[ot_207]

Originally Posted by Arkshine I have noticed too, but I was not sure because the content send a SVC_SPAWNSTATICSOUND message. I thought it was for others things.

I have tried testing all the message modes SVC_SOUND SVC_SOUNDFADE SVC_STOPSOUND and it seems from what I have read from the code these messages can not be sent through message_begin functions, but can be sent through MSG_WRITEBYTE(&datagram, SVC_SOUND), this means that they are differently transmitted, does SPAWNSTATICSOUND respond good? Has anyone tested it? From what I have studied it is almost impossible to send a sound only to a client.

Edit: And about SV_MultiCast it seems that it is the function that selects which player gets the sound message, unfortunately it does not have structs that Orpheu does support.

Am I right? Or am I following a stupid path?

[joropito]

OT, are you starting the message with SVC_TEMPENTITY type, then send SVC_SPAWNSTATICSOUND as the first byte?

[ot_207]

Originally Posted by joropito OT, are you starting the message with SVC_TEMPENTITY type, then send SVC_SPAWNSTATICSOUND as the first byte?

Nope, I am using something like this:

PHP Code:

message_begin(MSG_ONE, SVC_SPAWNSTATICSOUND, _, id)
// ETC
message_end() 


Does it work with tempentity?

[Arkshine]

What I don't understand is, in the decompiled code, the sound is not passed into the message.

[ot_207]

Originally Posted by Arkshine What I don't understand is, in the decompiled code, the sound is not passed into the message.

It is passed its precache number, it is used for smaller bandwidth usage.

[Arkshine]

Yes, you're right, the index is passed. ^^

[ot_207]

Originally Posted by Arkshine Yes, you're right, the index is passed. ^^

Unfortunately the info we just found gets us nowhere. I have tested all the possible and impossible methods to send a sound to client and what I have found:
CS does not accept SVC_SOUND SVC_SOUNDFADE SVC_STOPSOUND SVC_SPAWNSTATICSOUND if they are sent through message_begin(), when I run one of these 4 it disconnects and gives the error.
However if I type in console the command cl_messages
I get something like this:

-------- Message Load --------- 1:svc_nop: 688 msgs:0.67K 5:svc_setview: 8 msgs:0.02K 6:svc_sound: 234 msgs:2.78K 7:svc_time: 44467 msgs:217.12K 8:svc_print: 5 msgs:0.18K 9:svc_stufftext: 16 msgs:0.44K 10:svc_setangle: 11 msgs:0.08K 11:svc_serverinfo: 2 msgs:0.49K 12:svc_lightstyle: 128 msgs:1.02K 13:svc_updateuserinfo: 101 msgs:17.12K 14:svc_deltadescription: 14 msgs:11.77K 15:svc_clientdata: 44467 msgs:527.07K 17:svc_pings: 1983 msgs:50.26K 21:svc_event_reliable: 8 msgs:0.02K 22:svc_spawnbaseline: 2 msgs:4.47K 23:svc_temp_entity: 2290 msgs:146.69K 25:svc_signonnum: 2 msgs:0.00K 29:svc_spawnstaticsound: 6384 msgs:93.52K 32:svc_cdtrack: 2 msgs:0.01K 35:svc_weaponanim: 9 msgs:0.03K 37:svc_roomtype: 2 msgs:0.01K 39:svc_newusermsg: 164 msgs.04K 40:svc_packetentities: 5 msgs:0.39K 41:svc_deltapacketentities: 44460 msgs:2602.64K 42:svc_choke: 79 msgs:0.08K 43:svc_resourcelist: 2 msgs9.83K 44:svc_newmovevars: 5 msgs:0.50K 45:svc_resourcerequest: 2 msgs:0.02K 46:svc_customization: 2 msgs:0.08K 52:svc_voiceinit: 2 msgs:0.03K 54:svc_sendextrainfo: 2 msgs:0.01K User messages: 2666:19K

That means that there is another way that these sounds are transmited and we do not have access to it. swds.dll is the way to make this work, and I think that SV_Multicast is the function responsible for selecting the clients who receive the messages.

