MikMod sound library


Next: , Previous: (dir), Up: (dir)

MikMod Sound Library

This manual documents the MikMod Sound Library, version 3.1.20.


Next: , Previous: Top, Up: Top

1 Introduction

The MikMod sound library is an excellent way for a programmer to add music and sound effects to an application. It is a powerful and flexible library, with a simple and easy-to-learn API.

Besides, the library is very portable and runs under a lot of Unices, as well as under OS/2, MacOS and Windows. Third party individuals also maintain ports on other systems, including MS-DOS, and BeOS.

MikMod is able to play a wide range of module formats, as well as digital sound files. It can take advantage of particular features of your system, such as sound redirection over the network. And due to its modular nature, the library can be extended to support more sound or module formats, as well as new hardware or other sound output capabilities, as they appear.


Next: , Previous: Introduction, Up: Top

2 Tutorial

This chapter will describe how to quickly incorporate MikMod's power into your programs. It doesn't cover everything, but that's a start and I hope it will help you understand the library philosophy.

If you have a real tutorial to put here, you're welcome ! Please send it to me...


Next: , Previous: Tutorial, Up: Tutorial

2.1 MikMod Concepts

MikMod's sound output is composed of several sound voices which are mixed, either in software or in hardware, depending of your hardware configuration. Simple sounds, like sound effects, use only one voice, whereas sound modules, which are complex arrangements of sound effects, use several voices.

MikMod's functions operate either globally, or at the voice level. Differences in the handling of sound effects and modules are kept minimal, at least for the programmer.

The sound playback is done by a sound driver. MikMod provides several sound drivers: different hardware drivers, and some software drivers to redirect sound in a file, or over the network. You can even add your own driver, register it to make it known by the library, and select it (this is exactly what the module plugin of xmms does).


Next: , Previous: MikMod Concepts, Up: Tutorial

2.2 A Skeleton Program

To use MikMod in your program, there are a few steps required:

Here's a program which meets all those conditions:

     /* MikMod Sound Library example program: a skeleton */
     
     #include <mikmod.h>
     
     main()
     {
     	/* register all the drivers */
     	MikMod_RegisterAllDrivers();
     
     	/* initialize the library */
     	MikMod_Init("");
     
     	/* we could play some sound here... */
     
     	/* give up */
     	MikMod_Exit();
     }

This program would be compiled with the following command line: cc -o example example.c `libmikmod-config --cflags` `libmikmod-config --libs`

Although this programs produces no useful result, many things happen when you run it. The call to MikMod_RegisterAllDrivers registers all the drivers embedded in the MikMod library. Then, MikMod_Init chooses the more adequate driver and initializes it. The program is now ready to produce sound. When sound is not needed any more, MikMod_Exit is used to relinquish memory and let other programs have access to the sound hardware.


Next: , Previous: A Skeleton Program, Up: Tutorial

2.3 Playing Modules

Our program is not really useful if it doesn't produce sound. Let's suppose you've got this good old module, “Beyond music”, in the file beyond music.mod. How about playing it ?

To do this, we'll use the following code:

     /* MikMod Sound Library example program: a simple module player */
     
     #include <unistd.h>
     #include <mikmod.h>
     
     main()
     {
         MODULE *module;
     
         /* register all the drivers */
         MikMod_RegisterAllDrivers();
     
         /* register all the module loaders */
         MikMod_RegisterAllLoaders();
     
         /* initialize the library */
         md_mode |= DMODE_SOFT_MUSIC;
         if (MikMod_Init("")) {
             fprintf(stderr, "Could not initialize sound, reason: %s\n",
                     MikMod_strerror(MikMod_errno));
             return;
         }
     
         /* load module */
         module = Player_Load("beyond music.mod", 64, 0);
         if (module) {
             /* start module */
             Player_Start(module);
     
             while (Player_Active()) {
                 /* we're playing */
                 usleep(10000);
                 MikMod_Update();
             }
     
             Player_Stop();
             Player_Free(module);
         } else
             fprintf(stderr, "Could not load module, reason: %s\n",
                     MikMod_strerror(MikMod_errno));
     
         /* give up */
         MikMod_Exit();
     }

What's new here ? First, we've not only registered MikMod's device driver, but also the module loaders. MikMod comes with a large choice of module loaders, each one for a different module type. Since every loader is called to determine the type of the module when we try to load them, you may want to register only a few of them to save time. In our case, we don't matter, so we happily register every module loader.

Then, there's an extra line before calling MikMod_Init. We change the value of MikMod's variable md_mode to tell the library that we want the module to be processed by the software. If you're the happy owner of a GUS-type card, you could use the specific hardware driver for this card, but in this case you should not set the DMODE_SOFT_MUSIC flag.

We'll ensure that MikMod_Init was successful. Note that, in case of error, MikMod provides the variable MikMod_errno, an equivalent of the C library errno for MikMod errors, and the function MikMod_strerror, an equivalent to strerror.

Now onto serious business ! The module is loaded with the Player_Load function, which takes the name of the module file, and the number of voices afforded to the module. In this case, the module has only 4 channels, so 4 voices, but complex Impulse Tracker modules can have a lot of voices (as they can have as many as 256 virtual channels with so-called “new note actions”). Since empty voices don't cost time to be processed, it is safe to use a big value, such as 64 or 128. The third parameter is the “curiosity” of the loader: if nonzero, the loader will search for hidden parts in the module. However, only a few module formats can embed hidden or non played parts, so we'll use 0 here.

Now that the module is ready to play, let's play it. We inform the player that the current module is module with Player_Start. Playback starts, but we have to update it on a regular basis. So there's a loop on the result of the Player_Active function, which will tell us if the module has finished. To update the sound, we simply call MikMod_Update.

After the module has finished, we tell the player its job is done with Player_Stop, and we free the module with Player_Free.


Next: , Previous: Playing Modules, Up: Tutorial

2.4 Playing Sound Effects

MikMod is not limited to playing modules, it can also play sound effects, that is, module samples. It's a bit more complex than playing a module, because the module player does a lot of things for us, but here we'll get more control over what is actually played by the program. Let's look at an example:

     /* MikMod Sound Library example program: sound effects */
     
     #include <unistd.h>
     #include <mikmod.h>
     
     main()
     {
         int i;
         /* sound effects */
         SAMPLE *sfx1, *sfx2;
         /* voices */
         int v1, v2;
     
         /* register all the drivers */
         MikMod_RegisterAllDrivers();
     
         /* initialize the library */
         md_mode |= DMODE_SOFT_SNDFX;
         if (MikMod_Init("")) {
             fprintf(stderr, "Could not initialize sound, reason: %s\n",
                     MikMod_strerror(MikMod_errno));
             return;
         }
     
         /* load samples */
         sfx1 = Sample_Load("first.wav");
         if (!sfx1) {
             MikMod_Exit();
             fprintf(stderr, "Could not load the first sound, reason: %s\n",
                     MikMod_strerror(MikMod_errno));
             return;
         }
         sfx2 = Sample_Load("second.wav");
         if (!sfx2) {
             Sample_Free(sfx1);
             MikMod_Exit();
             fprintf(stderr, "Could not load the second sound, reason: %s\n",
                     MikMod_strerror(MikMod_errno));
             return;
         }
     
         /* reserve 2 voices for sound effects */
         MikMod_SetNumVoices(-1, 2);
     
         /* get ready to play */
         MikMod_EnableOutput();
     
         /* play first sample */
         v1 = Sample_Play(sfx1, 0, 0);
         for(i = 0; i < 5; i++) {
             MikMod_Update();
             usleep(100000);
         }
     
         /* half a second later, play second sample */
         v2 = Sample_Play(sfx2, 0, 0);
         do {
             MikMod_Update();
             usleep(100000);
         } while (!Voice_Stopped(v2));
     
         MikMod_DisableOutput();
     
         Sample_Free(sfx2);
         Sample_Free(sfx1);
     
         MikMod_Exit();
     }

As in the previous example, we begin by registering the sound drivers and initializing the library. We also ask for software mixing by modifying the variable md_mode.

It's time to load our files, with the Sample_Load function. Don't forget to test the return value — it looks ugly here on such a small example, but it's a good practice...

Since we want to play two samples, we have to use at least two voices for this, so we reserve them with a MikMod_SetNumVoices call. The first parameter sets the number of module voices, and the second parameter the number of sound effect voices. We don't want to set the number of module voices here (it's part of the module player's duty), so we use the value -1 to keep the current value, and we reserve two sound effect voices.

Now we're ready to play, so we call MikMod_EnableOutput to make the driver ready. Sound effects are played by the Sample_Play function. You just have to specify which sample you want to play, the offset from which you want to start, and the playback flags. More on this later. The function returns the number of the voice associated to the sample.

We play the first sample for half a second, then we start to play the second sample. Since we've reserved two channels, both samples play simultaneously. We use the Voice_Stopped function to stop the playback: it returns the current status of the voice argument, which is zero when the sample plays and nonzero when it has finished. So the do loop will stop exactly when the second sample is finished, regardless of the length of the first sample.

To finish, we get rid of the samples with Sample_Free.


Previous: Playing Sound Effects, Up: Tutorial

2.5 More Sound Effects

Sound effects have some attributes that can be affected to control the playback. These are speed, panning, and volume. Given a voice number, you can affect these attributes with the Voice_SetFrequency, Voice_SetPanning and Voice_SetVolume functions.

