SEGA DTS document on sound development
Contents
1. firing Refer to ST 068 for a more detailed description of its operation If you are new to patch editing please read Section 7 and Section 8 of this document Getting Started The TONE folder contains versions 2 07 2 06 and 1 13E of the Tone Editor For your reference version 1 13E is still included since many users have found it to be more stable than 2 06 Wave Editor A sample editor which should used as a final step before importing the sample into the Tone Editor It is suggested that you use a Third party sample editor for more versatility and editing power Refer to ST 99 for use of this application note make sure the loop start and end points are set on even numbers The loop end must be set to the end of the sample even if it is a non looping sample Linker Object oriented DSP programming application It allows to click and drag DSP routines and to direct those DSP algorithms to outputs of your choise For 3 D sound DSP like QSound or YSound by Yamaha you must use eLinker not included on this disk Contact your Third Party Account Executive to obtain the eLinker Please refer to ST 69 for instructions on how to use this application For more detail specifications on the DSP routines see ST 70 11 11 1996 Sound Driver Version 1 30 and up i SoundSimulator Version 2 0 and up Program Debugger Version 1 0 is 68000 Assembler Environment 1 14 Tone Editor Version 2 0 and up Wave Editor Version 1 13 DSP Link2. the map NEW CHOICE TONE bin and CHOICE TUNE more EXB Don t do anything with the space for DSP work RAM Choose SAVE from the File menu The map is now saved with it s new paths From now on when you open it the files will automatically be loaded 4 Select DOWNLOAD SYSTEM from the File menu 5 Select STARTUP SYSTEM from the File menu 6 Select MAKE CURRENT from the File menu 7 Select SOUND SIMULATOR from the Option menu The Sound Simulator window should open 8 Click on the blank box to the right of EFFECT CHANGE A file name with the extension EXB should appear in the box 9 Once the name of the EXB file is in the box click on EFFECT CHANGE 10 Click on the blank box to the right of MIXER CHANGE A file name with the extension bin should appear in the box 11 Once the name of the bin file is in the box click on MIXER CHANGE 12 Click on SEQUENCE 0 so that it is highlighted 13 Click on START The sequence should start playing 14 To stop the sequence click on STOP while SEQUENCE 0 is highlighted Running a MIDI sequencer At this point you should be able to boot up a MIDI sequencing program and with MIDI thru enabled play the sounds from the Saturn with your midi controller e The Tone Application General Description This is where you construct voices using AIFF samples map them to MIDI note numbers set output levels set DSP effects sends and returns It is similar to patch editors for multi timbral synthesizer
3. to get of now but very nice or the new Cross Products Midi Modified Saturns not as nice but cheaper and smaller You cannot run the Sound Simulator or generate DSP EXB files you have to generate the map files by hand which is not very difficult Using Tone Editor 2 06 In this example we will generate a very simple tone bank containing 1 voice that in it contains 1 layer with a non looping sample that triggers on only one note 1 Load up Tone Editor 2 Click OK on Target Board Not Connect or any other dialogue box for that matter 3 Click Cancel on File Requester that appears 4 Select File New menu 5 Requester appears asking for number of Voices So enter the number of voices you want in the tone bank For this example we will enter 1 6 Window called untitled will appear With No Voice Name Bend Range Portamento Vol Bias Play Mode and Size Ignore pratically all of these Enter a name in the Voice Name area only for your own reference say Voice 1 7 Double Click on 0 of No area 8 New requester appears asking for the number of Layers to be inside this voice For this example enter 1 9 New window opens called Voice 1 containing No Layer Name Start End Direct Level Effect Send etc Lets name our new Layer Layer 1 in the same way as our Voice 10 Click on Start Box here is the lowest note value that our layer will extend to We will enter 60 C4 ie Note C in octave 4 11 Cl
4. Don t touch any other boxes for the moment 22 Okay Save your file Also select File SaveBinFile Yeah We ve created a tone bank Creating a Map File Reference page 12 Sound Driver System Interface Manual Map contains the following information 1 Data Type ie Tone Bank Sequence DSP Prog DSP WorkARea 2 Unique identifier within each data Type 3 Start Address actually an offset in sound Ram 4 Data Size See reference for bit register assign Note 1 Lowest Start Address BOOOH 2 Highest data address 7FFFFH 3 That makes roughly 474K of data space Remember that sound driver uses area from 0 B000 4 Finish map off with Oxffff Oxffff Map for our example char map 0x00 0x00 OxBO 0x00 0x00 Ox0s Oxss Oxss Oxff Oxff Oxff Oxff lt gt lt gt lt a gt Tone Data Start Size of Map Terminator 0 first Address Tone Bank Triggering it from Code Principles 1 MIDI CHANNEL There are 32 logical channels each channel can be configured to play back separate voices with its own volume pan pitch bend settings and can be reconfigured at any point Each channel can play many notes as long as they are of the same voice configuration 2 SCSP SLOTS you don t have to worry about these These are the actually physical slots that the sound chip has each one can generate one sound at a time there are 32 of these These slots are assigned by the sou