Edit: Also about the cvars that are changed we also have this messages in console:

25296 0027 svc_clientdata 25296 0055 svc_deltapacketentities 25296 0198 svc_sound 25296 0211 svc_sound 25296 0224 svc_sound 25296 0237 svc_sound 25296 0250 svc_temp_entity 25296 0263 svc_sound 25296 0275 svc_temp_entity 25302 0008 AmmoPickup 25302 0011 AmmoPickup 25302 0014 AmmoPickup 25302 0017 AmmoPickup 25302 0020 Health 25302 0022 Damage 25302 0035 Damage 25302 0048 svc_customization 25302 0088 svc_updateuserinfo 25302 0261 svc_time 25302 0266 svc_choke 25302 0267 svc_clientdata 25302 0312svc_deltapacketentities 25302 0445 svc_sound 25302 0457 svc_sound 25302 0470 svc_temp_entity 25302 0542 svc_sound 25302 0554 svc_sound 25311 0008 TeamInfo 25311 0016 SelAmmo 25311 0021 GameTitle 25311 0023 svc_sendcvarvalue

It means that we have another way that these are transmited...

Edit2: Also SVC_DELTAPACKETENTITIES is sent through a different channel.

Edit3:
I have got some code from quake engine so you can see some info regarding sound transmition.

PHP Code:

/*
=================
SV_Multicast

Sends the contents of sv.multicast to a subset of the clients,
then clears sv.multicast.

MULTICAST_ALL    same as broadcast
MULTICAST_PVS    send to clients potentially visible from org
MULTICAST_PHS    send to clients potentially hearable from org
=================
*/
void SV_Multicast (vec3_t origin, int to)
{
    client_t    *client;
    byte        *mask;
    mleaf_t        *leaf;
    int            leafnum;
    int            j;
    qboolean    reliable;

    leaf = Mod_PointInLeaf (origin, sv.worldmodel);
    if (!leaf)
        leafnum = 0;
    else
        leafnum = leaf - sv.worldmodel->leafs;

    reliable = qfalse;

    switch (to)
    {
    case MULTICAST_ALL_R:
        reliable = qtrue;    // intentional fallthrough
    case MULTICAST_ALL:
        mask = sv.pvs;        // leaf 0 is everything;
        break;

    case MULTICAST_PHS_R:
        reliable = qtrue;    // intentional fallthrough
    case MULTICAST_PHS:
        mask = sv.phs + leafnum * 4*((sv.worldmodel->numleafs+31)>>5);
        break;

    case MULTICAST_PVS_R:
        reliable = qtrue;    // intentional fallthrough
    case MULTICAST_PVS:
        mask = sv.pvs + leafnum * 4*((sv.worldmodel->numleafs+31)>>5);
        break;

    default:
        mask = NULL;
        SV_Error ("SV_Multicast: bad to:%i", to);
    }

    // send the data to all relevent clients
    for (j = 0, client = svs.clients; j < MAX_CLIENTS; j++, client++)
    {
        if (client->state != cs_spawned)
            continue;

        if (to == MULTICAST_PHS_R || to == MULTICAST_PHS) {
            vec3_t delta;
            VectorSubtract(origin, client->edict->v.origin, delta);
            if (Length(delta) <= 1024)
                goto inrange;
        }

        leaf = Mod_PointInLeaf (client->edict->v.origin, sv.worldmodel);
        if (leaf)
        {
            // -1 is because pvs rows are 1 based, not 0 based like leafs
            leafnum = leaf - sv.worldmodel->leafs - 1;
            if ( !(mask[leafnum>>3] & (1<<(leafnum&7)) ) )
            {
//                Con_Printf ("supressed multicast\n");
                continue;
            }
        }

inrange:
        if (reliable) {
            ClientReliableCheckBlock(client, sv.multicast.cursize);
            ClientReliableWrite_SZ(client, sv.multicast.data, sv.multicast.cursize);
        } else
            SZ_Write (&client->datagram, sv.multicast.data, sv.multicast.cursize);
    }

    SZ_Clear (&sv.multicast);
}


/*  
==================
SV_StartSound

Each entity can have eight independant sound sources, like voice,
weapon, feet, etc.