In the previous example, we'll replace the actual sound code, located between the calls to MikMod_EnableOutput and MikMod_DisableOutput, with the following code:

         Sample_Play(sfx1, 0, 0);
         for(i = 0; i < 5; i++) {
             MikMod_Update();
             usleep(100000);
         }
         v2 = Sample_Play(sfx2, 0, SFX_CRITICAL);
         i = 0;
         do {
             MikMod_Update();
             usleep(100000);
             v1 = Sample_Play(sfx1, 0, 0);
             Voice_SetVolume(v1, 160);
             Voice_SetFrequency(v1, (sfx1->speed * (100 + i)) / 100);
             Voice_SetPanning(v2, (i++ & 1) ? PAN_LEFT : PAN_RIGHT);
         } while (!Voice_Stopped(v2));

The first thing you'll notice, is the SFX_CRITICAL flag used to play the second sample. Since the do loop will add another sample every 100 milliseconds, and we reserved only two voices, the oldest voice will be cut each time this is necessary. Doing this would cut the second sample in the second iteration of the loop. However, since we flagged this sound as “critical”, it won't be cut until it is finished or we stop it with a Voice_Stop call. So the second sample will play fine, whereas the first sample will be stopped every loop iteration.

Then, we choose to play the first sample a bit lower, with Voice_SetVolume. Volume voices range from 0 (silence) to 256. In this case we play the sample at 160. To make the sound look weird, we also change its frequency with Voice_SetFrequency. The computation in the example code makes the frequency more and more high (starting from the sample frequency and then increasing from 1% each iteration).

And to demonstrate the Voice_SetPanning function, we change the panning of the second sample at each iteration from the left to the right. The argument can be one of the standard panning PAN_LEFT, PAN_RIGHT, PAN_CENTER and PAN_SURROUND1, or a numeric value between 0 (PAN_LEFT) and 255 (PAN_RIGHT).


Next: , Previous: Tutorial, Up: Top

3 Using the Library

This chapter describes the various parts of the library and their uses.


Next: , Previous: Using the Library, Up: Using the Library

3.1 Library Version

If your program is dynamically linked with the MikMod library, you should check which version of the library you're working with. To do this, the library defines a few constants and a function to help you determine if the current library is adequate for your needs or if it has to be upgraded.

When your program includes mikmod.h, the following constants are defined:

So your program can tell with which version of the library it has been compiled this way:

     printf("Compiled with MikMod Sound Library version %ld.%ld.%ld\n",
            LIBMIKMOD_VERSION_MAJOR,
            LIBMIKMOD_VERSION_MINOR,
            LIBMIKMOD_REVISION);

The library defines the function MikMod_GetVersion which returns the value of LIBMIKMOD_VERSION for the library. If this value is greater than or equal to the value of LIBMIKMOD_VERSION for your program, your program will work; otherwise, you'll have to inform the user that he has to upgrade the library:

     {
         long engineversion = MikMod_GetVersion();
     
         if (engineversion < LIBMIKMOD_VERSION) {
             printf("MikMod library version (%ld.%ld.%ld) is too old.\n",
                    (engineversion >> 16) & 255,
                    (engineversion >> 8) & 255,
                    (engineversion) & 255);
             printf("This programs requires at least version %ld.%ld.%ld\n",
                    LIBMIKMOD_VERSION_MAJOR,
                    LIBMIKMOD_VERSION_MINOR,
                    LIBMIKMOD_REVISION);
             puts("Please upgrade your MikMod library.");
             exit(1);
         }
     }


Next: , Previous: Library Version, Up: Using the Library

3.2 Type Definitions

MikMod defines several data types to deal with modules and sample data. These types have the same memory size on every platform MikMod has been ported to.

These types are:


Next: , Previous: Type Definitions, Up: Using the Library

3.3 Error Handling

Although MikMod does its best to do its work, there are times where it can't. For example, if you're trying to play a corrupted file, well, it can't.

A lot of MikMod functions return pointers or BOOL values. If the pointer is NULL or the BOOL is 0 (false), an error has occurred.

MikMod errors are returned in the variable MikMod_errno. Each possible error has a symbolic error code, beginning with MMERR_. For example, if MikMod can't open a file, MikMod_errno will receive the value MMERR_OPENING_FILE.

You can get an appropriate error message to display from the function MikMod_strerror.

There is a second error variable named MikMod_critical. As its name suggests, it is only set if the error lets the library in an unstable state. This variable can only be set by the functions MikMod_Init, MikMod_SetNumVoices and MikMod_EnableOutput. If one of these functions return an error and MikMod_critical is set, the library is left in the uninitialized state (i.e. it was not initialized, or MikMod_Exit was called).

If you prefer, you can use a callback function to get notified of errors. This function must be prototyped as void MyFunction(void). Then, call MikMod_RegisterHandler with your function as argument to have it notified when an error occurs. There can only be one callback function registered, but MikMod_RegisterHandler will return you the previous handler, so you can chain handlers if you want to.


Next: , Previous: Error Handling, Up: Using the Library

3.4 Library Initialization and Core Functions

To initialize the library, you must register some sound drivers first. You can either register all the drivers embedded in the library for your platform with MikMod_RegisterAllDrivers, or register only some of them with MikMod_RegisterDriver. If you choose to register the drivers manually, you must be careful in their order, since MikMod_Init will try them in the order you registered them. The MikMod_RegisterAllDrivers function registers the network drivers first (for playing sound over the network), then the hardware drivers, then the disk writers, and in last resort, the nosound driver. Registering the nosound driver first would not be a very good idea...

You can get some printable information regarding the registered drivers with MikMod_InfoDriver; don't forget to call free on the returned string when you don't need it anymore.

After you've registered your drivers, you can initialize the sound playback with MikMod_Init, passing specific information to the driver if necessary. If you set the variable md_device to zero, which is its default value, the driver will be autodetected, that is, the first driver in the list that is available on the system will be used; otherwise only the driver whose order in the list of the registered drivers is equal to md_device will be tried. If your playback settings, in the variables md_mixfreq and md_mode, are not supported by the device, MikMod_Init will fail.

You can then choose the number of voices you need with MikMod_SetNumVoices, and activate the playback with MikMod_EnableOutput.

Don't forget to call MikMod_Update as often as possible to process the sound mixing. If necessary, fork a dedicated process to do this, or if the library is thread-safe on your system, use a dedicated thread.

If you want to change playback settings, most of them can't be changed on the fly. You'll need to stop the playback and reinitialize the driver. Use MikMod_Active to check if there is still sound playing; in this case, call MikMod_DisableOutput to end playback. Then, change your settings and call MikMod_Reset. You're now ready to select your number of voices and restart playback.

When your program ends, don't forget to stop playback and call MikMod_Exit to leave the sound hardware in a coherent state.

On systems that have pthreads, libmikmod is thread-safe2. You can check this in your programs with the MikMod_InitThreads function. If this function returns 1, the library is thread-safe.

The main benefit of thread-safety is that MikMod_Update can be called from a separate thread, which often makes application design easier. However, several libmikmod global variables are accessible from all your threads, so when more than one thread need to access libmikmod variables, you'll have to protect these access with the MikMod_Lock and MikMod_Unlock functions. If libmikmod is not thread-safe, these functions are no-ops.


Next: , Previous: Library Initialization, Up: Using the Library

3.5 Samples and Voice Control

Currently, MikMod only supports uncompressed mono WAV files as samples. You can load a sample by calling Sample_Load with a filename, or by calling Sample_LoadFP with an open FILE* pointer. These functions return a pointer to a SAMPLE structure, or NULL in case of error.

The SAMPLE structure has a few interesting fields:

Altering one of those fields will affect all voices currently playing the sample. You can achieve the same result on a single voice with the functions Voice_SetFrequency, Voice_SetVolume and Voice_SetPanning. Since the same sample can be played with different frequency, volume and panning parameters on each voice, you can get voice specific information with Voice_GetFrequency, Voice_GetVolume and Voice_GetPanning.

You can also make your sample loop by setting the fields loopstart and loopend and or'ing flags with SF_LOOP. To compute your loop values, the field length will be useful. However, you must know that all the sample length are expressed in samples, i.e. 8 bits for an 8 bit sample, and 16 bit for a 16 bit sample... Test flags for the value SF_16BITS to know this.

Speaking of flags, if you're curious and want to know the original format of the sample on disk (since libmikmod does some work on the sample data internally), refer to the inflags field.

If the common forward loop isn't enough, you can play with some other flags: SF_BIDI will make your sample loop “ping pong” (back and forth), and SF_REVERSE will make it play backwards.

To play your sample, use the Sample_Play function. This function will return a voice number which enable you to use the Voice_xx functions.

The sample will play until another sample takes over its voice (when you play more samples than you reserved sound effect voices), unless it has been flagged as SFX_CRITICAL. You can force it to stop with Voice_Stop, or you can force another sample to take over this voice with Voice_Play; however Voice_Play doesn't let you flag the new sample as critical.

Non looping samples will free their voice channel as soon as they are finished; you can know the current playback position of your sample with Voice_GetPosition. If it is zero, either the sample has finished playing or it is just beginning; use Voice_Stopped to know.

When you don't need a sample anymore, don't forget to free its memory with Sample_Free.


Next: , Previous: Samples and Voice Control, Up: Using the Library

3.6 Modules and Player Control

As for the sound drivers, you have to register the module loaders you want to use for MikMod to be able to load modules. You can either register all the module loaders with MikMod_RegisterAllLoaders, or only a few of them with MikMod_RegisterLoader. Be careful if you choose this solution, as the 15 instrument MOD loader has to be registered last, since loaders are called in the order they were register to identify modules, and the detection of this format is not fully reliable, so other modules might be mistaken as 15 instrument MOD files.

You can get some printable information regarding the registered loaders with MikMod_InfoLoader; don't forget to call free on the returned string when you don't need it anymore.

Note that, contrary to the sound drivers, you can register module loaders at any time, it doesn't matter.

For playlists, you might be interested in knowing the module title first, and Player_LoadTitle will give you this information. Don't forget to free the returned text when you don't need it anymore.

You can load a module either with Player_Load and the name of the module, or with Player_LoadFP and an open FILE* pointer. These functions also expect a maximal number of voices, and a curiosity flag. Unless you have excellent reasons not to do so, choose a big limit, such as 64 or even 128 for complex Impulse Tracker modules. Both functions return a pointer to an MODULE structure, or NULL if an error occurs.

