-Mini-HOWTO and linux-kernel/Documetation/sound/AWE32.
+If you're using PnP cards, the card must be initialized before loading
+the sound driver. There're several options to do this:
+ - Initialize the card via ISA PnP tools, and load the sound module.
+ - Initialize the card on DOS, and load linux by loadlin.exe
+ - Use PnP kernel driver (for Linux-2.x.x)
+The detailed instruction for the solution using isapnp tools is found
+in many documents like above. A brief instruction is also included in
+the installation document of this package.
+For PnP driver project, please refer to the following URL:
+The awedrv has several different playing modes to realize easy channel
+allocation for MIDI songs. To hear the exact sound quality, you need
+to obtain the extended sequencer program, drvmidi or playmidi-2.5.
+For playing MIDI files, you *MUST* load the soundfont file on the
+driver previously by sfxload utility. Otherwise you'll here no sounds
+at all! All the utilities and driver source packages are found in the
+above URL. The sfxload program is included in the package
+awesfx-0.4.3.tgz. Binary packages are available there, too. See the
+instruction in each package for installation.
+Loading a soundfont file is very simple. Just execute the command
+ % sfxload synthgm.sbk
+Then, sfxload transfers the file "synthgm.sbk" to the driver.
+Both SF1 and SF2 formats are accepted.
+Now you can hear midi musics by a midi player.
+ % drvmidi foo.mid
+If you run MIDI player after MOD player, you need to load soundfont
+files again, since MOD player programs clear the previous loaded
+samples by their own data.
+If you have only 512kb on the sound card, I recommend to use dynamic
+sample loading via -L option of drvmidi. 2MB GM/GS soundfont file is
+available in most midi files.
+ % sfxload synthgm
+ % drvmidi -L 2mbgmgs foo.mid
+This makes a big difference (believe me)! For more details, please
+refer to the FAQ list which is available on the URL above.
+The current chorus, reverb and equalizer status can be changed by
+aweset utility program (included in awesfx package). Note that
+some awedrv-native programs (like drvmidi and xmp) will change the
+current settings by themselves. The aweset program is effective
+only for other programs like playmidi.
+Compile conditions are defined in awe_config.h.
+[Compatibility Conditions]
+The following flags are defined automatically when using installation
+shell script.
+ indicates your Linux kernel supports module for each sound card
+ (in recent 2.1 or 2.2 kernels and unofficial patched 2.0 kernels
+ as distributed in the RH5.0 package).
+ This flag is automatically set when you're using 2.1.x kernels.
+ You can pass the base address and memory size via the following
+ module options,
+ io = base I/O port address (eg. 0x620)
+ memsize = DRAM size in kilobytes (eg. 512)
+ As default, AWE driver probes these values automatically.
+[Hardware Conditions]
+You DON'T have to define the following two values.
+Define them only when the driver couldn't detect the card properly.
+- AWE_DEFAULT_BASE_ADDR (default: not defined)
+ specifies the base port address of your AWE32 card.
+ 0 means to autodetect the address.
+- AWE_DEFAULT_MEM_SIZE (default: not defined)
+ specifies the memory size of your AWE32 card in kilobytes.
+ -1 means to autodetect its size.
+[Sample Table Size]
+From ver.0.4.0, sample tables are allocated dynamically (except
+Linux-1.2.x system), so you need NOT to touch these parameters.
+Linux-1.2.x users may need to increase these values to appropriate size
+if the sound card is equipped with more DRAM.
+[Other Conditions]
+- AWE_ALWAYS_INIT_FM (default: not defined)
+ indicates the AWE driver always initialize FM passthrough even
+ without DRAM on board. Emu8000 chip has a restriction for playing
+ samples on DRAM that at least two channels must be occupied as
+ passthrough channels.
+- AWE_DEBUG_ON (default: defined)
+ turns on debugging messages if defined.
+- AWE_HAS_GUS_COMPATIBILITY (default: defined)
+ Enables GUS compatibility mode if defined, reading GUS patches and
+ GUS control commands. Define this option to use GMOD or other
+ GUS module players.
+- CONFIG_AWE32_MIDIEMU (default: defined)
+ Adds a MIDI emulation device by Emu8000 wavetable. The emulation
+ device can be accessed as an external MIDI, and sends the MIDI
+ control codes directly. XG and GS sysex/NRPN are accepted.
+ No MIDI input is supported.
+- CONFIG_AWE32_MIXER (default: not defined)
+ Adds a mixer device for AWE32 bass/treble equalizer control.
+ You can access this device using /dev/mixer?? (usually mixer01).
+- AWE_USE_NEW_VOLUME_CALC (default: defined)
+ Use the new method to calculate the volume change as compatible
+ with DOS/Win drivers. This option can be toggled via aweset
+ program, or drvmidi player.
+- AWE_CHECK_VTARGET (default: defined)
+ Check the current volume target value when searching for an
+ empty channel to allocate a new voice. This is experimentally
+ implemented in this version. (probably, this option doesn't
+ affect the sound quality severely...)
+- AWE_ALLOW_SAMPLE_SHARING (default: defined)
+ Allow sample sharing for differently loaded patches.
+ This function is available only together with awesfx-0.4.3p3.
+ Note that this is still an experimental option.
+- DEF_FM_CHORUS_DEPTH (default: 0x10)
+ The default strength to be sent to the chorus effect engine.
+ From 0 to 0xff. Larger numbers may often cause weird sounds.
+- DEF_FM_REVERB_DEPTH (default: 0x10)
+ The default strength to be sent to the reverb effect engine.
+ From 0 to 0xff. Larger numbers may often cause weird sounds.
+Thanks to Witold Jachimczyk ( for much advice
+on programming of AWE32. Much code is brought from his AWE32-native
+MOD player, ALMP.
+The port of awedrv to FreeBSD is done by Randall Hopper
+The new volume calculation routine was derived from Mark Weaver's
+ADIP compatible routines.
+I also thank linux-awe-ml members for their efforts
+to reboot their system many times :-)
+- Complete DOS/Win compatibility
+- DSP-like output
+Copyright (C) 1996-1998 Takashi Iwai
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+GNU General Public License for more details.
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.