Channel 0 is an auto-allocate channel, the others override anything
allready running on that entity/channel pair.

An attenuation of 0 will play full volume everywhere in the level.
Larger attenuations will drop off.  (max 4 attenuation)

==================
*/  
void SV_StartSound (edict_t *entity, int channel, char *sample, int volume,
    float attenuation)
{       
    int         sound_num;
    int            field_mask;
    int            i;
    int            ent;
    vec3_t        origin;
    qboolean    use_phs;
    qboolean    reliable = qfalse;

    if (volume < 0 || volume > 255)
        SV_Error ("SV_StartSound: volume = %i", volume);

    if (attenuation < 0 || attenuation > 4)
        SV_Error ("SV_StartSound: attenuation = %f", attenuation);

    if (channel < 0 || channel > 15)
        SV_Error ("SV_StartSound: channel = %i", channel);

// find precache number for sound
    for (sound_num=1 ; sound_num<MAX_SOUNDS
        && sv.sound_precache[sound_num] ; sound_num++)
        if (!strcmp(sample, sv.sound_precache[sound_num]))
            break;
    
    if ( sound_num == MAX_SOUNDS || !sv.sound_precache[sound_num] )
    {
        Con_Printf ("SV_StartSound: %s not precacheed\n", sample);
        return;
    }
    
    ent = NUM_FOR_EDICT(entity);

    if ((channel & 8) || !sv_phs.value)    // no PHS flag
    {
        if (channel & 8)
            reliable = qtrue; // sounds that break the phs are reliable
        use_phs = qfalse;
        channel &= 7;
    }
    else
        use_phs = qtrue;

//    if (channel == CHAN_BODY || channel == CHAN_VOICE)
//        reliable = qtrue;

    channel = (ent<<3) | channel;

    field_mask = 0;
    if (volume != DEFAULT_SOUND_PACKET_VOLUME)
        channel |= SND_VOLUME;
    if (attenuation != DEFAULT_SOUND_PACKET_ATTENUATION)
        channel |= SND_ATTENUATION;

    // use the entity origin unless it is a bmodel
    if (entity->v.solid == SOLID_BSP)
    {
        for (i=0 ; i<3 ; i++)
            origin[i] = entity->v.origin[i]+0.5*(entity->v.mins[i]+entity->v.maxs[i]);
    }
    else
    {
        VectorCopy (entity->v.origin, origin);
    }

    MSG_WriteByte (&sv.multicast, svc_sound);
    MSG_WriteShort (&sv.multicast, channel);
    if (channel & SND_VOLUME)
        MSG_WriteByte (&sv.multicast, volume);
    if (channel & SND_ATTENUATION)
        MSG_WriteByte (&sv.multicast, attenuation*64);
    MSG_WriteByte (&sv.multicast, sound_num);
    for (i=0 ; i<3 ; i++)
        MSG_WriteCoord (&sv.multicast, origin[i]);

    if (use_phs)
        SV_Multicast (origin, reliable ? MULTICAST_PHS_R : MULTICAST_PHS);
    else
        SV_Multicast (origin, reliable ? MULTICAST_ALL_R : MULTICAST_ALL);
} 


[joropito]

Maybe this can help

http://forums.alliedmods.net/showpos...14&postcount=5

[padilha007]

with orpheus is possible remove voice icon above the head?

[Arkshine]

It's handled client-side, I doubt you can.