You'll find some useful information in this structure:

Now that the module is loaded, you need to tell the module player that you want to play this particular module with Player_Start (the player can only play one module, but you can have several modules in memory). The playback begins. Should you forget which module is playing, Player_GetModule will return it to you.

You can change the current song position with the functions Player_NextPosition, Player_PrevPosition and Player_SetPosition, the speed with Player_SetSpeed and Player_SetTempo, and the volume (ranging from 0 to 128) with Player_SetVolume.

Playback can be paused or resumed with Player_TogglePause. Be sure to check with Player_Paused that it isn't already in the state you want !

Fine player control is achieved by the functions Player_Mute, Player_UnMute and Player_ToggleMute which can silence or resume a set of module channels. The function Player_Muted will return the state of a given channel. And if you want even more control, you can get the voice corresponding to a module channel with Player_GetChannelVoice and act directly on the voice.

Modules play only once, but can loop indefinitely if they are designed to do so. You can change this behavior with the wrap and loop of the MODULE structure; the first one, if set, will make the module restart when it's finished, and the second one, if set, will prevent the module from jumping backwards.

You can test if the module is still playing with Player_Active, and you can stop it at any time with Player_Stop. When the module isn't needed anymore, get rid of it with Player_Free.


Previous: Modules and Player Control, Up: Using the Library

3.7 Loading Data from Memory

If you need to load modules or sound effects from other places than plain files, you can use the MREADER and MWRITER objects to achieve this.

The MREADER and MWRITER structures contain a list of function pointers, which emulate the behaviour of a regular FILE * object. In fact, all functions which take filenames or FILE * as arguments are only wrappers to a real function which takes an MREADER or an MWRITER argument.

So, if you need to load a module from memory, or for a multi-file archive, for example, all you need is to build an adequate MREADER object, and use Player_LoadGeneric instead of Player_Load or Player_LoadFP. For samples, use Sample_LoadGeneric instead of Sample_Load or Sample_LoadFP.


Next: , Previous: Using the Library, Up: Top

4 Library Reference

This chapter describes in more detail all the functions and variables provided by the library. See Type Definitions, for the basic type reference.


Next: , Previous: Library Reference, Up: Library Reference

4.1 Variable Reference

4.1.1 Error Variables

The following variables are set by the library to return error information.

int MikMod_errno
When an error occurs, this variable contains the error code. See Error Reference, for more information.
BOOL MikMod_critical
When an error occurs, this variable informs of the severity of the error. Its value has sense only if the value of MikMod_errno is different from zero. If the value of MikMod_critical is zero, the error wasn't fatal and the library is in a stable state. However, if it is nonzero, then the library can't be used and has reseted itself to the uninitialized state. This often means that the mixing parameters you choose were not supported by the driver, or that it doesn't has enough voices for your needs if you called MikMod_SetNumVoices.

4.1.2 Sound Settings

The following variables control the sound output parameters and their changes take effect immediately.

UBYTE md_musicvolume
Volume of the module. Allowed values range from 0 to 128. The default value is 128.
UBYTE md_pansep
Stereo channels separation. Allowed values range from 0 (no separation, thus mono sound) to 128 (full channel separation). The default value is 128.
UBYTE md_reverb
Amount of sound reverberation. Allowed values range from 0 (no reverberation) to 15 (a rough estimate for chaos...). The default value is 0.
UBYTE md_sndfxvolume
Volume of the sound effects. Allowed values range from 0 to 128. The default value is 128.
UBYTE md_volume
Overall sound volume. Allowed values range from 0 to 128. The default value is 128.

4.1.3 Driver Settings

The following variables control more in-depth sound output parameters. Except for some md_mode flags, their changes do not have any effect until you call MikMod_Init or MikMod_Reset.

UWORD md_device
This variable contains the order, in the list of the registered drivers, of the sound driver which will be used for sound playback. This order is one-based; if this variable is set to zero, the driver is autodetected, which means the list is tested until a driver is present on the system. The default value is 0, thus driver is autodetected.
MDRIVER* md_driver
This variable points to the driver which is being used for sound playback, and is undefined when the library is uninitialized (before MikMod_Init and after MikMod_Exit). This variable is for information only, you should never attempt to change its value. Use md_driver and MikMod_Init (or MikMod_Reset) instead.
UWORD md_mixfreq
Sound playback frequency, in hertz. High values yield high sound quality, but need more computing power than lower values. The default value is 44100 Hz, which is compact disc quality. Other common values are 22100 Hz (radio quality), 11025 Hz (phone quality), and 8000 Hz (mu-law quality).
UWORD md_mode
This variable is a combination of several flags, to select which output mode to select. The following flags have a direct action to the sound output (i.e. changes take effect immediately):
DMODE_INTERP
This flag, if set, enables the interpolated mixers. Interpolated mixing gives better sound but takes a bit more time than standard mixing. If the library is built with the high quality mixer, interpolated mixing is always enabled, regardless of this flag.
DMODE_REVERSE
This flag, if set, exchanges the left and right stereo channels.
DMODE_SURROUND
This flag, if set, enables the surround mixers. Since surround mixing works only for stereo sound, this flag has no effect if the sound playback is in mono.

The following flags aren't taken in account until the sound driver is changed or reset:

DMODE_16BIT
This flag, if set, selects 16 bit sound mode. This mode yields better sound quality, but needs twice more mixing time.
DMODE_HQMIXER
This flag, if set, selects the high-quality software mixer. This mode yields better sound quality, but needs more mixing time. Of course, this flag has no effect if no DMODE_SOFT_xx flag is set.
DMODE_SOFT_MUSIC
This flag, if set, selects software mixing of the module.
DMODE_SOFT_SNDFX
This flag, if set, selects software mixing of the sound effects.
DMODE_STEREO
This flag, if set, selects stereo sound.

The default value of this variable is ‘DMODE_STEREO | DMODE_SURROUND | DMODE_16BITS | DMODE_SOFT_MUSIC | DMODE_SOFT_SNDFX’.


Next: , Previous: Variable Reference, Up: Library Reference

4.2 Structure Reference

Only the useful fields are described here; if a structure field is not described, you must assume that it's an internal field which must not be modified.

4.2.1 Drivers

The MDRIVER structure is not meant to be used by anything else than the core of the library, but its first four fields contain useful information for your programs:

CHAR* Name
Name of the driver, usually never more than 20 characters.
CHAR* Description
Description of the driver, usually never more than 50 characters.
UBYTE HardVoiceLimit
Maximum number of hardware voices for this driver, 0 if the driver has no hardware mixing support.
UBYTE SoftVoiceLimit
Maximum number of software voices for this driver, 0 if the driver has no software mixing support.
CHAR* Alias
A short name for the driver, without spaces, usually never more than 10 characters.

4.2.2 Modules

The MODULE structure gathers all the necessary information needed to play a module file, regardless of its initial format.

4.2.2.1 General Module Information

The fields described in this section contain general information about the module and should not be modified.

CHAR* songname
Name of the module.
CHAR* modtype
Type of the module (which tracker format).
CHAR* comment
Either the module comments, or NULL if the module doesn't have comments.
UWORD flags
Several module flags or'ed together.
UF_ARPMEM
If set, arpeggio effects have memory.
UF_BGSLIDES
If set, volume slide effects continue until a new note or a new effect is played.
UF_HIGHBPM
If set, the module is allowed to have its tempo value (bpm) over 255.
UF_INST
If set, the module has instruments and samples; otherwise, the module has only samples.
UF_LINEAR
If set, slide periods are linear; otherwise, they are logarithmic.
UF_NNA
If set, module uses new note actions (NNA) and the numvoices field is valid.
UF_NOWRAP
If set, pattern break on the last pattern does not continue to the first pattern.
UF_S3MSLIDES
If set, module uses old-S3M style volume slides (slides processed every tick); otherwise, it uses the standard style (slides processed every tick except the first).
UF_XMPERIODS
If set, module uses XM-type periods; otherwise, it uses Amiga periods.
UF_FT2QUIRKS
If set, module player will reproduce some FastTracker 2 quirks during playback.
UF_PANNING
If set, module use panning commands.

UBYTE numchn
The number of channels in the module.
UBYTE numvoices
If the module uses NNA, and this variable is not zero, it contains the limit of module voices; otherwise, the limit is set to the maxchan parameter of the Player_Loadxx functions.
UWORD numpos
The number of sound positions in the module.
UWORD numpat
The number of patterns.
UWORD numins
The number of instruments.
UWORD numsmp
The number of samples.
INSTRUMENT* instruments
Points to an array of instrument structures.
SAMPLE* samples
Points to an array of sample structures.
UBYTE realchn
During playback, this variable contains the number of active channels (not counting NNA channels).
UBYTE totalchn
During playback, this variable contains the total number of channels (including NNA channels).
ULONG sngtime
Elapsed song time, in 2^−10 seconds units (not exactly a millisecond). To convert this value to seconds, divide by 1024, not 1000 !
4.2.2.2 Playback Settings

The fields described here control the module playback and can be modified at any time, unless otherwise specified.

UBYTE initspeed
The initial speed of the module (Protracker compatible). Valid range is 1-32.
UBYTE inittempo
The initial tempo of the module (Protracker compatible). Valid range is 32-255.
UBYTE initvolume
The initial overall volume of the module. Valid range is 0-128.
UWORD panning[]
The current channel panning positions. Only the first numchn values are defined.
UBYTE chanvol[]
The current channel volumes. Only the first numchn values are defined.
UWORD bpm
The current tempo of the module. Use Player_SetTempo to change its value.
UBYTE sngspd
The current speed of the module. Use Player_SetSpeed to change its value.
UBYTE volume
The current overall volume of the module, in range 0-128. Use Player_SetVolume to change its value.
BOOL extspd
If zero, Protracker extended speed effect (in-module tempo modification) is not processed. The default value is 1, which causes this effect to be processed. However, some old modules might not play correctly if this effect is not neutralized.
BOOL panflag
If zero, panning effects are not processed. The default value is 1, which cause all panning effects to be processed. However, some old modules might not play correctly if panning is not neutralized.
BOOL wrap
If nonzero, module wraps to its restart position when it is finished, to play continuously. Default value is zero (play only once).
UBYTE reppos
The restart position of the module, when it wraps.
BOOL loop
If nonzero, all in-module loops are processed; otherwise, backward loops which decrease the current position are not processed (i.e. only forward loops, and backward loops in the same pattern, are processed). This ensures that the module never loops endlessly. The default value is 1 (all loops are processed).
BOOL fadeout
If nonzero, volume fades out during when last position of the module is being played. Default value us zero (no fadeout).
UWORD patpos
Current position (row) in the pattern being played. Must not be changed.
SWORD sngpos
Current song position. Do not change this variable directly, use Player_NextPosition, Player_PrevPosition or Player_SetPosition instead.
SWORD relspd
Relative playback speed. The value of this variable is added to the module tempo to define the actual playback speed. The default value is 0, which make modules play at their intended speed.

4.2.3 Module Instruments

Although the INSTRUMENT structure is intended for internal use, you might need to know its name:

CHAR* insname
The instrument text, theoretically its name, but often a message line.

4.2.4 Samples

The SAMPLE structure is used for sound effects and module samples as well. You can play with the following fields:

SWORD panning
Panning value of the sample. Valid values range from PAN_LEFT (0) to PAN_RIGHT (255), or PAN_SURROUND.
ULONG speed
Playing frequency of the sample, it hertz.
UBYTE volume
Sample volume. Valid range is 0-64.
UWORD flags
Several format flags or'ed together describing the format of the sample in memory.

Format flags:

SF_16BITS
If set, sample data is 16 bit wide; otherwise, it is 8 bit wide.
SF_BIG_ENDIAN
If set, sample data is in big–endian (Motorola) format; otherwise, it is in little–endian (Intel) format.
SF_DELTA
If set, sample is stored as delta values (differences between two consecutive samples); otherwise, sample is stored as sample values.
SF_ITPACKED
If set, sample data is packed with Impulse Tracker's compression method; otherwise, sample is not packed.
SF_SIGNED
If set, sample data is made of signed values; otherwise, it is made of unsigned values.
SF_STEREO
If set, sample data is stereo (two channels); otherwise, it is mono.

Playback flags:

SF_BIDI
If set, sample loops “ping pong” (back and forth).
SF_LOOP
If set, sample loops forward.
SF_REVERSE
If set, sample plays backwards.

UWORD inflags
Same as “flags”, but describing the format of the sample on disk.
ULONG length
Length of the sample, in samples. The length of a sample is 8 bits (1 byte) for a 8 bit sample, and 16 bits (2 bytes) for a 16 bit sample.
ULONG loopstart
Loop starting position, relative to the start of the sample, in samples.
ULONG loopend
Loop ending position, relative to the start of the sample, in samples.

4.2.5 MREADER

The MREADER contains the following function pointers:

int (*Seek)(struct MREADER*, long offset, int whence)
This function should have the same behaviour as fseek, with offset 0 meaning the start of the object (module, sample) being loaded.
long (*Tell)(struct MREADER*)
This function should have the same behaviour as ftell, with offset 0 meaning the start of the object being loaded.
BOOL (*Read)(struct MREADER*, void *dest, size_t length)
This function should copy length bytes of data into dest, and return zero if an error occured, and any nonzero value otherwise. Note that an end-of-file condition will not be considered as an error in this case.
int (*Get)(struct MREADER*)
This function should have the same behaviour as fgetc.
BOOL (*Eof)(struct MREADER*)
This function should have the same behaviour as feof.

For an example of how to build an MREADER object, please refer to the MFILEREADER object in file mmio/mmio.c in the library sources.

4.2.6 MWRITER

The MWRITER contains the following function pointers:

int (*Seek)(struct MWRITER*, long offset, int whence);
This function should have the same behaviour as fseek, with offset 0 meaning the start of the object being written.
long (*Tell)(struct MWRITER*);
This function should have the same behaviour as ftell, with offset 0 meaning the start of the object being written.
BOOL (*Write)(struct MWRITER*, void *src, size_t length);
This function should copy length bytes of data from src, and return zero if an error occured, and any nonzero value otherwise.
int (*Put)(struct MWRITER*, int data);
This function should have the same behaviour as fputc.

For an example of how to build an MWRITER object, please refer to the MFILEWRITER object in file mmio/mmio.c in the library sources.


Next: , Previous: Structure Reference, Up: Library Reference

4.3 Error Reference

The following errors are currently defined:

4.3.1 General Errors

MMERR_DYNAMIC_LINKING
This error occurs when a specific driver was requested, but the support shared library couldn't be loaded. Currently, the only drivers which can yield this error are the ALSA, EsounD and Ultra drivers.
MMERR_OPENING_FILE
This error occurs when a file can not be opened, either for read access from a xx_Loadxx function, or for write access from the disk writer drivers.
MMERR_OUT_OF_MEMORY
This error occurs when there is not enough virtual memory available to complete the operation, or there is enough memory but the calling process would exceed its memory limit. MikMod does not do any resource tuning, your program has to use the setrlimit function to do this if it needs to load very huge samples.

4.3.2 Sample Errors

MMERR_SAMPLE_TOO_BIG
This error occurs when the memory allocation of the sample data yields the error MMERR_OUT_OF_MEMORY.
MMERR_OUT_OF_HANDLES
This error occurs when your program reaches the limit of loaded samples, currently defined as 384, which should be sufficient for most cases.
MMERR_UNKNOWN_WAVE_TYPE
This error occurs when you're trying to load a sample which format is not recognized.

4.3.3 Module Errors

MMERR_ITPACK_INVALID_DATA
This error occurs when a compressed module sample is corrupt.
MMERR_LOADING_HEADER
This error occurs when you're trying to load a module which has a corrupted header, or is truncated.
MMERR_LOADING_PATTERN
This error occurs when you're trying to load a module which has corrupted pattern data, or is truncated.
MMERR_LOADING_SAMPLEINFO
This error occurs when you're trying to load a module which has corrupted sample information, or is truncated.
MMERR_LOADING_TRACK
This error occurs when you're trying to load a module which has corrupted track data, or is truncated.
MMERR_MED_SYNTHSAMPLES
This error occurs when you're trying to load a MED module which has synthsounds samples, which are currently not supported.3
MMERR_NOT_A_MODULE
This error occurs when you're trying to load a module which format is not recognized.
MMERR_NOT_A_STREAM
This error occurs when you're trying to load a sample with a sample which format is not recognized.

4.3.4 Driver Errors

4.3.4.1 Generic Driver Errors
MMERR_16BIT_ONLY
This error occurs when the sound device doesn't support non-16 bit linear sound output, which are the requested settings.
MMERR_8BIT_ONLY
This error occurs when the sound device doesn't support non-8 bit linear sound output, which are the requested settings.
MMERR_DETECTING_DEVICE
This error occurs when the driver's sound device has not been detected.
MMERR_INITIALIZING_MIXER
This error occurs when MikMod's internal software mixer could not be initialized properly.
MMERR_INVALID_DEVICE
This error occurs when the driver number (in md_device) is out of range.
MMERR_NON_BLOCK
This error occurs when the driver is unable to set the audio device in non blocking mode.
MMERR_OPENING_AUDIO
This error occurs when the driver can not open sound device.
MMERR_STEREO_ONLY
This error occurs when the sound device doesn't support mono sound output, which is the requested setting.
MMERR_ULAW
This error occurs when the sound device only supports uLaw output (which implies mono, 8 bit, and 8000 Hz sampling rate), which isn't the requested setting.
4.3.4.2 AudioFile Driver Specific Error
MMERR_AF_AUDIO_PORT
This error occurs when the AudioFile driver can not find a suitable AudioFile port.
4.3.4.3 AIX Driver Specific Errors
MMERR_AIX_CONFIG_CONTROL
This error occurs when the “Control” step of the device configuration has failed.
MMERR_AIX_CONFIG_INIT
This error occurs when the “Init” step of the device configuration has failed.
MMERR_AIX_CONFIG_START
This error occurs when the “Start” step of the device configuration has failed.
4.3.4.4 Ultra Driver Specific Errors
MMERR_GUS_RESET
This error occurs when the sound device couldn't be reset.
MMERR_GUS_SETTINGS
This error occurs because the sound device only works in 16 bit linear stereo sound at 44100 Hz, which is not the requested settings.
MMERR_GUS_TIMER
This error occurs when the ultra driver could not setup the playback timer.
4.3.4.5 HP-UX Driver Specific Errors
MMERR_HP_AUDIO_DESC
This error occurs when the HP driver can not get the audio hardware description.
MMERR_HP_AUDIO_OUTPUT
This error occurs when the HP driver can not select the audio output.
MMERR_HP_BUFFERSIZE
This error occurs when the HP driver can not set the transmission buffer size.
MMERR_HP_CHANNELS
This error occurs when the HP driver can not set the requested number of channels.
MMERR_HP_SETSAMPLESIZE
This error occurs when the HP driver can not set the requested sample size.
MMERR_HP_SETSPEED
This error occurs when the HP driver can not set the requested sample rate.
4.3.4.6 ALSA Driver Specific Errors
MMERR_ALSA_NOCONFIG
This error occurs when no ALSA playback configuration is available.
MMERR_ALSA_SETPARAMS
This error occurs when the ALSA driver can not set the requested sample format, sample rate, number of channels, access type, or latency values.
MMERR_ALSA_SETFORMAT
This error occurs when the ALSA driver can not set the requested sample format.
MMERR_ALSA_SETRATE
This error occurs when the ALSA driver does not support the requested sample rate.
MMERR_ALSA_SETCHANNELS
This error occurs when the ALSA driver does not support the requested number of channels.
MMERR_ALSA_BUFFERSIZE
This error occurs when the ALSA driver can not retrieve the buffer or period size.
MMERR_ALSA_PCM_START
This error occurs when the ALSA driver can not start the pcm playback.
MMERR_ALSA_PCM_WRITE
This error occurs when the ALSA driver encounters a write error.
MMERR_ALSA_PCM_RECOVER
This error occurs when the ALSA driver encounters an error recovery failure.
4.3.4.7 Open Sound System Driver Specific Errors
MMERR_OSS_SETFRAGMENT
This error occurs when the OSS driver can not set audio fragment size.
MMERR_OSS_SETSAMPLESIZE
This error occurs when the OSS driver can not set the requested sample size.
MMERR_OSS_SETSPEED
This error occurs when the OSS driver can not set the requested sample rate.
MMERR_OSS_SETSTEREO
This error occurs when the OSS driver can not set the requested number of channels.
4.3.4.8 SGI Driver Specific Errors
MMERR_SGI_MONO
This error occurs when the hardware only supports stereo sound.
MMERR_SGI_SPEED
This error occurs when the hardware does not support the requested sample rate.
MMERR_SGI_STEREO
This error occurs when the hardware only supports mono sound.
MMERR_SGI_16BIT
This error occurs when the hardware only supports 16 bit sound.
MMERR_SGI_8BIT
This error occurs when the hardware only supports 8 bit sound.
4.3.4.9 Sun Driver Specific Error
MMERR_SUN_INIT
This error occurs when the sound device initialization failed.
4.3.4.10 OS/2 Driver Specific Errors
MMERR_OS2_MIXSETUP
This error occurs when the DART driver can not set the mixing parameters.
MMERR_OS2_SEMAPHORE
This error occurs when the MMPM/2 driver can not create the semaphores needed for playback.
MMERR_OS2_THREAD
This error occurs when the MMPM/2 driver can not create the thread needed for playback.
MMERR_OS2_TIMER
This error occurs when the MMPM/2 driver can not create the timer needed for playback.
4.3.4.11 DirectX Driver Specific Errors
MMERR_DS_BUFFER
This error occurs when the DirectX driver can not allocate the playback buffers.
MMERR_DS_EVENT
This error occurs when the DirectX driver can not register the playback event.
MMERR_DS_FORMAT
This error occurs when the DirectX driver can not set the playback format.
MMERR_DS_NOTIFY
This error occurs when the DirectX driver can not register the playback callback.
MMERR_DS_PRIORITY
This error occurs when the DirectX driver can not set the playback priority.
MMERR_DS_THREAD
This error occurs when the DirectX driver can not create the playback thread.
MMERR_DS_UPDATE
This error occurs when the DirectX driver can not initialize the playback thread.
4.3.4.12 Windows Multimedia API Driver Specific Errors
MMERR_WINMM_ALLOCATED
This error occurs when the playback resource is already allocated by another application.
MMERR_WINMM_DEVICEID
This error occurs when the Multimedia API Driver is given an invalid audio device identificator.
MMERR_WINMM_FORMAT
This error occurs when the playback output format is not supported by the audio device.
MMERR_WINMM_HANDLE
This error occurs when the Multimedia API Driver is given an invalid handle.
MMERR_WINMM_UNKNOWN
This error should not occur ! If you get this error, please contact the libmikmod development mailing list.
4.3.4.13 MacOS Driver Specific Errors
MMERR_MAC_SPEED
This error occurs when the playback speed is not supported by the audio device.
MMERR_MAC_START
This error occurs when the MacOS driver can not start playback.


Next: , Previous: Error Reference, Up: Library Reference

4.4 Function Reference


Next: , Previous: Function Reference, Up: Function Reference

4.4.1 Library Core Functions

4.4.1.1 MikMod_Active

BOOL MikMod_Active(void)

Description
This function returns whether sound output is enabled or not.
Result
0
Sound output is disabled.
1
Sound output is enabled.
Notes
Calls to MikMod_Update will be ignored when sound output is disabled.
See also
MikMod_DisableOutput, MikMod_EnableOutput.
4.4.1.2 MikMod_DisableOutput

void MikMod_DisableOutput(void)

Description
This function stops the sound mixing.
Notes
Calls to MikMod_Update will be ignored when sound output is disabled.
See also
MikMod_Active, MikMod_EnableOutput.
4.4.1.3 MikMod_EnableOutput

int MikMod_EnableOutput(void)

Description
This function starts the sound mixing.
Result
0
Sound mixing is ready.
nonzero
An error occurred during the operation.
Notes
Calls to MikMod_Update will be ignored when sound output is disabled.
See also
MikMod_Active, MikMod_DisableOutput.
4.4.1.4 MikMod_Exit

void MikMod_Exit(void)

Description
This function deinitializes the sound hardware and frees all the memory and resources used by MikMod.
See also
MikMod_Init, MikMod_Reset.
4.4.1.5 MikMod_GetVersion

long MikMod_GetVersion(void)

Description
This function returns the version number of the library.
Result
The version number, encoded as follows: (maj<<16)|(min<<8)|(rev), where ‘maj’ is the major version number, ‘min’ is the minor version number, and ‘rev’ is the revision number.
4.4.1.6 MikMod_InfoDriver

CHAR* MikMod_InfoDriver(void)

Description
This function returns a formatted list of the registered drivers in a buffer.
Result
A pointer to a text buffer, or NULL if no drivers are registered.
Notes
The buffer is created with malloc; the caller must free it when it is no longer necessary.
See also
MikMod_RegisterDriver, MikMod_RegisterAllDrivers.
4.4.1.7 MikMod_InfoLoader

CHAR* MikMod_InfoLoader(void)

Description
This function returns a formatted list of the registered module loaders in a buffer.
Result
A pointer to a text buffer, or NULL if no loaders are registered.
Notes
The buffer is created with malloc; the caller must free it when it is no longer necessary.
See also
MikMod_RegisterLoader, MikMod_RegisterAllLoaders.
4.4.1.8 MikMod_Init

int MikMod_Init(CHAR *parameters)

Description
This function performs the library initialization, including the sound driver choice and configuration, and all the necessary memory allocations.
Parameters
parameters
Optional parameters given to the sound driver. These parameters are ignored if the value of md_device is zero (autodetection).
Result
0
Initialization was successful.
nonzero
An error occurred during initialization.
Notes
When the initialization fails, the library uses the nosound sound driver to let other the other MikMod functions work without crashing the application.
See also
MikMod_Exit, MikMod_InitThreads, MikMod_Reset.
4.4.1.9 MikMod_InitThreads

BOOL MikMod_InitThreads(void)

Description
This function returns whether libmikmod is thread-safe.
Result
0
The library is not thread-safe.
1
The library is thread-safe.
Notes
This function should be called before any call to MikMod_Lock or MikMod_Unlock is made.
See also
MikMod_Lock, MikMod_Unlock.
4.4.1.10 MikMod_Lock

void MikMod_Lock(void)

Description
This function obtains exclusive access to libmikmod's variables.
Notes
This function locks an internal mutex. If the mutex is already locked, it will block the calling thread until the mutex is unlocked.
Every MikMod_Unlock call should be associated to a MikMod_Lock call. To be sure this is the case, we advise you to define and use the following macros:
#define MIKMOD_LOCK MikMod_Lock();{
#define MIKMOD_UNLOCK }MikMod_Unlock();
The function MikMod_InitThreads must have been invoked before any call to MikMod_Lock in made.

See also
MikMod_InitThreads, MikMod_Unlock.
4.4.1.11 MikMod_RegisterAllDrivers

void MikMod_RegisterAllDrivers(void)

Description
This function registers all the available drivers.
See also
MikMod_InfoDriver, MikMod_RegisterDriver.
4.4.1.12 MikMod_RegisterAllLoaders

void MikMod_RegisterAllLoaders(void)

Description
This function registers all the available module loaders.
See also
MikMod_InfoLoader, MikMod_RegisterLoader.
4.4.1.13 MikMod_RegisterDriver

void MikMod_RegisterDriver(struct MDRIVER* newdriver)

Description
This function adds the specified driver to the internal list of usable drivers.
Parameters
newdriver
A pointer to the MDRIVER structure identifying the driver.
Notes
It is safe to register the same driver several times, although it will not be duplicated in the list.
You should register all the drivers you need before calling MikMod_Init. If you want to register all the available drivers, use MikMod_RegisterAllDrivers instead.
See also
MikMod_InfoDriver, MikMod_RegisterAllDrivers.
4.4.1.14 MikMod_RegisterErrorHandler

MikMod_handler_t MikMod_RegisterErrorHandler(MikMod_handler_t newhandler)

Description
This function selects the function which should be called in case of error.
Parameters
newhandler
The new error callback function.
Result
The previous error callback function, or NULL if there was none.
Notes
MikMod_handler_t is defined as void(*function)(void), this means your error function has the following prototype: void MyErrorHandler(void)
The error callback function is called when errors are detected, but not always immediately (the library has to resume to a stable state before calling your callback).
4.4.1.15 MikMod_RegisterLoader

void MikMod_RegisterLoader(struct MLOADER* newloader)

Description
This function adds the specified module loader to the internal list of usable module loaders.
Parameters
newloader
A pointer to the MLOADER structure identifying the loader.
Notes
It is safe to register the same loader several times, although it will not be duplicated in the list.
You should register all the loaders you need before calling Player_Load or Player_LoadFP. If you want to register all the available module loaders, use MikMod_RegisterAllLoaders instead.
The 15 instrument module loader (load_m15) should always be registered last.
See also
MikMod_InfoLoader, MikMod_RegisterAllLoaders.
4.4.1.16 MikMod_RegisterPlayer

MikMod_player_t MikMod_RegisterPlayer(MikMod_player_t newplayer)

Description
This function selects the function which should be used to process module playback.
Parameters
newplayer
The new playback function
Result
The previous playback function.
Notes
MikMod_player_t is defined as void(*function)(void), this means your player function has the following prototype: void MyPlayer(void)
The player function is called every module tick to process module playback. The rate at which the player function is called is controlled by the sound driver, and is computed by the following equation:
(bpm*50)/125 calls per second, which means every 125000/(bpm*50) milliseconds. The bpm value is the tempo of the module and can change from its initial value when requested by the module.
When changing the playback function, you should make sure that you chain to the default MikMod playback function, otherwise you won't get module sound anymore...
Example
              MikMod_player_t oldroutine;
          
              void MyPlayer(void)
              {
                  oldroutine();
                  /* your stuff here */
                  ...
              }
          
              main()
              {
                  ...
                  /* Register our player */
                  oldroutine = MikMod_RegisterPlayer(MyPlayer);
                  ...
              }
4.4.1.17 MikMod_Reset

int MikMod_Reset(CHAR *parameters)

Description
This function resets MikMod and reinitialize the sound hardware.
Parameters
parameters
Optional parameters given to the sound driver. If you set the value of md_device to zero (autodetect), these parameters are ignored.
Result
0
Reinitialization was successful.
nonzero
An error occurred during reinitialization.
Notes
Use this function when you have changed the global configuration variables: md_device and md_mixfreq, or one of the md_mode flags which require sound reinitialization. Sound playback will continue as soon as the driver is ready.
See also
MikMod_Exit, MikMod_Init.
4.4.1.18 MikMod_SetNumVoices

int MikMod_SetNumVoices(int musicvoices, int samplevoices)

Description
This function sets the number of mixed voices which can be used for music and sound effects playback.
Parameters
musicvoices
The number of voices to reserve for music playback.
samplevoices
The number of voices to reserve for sound effects.
Result
0
Initialization was successful.
nonzero
An error occurred during initialization.
Notes
A value of -1 for any of the parameters will retain the current number of reserved voices.
The maximum number of voices vary from driver to driver (hardware drivers often have a limit of 32 to 64 voices, whereas the software drivers handle 255 voices). If your settings exceed the driver's limit, they will be truncated.
See also
MikMod_Init, MikMod_Reset.
4.4.1.19 MikMod_Unlock

void MikMod_Unlock(void)

Description
This function relinquishes exclusive access to libmikmod's variables.
Notes
This function unlocks an internal mutex, so that other threads waiting for the lock can be resumed.
Every MikMod_Unlock call should be associated to a MikMod_Lock call. To be sure this is the case, we advise you to define and use the following macros:
#define MIKMOD_LOCK MikMod_Lock();{
#define MIKMOD_UNLOCK }MikMod_Unlock();
The function MikMod_InitThreads must have been invoked before any call to MikMod_Unlock in made.

See also
MikMod_InitThreads, MikMod_Lock.
4.4.1.20 MikMod_Update

void MikMod_Update(void)

Description
This routine should be called on a regular basis to update the sound.
Notes
The sound output buffer is filled each time this function is called; if you use a large buffer, you don't need to call this routine as frequently as with a smaller buffer, but you get a bigger shift between the sound being played and the reported state of the player, since the player is always a buffer ahead of the playback.
If you play low quality sound (for example, mono 8 bit 11kHz sound), you only need to call this routine a few times per second. However, for high quality sound (stereo 16 bit 44kHz), this rate increases to a few hundred times per second, but never more, due to the minimal buffer size constraint imposed to the sound drivers.
If you plan on modifying voice information with the Voice_xx functions, you should do this before calling MikMod_Update.
4.4.1.21 MikMod_strerror

char* MikMod_strerror(int errnum)

Description
This function associates a descriptive message to an error code.
Parameters
errnum
The MikMod error code.
Result
A pointer to a string describing the error.


Next: , Previous: Library Core Functions, Up: Function Reference

4.4.2 Module Player Functions

4.4.2.1 Player_Active

BOOL Player_Active(void)

Description
This function returns whether the module player is active or not.
Result
0
No module is playing.
nonzero
A module is currently playing.
Notes
This function will still report that the player is active if the playing module is paused.
See also
Player_Paused, Player_TogglePause, Player_Start, Player_Stop
4.4.2.2 Player_Free

void Player_Free(MODULE* module)

Description
This function stops the module if it is playing and unloads it from memory.
Parameters
module
The module to free.
See also
Player_Load, Player_LoadFP.
4.4.2.3 Player_GetChannelVoice

int Player_GetChannelVoice(UBYTE channel)

Description
This function determines the voice corresponding to the specified module channel.
Parameters
channel
The module channel to use.
Result
The number of the voice corresponding to the module channel.
Notes
If the module channel has NNAs, the number will correspond to the main channel voice.
See also
Voice_SetPanning, Voice_SetVolume, Player_Mute, Player_ToggleMute, Player_Unmute.
4.4.2.4 Player_GetModule

MODULE* Player_GetModule(void)

Description
This function returns the module currently being played.
Result
A pointer to the MODULE being played, or NULL if no module is playing.
See also
Player_Stop, Player_Start.
4.4.2.5 Player_Load

MODULE* Player_Load(CHAR* filename, int maxchan, BOOL curious)

Description
This function loads a music module.
Parameters
filename
The name of the module file.
maxchan
The maximum number of channels the song is allowed to request from the mixer.
curious
The curiosity level to use.
Result
A pointer to a MODULE structure, or NULL if an error occurs.
Notes
If the curiosity level is set to zero, the module will be loaded normally. However, if it is nonzero, the following things occur:
See also
Player_Free, Player_LoadFP, Player_LoadTitle, Player_LoadTitleFP, Player_Start.
4.4.2.6 Player_LoadFP

MODULE* Player_LoadFP(FILE* file, int maxchan, BOOL curious)

Description
This function loads a music module.
Parameters
file
An open file, at the position where the module starts.
maxchan
The maximum number of channels the song is allowed to request from the mixer.
curious
The curiosity level to use.
Result
A pointer to a MODULE structure, or NULL if an error occurs.
Notes
The file is left open, at the same position as before the function call.
If the curiosity level is set to zero, the module will be loaded normally. However, if it is nonzero, the following things occur:
See also
Player_Free, Player_Load, Player_LoadTitle, Player_LoadTitleFP, Player_Start.
4.4.2.7 Player_LoadTitle

MODULE* Player_LoadTitle(CHAR* filename)

Description
This function retrieves the title of a module file.
Parameters
filename
The name of the module file.
Result
A pointer to the song title, or NULL if either the module has no title or an error has occurred.
Notes
The title buffer is created with malloc; the caller must free it when it is no longer necessary.
See also
Player_Load, Player_LoadFP, Player_LoadTitleFP.
4.4.2.8 Player_LoadTitleFP

MODULE* Player_LoadTitleFP(FILE* file)

Description
This function retrieves the title of a module file.
Parameters
file
An open file, at the position where the module starts.
Result
A pointer to the song title, or NULL if either the module has no title or an error has occurred.
Notes
The title buffer is created with malloc; the caller must free it when it is no longer necessary.
See also
Player_Load, Player_LoadFP, Player_LoadTitle.
4.4.2.9 Player_Mute

void Player_Mute(SLONG operation, ...)

Description
This function mutes a single module channel, or a range of module channels.
Parameters
operation
Either the number of a module channel to mute (starting from zero), or an operation code. In the latter case, two extra parameters are needed to determine the range of channels.
Notes
If the operation is MUTE_INCLUSIVE, the two channel numbers delimit the range and are part of the range ; otherwise, if the operation is MUTE_EXCLUSIVE, they are outside of the range.
Example
              /* mute channel 10 */
              Player_Mute(10);
              /* mute channels 2 to 5 */
              Player_Mute(MUTE_INCLUSIVE, 2, 5);
              /* mute channels 7 to 9 */
              Player_Mute(MUTE_EXCLUSIVE, 6, 10);

See also
Player_Muted, Player_ToggleMute, Player_Unmute.
4.4.2.10 Player_Muted

BOOL Player_Muted(UBYTE channel)

Description
This function determines whether a module channel is muted or not.
Parameters
channel
The module channel to test (starting from zero).
Result
0
The channel is not muted.
1
The channel is muted.
See also
Player_Mute, Player_ToggleMute, Player_Unmute.
4.4.2.11 Player_NextPosition

void Player_NextPosition(void)

Description
This function jumps to the next position in the module.
Notes
All playing samples and active song voices are cut to avoid hanging notes.
See also
Player_PrevPosition, Player_SetPosition.
4.4.2.12 Player_Paused

BOOL Player_Paused(void)

Description
This function determines whether the module is paused or not.
Result
0
The module is not paused.
1
The module is paused.
See also
Player_TogglePause.
4.4.2.13 Player_PrevPosition

void Player_PrevPosition(void)

Description
This function jumps to the previous position in the module.
Notes
All playing samples and active song voices are cut to avoid hanging notes.
See also
Player_NextPosition, Player_SetPosition.
4.4.2.14 Player_SetPosition

void Player_SetPosition(UWORD position)

Description
This function jumps to the specified position in the module.
Parameters
position
The pattern position to jump to.
Notes
All playing samples and active song voices are cut to avoid hanging notes.
See also
Player_NextPosition, Player_PrevPosition.
4.4.2.15 Player_SetSpeed

void Player_SetSpeed(UWORD speed)

Description
This function sets the module speed.
Parameters
speed
The new module speed, in the range 1-32.
See also
Player_SetTempo.
4.4.2.16 Player_SetTempo

void Player_SetTempo(UWORD tempo)

Description
This function sets the module tempo.
Parameters
tempo
The new module tempo, in the range 32-255.
See also
Player_SetSpeed.
4.4.2.17 Player_SetVolume

void Player_SetVolume(SWORD volume)

Description
This function sets the module volume.
Parameters
volume
The new overall module playback volume, in the range 0-128.
4.4.2.18 Player_Start

void Player_Start(MODULE* module)

Description
This function starts the specified module playback.
Parameters
module
The module to play.
Notes
If another module is playing, it will be stopped and the new module will play.
See also
Player_Stop.
4.4.2.19 Player_Stop

void Player_Stop(void)

Description
This function stops the currently playing module.
See also
Player_Start.
4.4.2.20 Player_ToggleMute

void Player_ToggleMute(SLONG operation, ...)

Description
This function changes the muted status of a single module channel, or a range of module channels.
Parameters
operation
Either the number of a module channel to work on (starting from zero), or an operation code. In the latter case, two extra parameters are needed to determine the range of channels.
Notes
If the operation is MUTE_INCLUSIVE, the two channel numbers delimit the range and are part of the range ; otherwise, if the operation is MUTE_EXCLUSIVE, they are outside of the range.
Example
              /* toggle mute on channel 10 */
              Player_ToggleMute(10);
              /* toggle mute on channels 2 to 5 */
              Player_ToggleMute(MUTE_INCLUSIVE, 2, 5);
              /* toggle mute on channels 7 to 9 */
              Player_ToggleMute(MUTE_EXCLUSIVE, 6, 10);

See also
Player_Mute, Player_Muted, Player_Unmute.
4.4.2.21 Player_TogglePause

void Player_TogglePause(void)

Description
This function toggles the playing/paused status of the module.
Notes
Calls to Player_xx functions still have effect when the module is paused.
See also
Player_Paused, Player_Start, Player_Stop.
4.4.2.22 Player_Unmute

void Player_Unmute(SLONG operation, ...)

Description
This function unmutes a single module channel, or a range of module channels.
Parameters
operation
Either the number of a module channel to unmute (starting from zero), or an operation code. In the latter case, two extra parameters are needed to determine the range of channels.
Notes
If the operation is MUTE_INCLUSIVE, the two channel numbers delimit the range and are part of the range ; otherwise, if the operation is MUTE_EXCLUSIVE, they are outside of the range.
Example
              /* unmute channel 10 */
              Player_Unmute(10);
              /* unmute channels 2 to 5 */
              Player_Unmute(MUTE_INCLUSIVE, 2, 5);
              /* unmute channels 7 to 9 */
              Player_Unmute(MUTE_EXCLUSIVE, 6, 10);

See also
Player_Mute, Player_Muted, Player_ToggleMute.


Next: , Previous: Module Player Functions, Up: Function Reference

4.4.3 Sample Functions

4.4.3.1 Sample_Free

void Sample_Free(SAMPLE* sample)

Description
This function unloads a sample from memory.
Parameters
sample
The sample to free.
See also
Sample_Load, Sample_LoadFP.
4.4.3.2 Sample_Load

SAMPLE* Sample_Load(CHAR* filename)

Description
This function loads a sample.
Parameters
filename
The sample filename.
Result
A pointer to a SAMPLE structure, or NULL if an error has occurred.
See also
Sample_Free, Sample_LoadFP.
4.4.3.3 Sample_LoadFP

SAMPLE* Sample_LoadFP(FILE* file)

Description
This function loads a sample.
Parameters
file
An open file, at the position where the sample starts.
Result
A pointer to a SAMPLE structure, or NULL if an error has occurred.
Notes
The file is left open, at the same position as before the function call.
See also
Sample_Free, Sample_Load.
4.4.3.4 Sample_Play

SBYTE Sample_Play(SAMPLE* sample, ULONG start, UBYTE flags)

Description
This function plays a sample as a sound effect.
Parameters
sample
The sample to play.
start
The starting position (in samples).
flags
Either zero, for normal sound effects, or SFX_CRITICAL, for critical sound effects which must not be interrupted.
Result
The voice number corresponding to the voice which will play the sample.
Notes
Each new sound effect is played on a new voice. When all voices are taken, the oldest sample which was not marked as critical is cut and its voice is used for the new sample. Critical samples are not cut unless all the voices are taken with critical samples and you attempt to play yet another critical sample. Use Voice_Stop to force the end of a critical sample.
See also
MikMod_SetNumVoices, Voice_Play, Voice_SetFrequency, Voice_SetPanning, Voice_SetVolume, Voice_Stop.


Next: , Previous: Sample Functions, Up: Function Reference

4.4.4 Voice Functions

4.4.4.1 Voice_GetFrequency

ULONG Voice_GetFrequency(SBYTE voice)

Description
This function returns the frequency of the sample currently playing on the specified voice.
Parameters
voice
The number of the voice to get frequency.
Result
The current frequency of the sample playing on the specified voice, or zero if no sample is currently playing on the voice.
See also
Voice_SetFrequency.
4.4.4.2 Voice_GetPanning

ULONG Voice_GetPanning(SBYTE voice)

Description
This function returns the panning position of the sample currently playing on the specified voice.
Parameters
voice
The number of the voice to get panning position.
Result
The current panning position of the sample playing on the specified voice, or PAN_CENTER if no sample is currently playing on the voice.
See also
Voice_SetPanning.
4.4.4.3 Voice_GetPosition

SLONG Voice_GetPosition(SBYTE voice)

Description
This function returns the sample position (in samples) of the sample currently playing on the specified voice.
Parameters
voice
The number of the voice to get sample position (starting from zero).
Result
The current play location of the sample playing on the specified voice, or zero if the position can not be determined or if no sample is currently playing on the voice.
Notes
This function may not work with some drivers (especially for hardware mixed voices). In this case, it returns always -1.
See also
Sample_Play, Voice_Play.
4.4.4.4 Voice_GetVolume

UWORD Voice_GetVolume(SBYTE voice)

Description
This function returns the volume of the sample currently playing on the specified voice.
Parameters
voice
The number of the voice to get volume.
Result
The current volume of the sample playing on the specified voice, or zero if no sample is currently playing on the voice.
See also
Voice_RealVolume, Voice_SetVolume.
4.4.4.5 Voice_Play

void Voice_Play(SBYTE voice, SAMPLE* sample, ULONG start)

Description
Start a new sample in the specified voice.
Parameters
voice
The number of the voice to be processed (starting from zero).
sample
The sample to play.
start
The starting position (in samples).
Notes
The sample will be played at the volume, panning and frequency settings of the voice, regardless or the sample characteristics.
The sample played this way gets the same “critical” status as the sample which was previously played on this voice.
See also
Sample_Play, Voice_SetFrequency, Voice_SetPanning, Voice_SetVolume.
4.4.4.6 Voice_RealVolume

ULONG Voice_RealVolume(SBYTE voice)

Description
This function returns the actual playing volume of the specified voice.
Parameters
voice
The number of the voice to analyze (starting from zero).
Result
The real volume of the voice when the function was called, in the range 0-65535, not related to the volume constraint specified with Voice_SetVolume.
Notes
This function may not work with some drivers (especially for hardware mixed voices). In this case, it always returns zero.
Also note that the real volume computation is not a trivial process and takes some CPU time.
See also
Sample_Play, Voice_GetVolume, Voice_Play, Voice_SetVolume.
4.4.4.7 Voice_SetFrequency

void Voice_SetFrequency(SBYTE voice, ULONG frequency)

Description
This function sets the frequency (pitch) of the specified voice.
Parameters
voice
The number of the voice to be processed (starting from zero).
frequency
The new frequency of the voice, in hertz.
See also
Sample_Play, Voice_GetFrequency, Voice_Play, Voice_SetPanning, Voice_SetVolume, Voice_Stop.
4.4.4.8 Voice_SetPanning

void Voice_SetPanning(SBYTE voice, ULONG panning)

Description
This function sets the panning position of the specified voice.
Parameters
voice
The number of the voice to be processed (starting from zero).
panning
The new panning position of the voice.
Notes
Panning can vary between 0 (PAN_LEFT) and 255 (PAN_RIGHT). Center is 127 (PAN_CENTER. Surround sound can be enabled by specifying the special value PAN_SURROUND.
See also
Sample_Play, Voice_GetPanning, Voice_Play, Voice_SetFrequency, Voice_SetVolume, Voice_Stop.
4.4.4.9 Voice_SetVolume

void Voice_SetVolume(SBYTE voice, UWORD volume)

Description
This function sets the volume of the specified voice.
Parameters
voice
The number of the voice to be processed (starting from zero).
volume
The new volume of the voice, in the range 0-256.
See also
Sample_Play, Voice_GetVolume, Voice_Play, Voice_SetFrequency, Voice_SetPanning, Voice_Stop.
4.4.4.10 Voice_Stop

void Voice_Stop(SBYTE voice)

Description
This function stops the playing sample of the specified voice.
Parameters
voice
The number of the voice to be processed (starting from zero).
Notes
After the call to Voice_Stop, the function Voice_Stopped will return nonzero (true) for the voice. If you want to silence the voice without stopping the playback, use Voice_SetVolume(voice, 0) instead.
See also
Sample_Play, Voice_Play, Voice_SetFrequency, Voice_SetPanning, Voice_SetVolume.
4.4.4.11 Voice_Stopped

BOOL Voice_Stopped(SBYTE voice)

Description
This function returns whether the voice is active or not.
Parameters
voice
The number of the voice to be checked (starting from zero).
Result
0
The voice is stopped or has no sample assigned.
nonzero
The voice is playing a sample.
Notes
This function may not work with some drivers (especially for hardware mixed voices). In this case, its return value is undefined.
See also
Voice_Stop.


Next: , Previous: Voice Functions, Up: Library Reference

4.5 Loader Reference


Next: , Previous: Loader Reference, Up: Loader Reference

4.5.1 Module Loaders

MikMod presents a large choice of module loaders, for the most common formats as well as for some less-known exotic formats.

load_669
This loader recognizes “Composer 669” and “Unis 669” modules. The 669 and “Extended 669” formats were among the first PC module formats. They do not have a wide range of effects, but support up to 32 channels.
“Composer 669” was written by Tran of Renaissance, a.k.a. Tomasz Pytel and released in 1992. “Unis 669 Composer” was written by Jason Nunn and released in 1994.
load_amf
This loader recognizes the “Advanced Module Format”, which is the internal module format of the “DOS Sound and Music Interface” (DSMI) library. This format has the same limitations as the S3M format. The most famous DSMI application was DMP, the Dual Module Player.
DMP and the DSMI library were written by Otto Chrons. DSMI was first released in 1993.
load_dsm
This loader recognizes the internal DSIK format, which is the internal module format of the “Digital Sound Interface Kit” (DSIK) library, the ancester of the SEAL library. This format has the same limitations as the S3M format.
The DSIK library was written by Carlos Hasan and released in 1994.
load_far
This loader recognizes “Farandole” modules. These modules can be up to 16 channels and have Protracker comparable effects.
The Farandole composer was written by Daniel Potter and released in 1994.
load_gdm
This loader recognizes the “General DigiMusic” format, which is the internal format of the “Bells, Whistles and Sound Boards” library. This format has the same limitations as the S3M format.
The BWSB library was written by Edward Schlunder and first released in 1993.
load_imf
This loader recognizes “Imago Orpheus” modules. This format is roughly equivalent to the XM format, but with two effects columns instead of a volume column and an effect column.
Imago Orpheus was written by Lutz Roeder and released in 1994.
load_it
This loader recognizes “Impulse Tracker” modules, currently the most powerful format. These modules support up to 64 real channels, and up to 256 virtual channels with the “New Note Action” feature. Besides, it has the widest range of effects, and supports 16 bit samples as well as surround sound.
“Impulse Tracker” was written by Jeffrey Lim and released in 1996.
load_med
This loader recognizes “OctaMED” modules. These modules are comparable to Protracker modules, but can embed “synthsounds”, which are midi-like instruments.
“MED” and later “OctaMED” were written by Teijo Kinnunen. “MED” was released in 1989, and “OctaMED” was released in 1992.
load_m15
This loader recognizes the old 15 instrument modules, created by “Ultimate Soundtracker”, “Soundtracker” and the first versions of “Protracker”.
Since this format was one of the first module formats, developed in 1987, it does not have any signature field, which makes it hard to detect reliably, because of its similarities with later module formats.
load_mod
This loader recognizes the standard 31 instrument modules, created by “Protracker” or Protracker-compatible programs. The original Protracker format was limited to 4 channels, but other trackers like “TakeTracker”, “StarTracker” or “Oktalyzer” afforded more channels.
Although it is now technically obsolete, this format is still widely used, due to its playback simplicity (on the adequate hardware, the Amiga).
load_mtm
This loader recognizes the “MultiTracker Module Editor” modules. The MTM format has up to 32 channels, and protracker comparable effects. It was intended to replace “Composer 669”. The “MultiTracker Module Editor” was written by Starscream of Renaissance, a.k.a. Daniel Goldstein and released in late 1993.
load_okt
This loader recognizes the “Amiga Oktalyzer” modules. The OKT format has up to 8 channels, and a few protracker compatible effects, as well as other OKT-specific effects, of which only a few are currently supported by libmikmod. “Oktalyzer” was written by Armin Sander and released in 1990.
load_stm
This loader recognizes “ScreamTracker” modules. “ScreamTracker” was the first PC tracker, as well as the first PC module format. Loosely inspired by the “SoundTracker” format, it does not have as many effects as Protracker, although it supports 31 instruments and 4 channels.
“ScreamTracker” was written by PSI of Future Crew, a.k.a. Sami Tammilehto.
load_stx
This loader recognizes “STMIK 0.2” modules. “STMIK” (the Scream Tracker Music Interface Kit) was a module playing library distributed by Future Crew to play Scream Tracker module in games and demos. It uses an intermediate format between STM and S3M and comes with a tool converting STM modules to STX.
“STMIK” was written by PSI of Future Crew, a.k.a. Sami Tammilehto.
load_s3m
This loader recognizes “ScreamTracker 3” modules. This version was a huge improvement over the original “ScreamTracker”. It supported 32 channels, up to 99 instruments, and a large choice of effects.
“ScreamTracker 3” was written by PSI of Future Crew, a.k.a. Sami Tammilehto, and released in 1994.
load_ult
This loader recognizes “UltraTracker” modules. They are mostly similar to Protracker modules, but support two effects per channel.
“UltraTracker” was written by MAS of Prophecy, a.k.a. Marc Andre Schallehn, and released in 1993.
load_uni
This loader recognizes “UNIMOD” modules. This is the internal format used by MikMod and APlayer. Use of this format is discouraged, this loader being provided for completeness.
load_xm
This loader recognizes “FastTracker 2” modules. This format was designed from scratch, instead of creating yet another Protracker variation. It was the first format using instruments as well as samples, and envelopes for finer effects.
FastTracker 2 was written by Fredrik Huss and Magnus Hogdahl, and released in 1994.


Next: , Previous: Module Loaders, Up: Loader Reference

4.5.2 Sample Loaders

Currently, the only file type than can be loaded as a sample is the RIFF WAVE file. Stereo or compressed WAVE files are not supported yet.


Next: , Previous: Sample Loaders, Up: Library Reference

4.6 Driver Reference


Next: , Previous: Driver Reference, Up: Driver Reference

4.6.1 Network Drivers

These drivers send the generated sound over the network to a server program, which sends the sound to the real sound hardware. The server program can be on the same machine than your program, but MikMod does not have access to the hardware. Network drivers only support software mixing.

drv_AF
This driver works with the “Digital AudioFile” library.
Start the server on the machine you want, set its hostname in the ‘AUDIOFILE’ environment variable, and MikMod is ready to send it sound.
drv_esd
This driver works with the “Enlightened Sound Daemon”.
Start the esd daemon on the machine you want, set its hostname in the ‘ESPEAKER’ environment variable, and MikMod is ready to send it sound.


Next: , Previous: Network Drivers, Up: Driver Reference

4.6.2 Hardware Drivers

These drivers access to the sound hardware of the machine they run on. Depending on your Unix flavor, you'll end with one or more drivers from this list:

drv_aix
This driver is only available under AIX, and access its audio device.
This driver only supports software mixing.
drv_alsa
This driver is only available under Linux, and requires the ALSA driver to be compiled for your current kernel.
This driver only supports software mixing, but a future version of the driver might be able to use the hardware capabilities of some sound cards.
drv_dart
This driver is only available under OS/2 version 3 and higher (Warp), and uses the “Direct Audio Real-Time” interface.
This driver only supports software mixing.
drv_ds
This driver is only available under WIN32, and uses the “DirectSound” api.
This driver only supports software mixing.
drv_hp
This driver is only available under HP-UX, and access its audio device.
This driver only supports software mixing.
drv_os2
This driver is only available under OS/2 version 3 and higher (Warp), and OS/2 2.x with MMPM/2.
This driver only supports software mixing.
drv_oss
This driver is available under any Unix with the Open Sound System drivers installed. Linux and FreeBSD also come with the OSS/Lite driver (the non-commercial version of OSS) and can make use of this driver.
This driver only supports software mixing.
drv_sam9407
This driver is only available under Linux, and requires the Linux sam9407 driver to be compiled for your current kernel.
This driver only supports hardware mixing.
drv_sgi
This driver is only available under IRIX, and uses the SGI audio library.
This driver only supports software mixing.
drv_sndio
This driver is only available under OpenBSD and uses the “sndio” interface.
This driver only supports software mixing.
drv_sun
This driver is only available under Unices which implement SunOS-like audio device interfaces, that is, SunOS, Solaris, NetBSD and OpenBSD.
This driver only supports software mixing.
drv_ultra
This driver is only available under Linux, and requires the Linux Ultrasound driver (the ancestor of ALSA) to be compiled for your current kernel.
This driver only supports hardware mixing.
drv_win
This driver is only available under WIN32, and uses the “multimedia API”.
This driver only supports software mixing.


Next: , Previous: Hardware Drivers, Up: Driver Reference

4.6.3 Disk Writer Drivers

These drivers work on any machine, since the generated sound is not sent to hardware, but written in a file. Disk writer drivers only support software mixing.

drv_raw
This driver outputs the sound data in a file by default named music.raw in the current directory. The file has no header and only contains the sound output.
drv_wav
This driver outputs the sound data in a RIFF WAVE file by default named music.wav in the current directory.


Previous: Disk Writer Drivers, Up: Driver Reference

4.6.4 Other Drivers

These drivers are of little interest, but are handy sometimes.

drv_stdout
This driver outputs the sound data to the program's standard output. To avoid inconvenience, the data will not be output if the standard output is a terminal, thus you have to pipe it through another command or to redirect it to a file. Using this driver and redirecting to a file is equivalent to using the drv_raw disk writer.
This driver only supports software mixing.
drv_pipe
This driver pipes the sound data to a command (which must be given in the driver commandline, via MikMod_Init).
This driver only supports software mixing.
drv_nos
This driver doesn't produce sound at all, and will work on any machine.
Since it does not have to produce sound, it supports both hardware and software mixing, with as many hardware voices as you like.


Next: , Previous: Library Reference, Up: Top

5 Index


Next: , Previous: Index, Up: Index

Function Index


Previous: Function Index, Up: Index

Type and Variable Index

Table of Contents


Footnotes

[1] PAN_SURROUND will be mapped to PAN_CENTER if the library is initialized without surround sound, that is, if the variable md_mode doesn't have the bit DMODE_SURROUND set.

[2] Unless you explicitely choose to create a non thread-safe version of libmikmod at compile-time.

[3] You can force libmikmod to load the module (without the synthsounds, of course) by setting the curious parameter to 1 when invoking Player_Loadxx.