5. Editor User s Manual Tone Editor User s Manual Addendum File Format SCSP DSP Effect Module Specifications DSP Linker User s Manual dAsms User s Manual Parameter Editor User s Manual Saturn Sound Tools Manual Supplement SATURN Sound Driver System Interface Version 3 03 Standard MIDI File Converter Specifications Sound Programming Debugger User s Manual Microcomputing Developing Int Environment for Macintosh NEW Saturn Sound Driver Implementation Manual ST 081 R5 062894 ST 168 R3 011895 ST 099 R1 042594 ST 067 121593 ST 068 R1 042594 ST 235 030795 ST 069 121693 ST 070 R1 031094 ST 228 R1 030595 ST 227 R1 030595 ST 198 R1 121594 ST 166 R4 012395 ST 066 121593 ST 065 R1 031494 ST 080 R2 050994 ST 241 042795 Sound for the Saturn is best dealt with by using the Saturn Custom Sound Processor SCSP because it has the versatility of standard midi commands The tools allow the sound designer to set up the memory map and how it is utilized It uses Standard MIDI files to trigger AIFF files that have been collected by the Tone Editor It can also trigger these samples For further assistance contact SEGA DTS by emailing dts sega com
6. SOUND pdf SEGA DTS This document contains information on the following 1 Overview amp Release Notes of the Sega Saturn Sound Development Tools a Sound Simulator amp Sound Driver b Tone Editor c Wave Editor d Linker 2 Cartdev Compatibility a Important Notes 3 Conversion Tools a AIFF to RedBook MAC b ADPCM Encoder Tool Demo to test Saturn Sound Box amp MIDI CartDev Code Demos for Saturn Programming Box New General MIDI and FM Sound Tone Banks Generation of sound data without use of Saturn Sound Box Getting Started a Loading a map amp running a demo sequence b System Startup c The Tone Application Voice Architecture Parameter Layers Loading an AIFF file 9 Notes on Development Hardware Tools Documentation CS oe The Saturn Sound Development Tools are all Macintosh applications This DTS CD contains all of the tools necessary to create and incorporate sound into the Saturn These are the four main development tools used for sound design on the Saturn They are used in conjunction with the Saturn Sound Box which is directly connected to the Macintosh via SCSI It is suggested that you add an in line SCSI terminator in between the sound box and the end of the SCSI cable Release Notes in English All the previously available Japanese release notes for the tools have been translated in to English for your reference purposes Look for the release notes inside each tool s folder Updates to this information for
7. alue may also be set to negative numbers with the reverse effect Play Mode As of version 1 13 this does not work Layers Double click on the number on the left side of the voice window to open the layer window If you are creating a new voice you will be asked how many layers to use You can always change this number later Each layer has the following parameters Name Double click on to edit Start End Two MIDI note numbers to determine the range of the layer These may overlap with other layers Direct Level Output level of the direct signal Effect Send Amount of signal sent to a DSP module Effect Select Number of the DSP module Direct Pan Panning position of the direct signal In addition to these parameters each layer has a slot window which contains settings for modulation and envelope data as well as the area to load the sample itself Double click on thenumber to the left of the layer name to open the slot window Loading an AIFF File Double click on the rectangle in the middle of the slot window You can then select an AIFF file to load Once you have selected a file it s name should appear in the rectangle Double click on the triangle below the filename This is the Inverter window and is where you can set the tuning of the sample along with the loop type Hardware Requirements Macintosh IIFX or better 16 MB RAM more is needed to simulate PCM streaming 300 MB Fast Hard Disk for sample stora
8. ck of a CD This is a collection of General MIDI and FM Sound Tone Banks ready to use with the Sound Tools IMPORTANT Do NOT use the previously released InVision Tone Banks If you have any questions regarding this please contact SEGA DTS by emailing dts sega com Author Thomas Szirtes Sega Europe Basic Principles For one shot sound effects What you need 1 A Macintosh 2 Sega Sound Tools on DTS disk namely Tone Editor 3 Sound Driver GSDDRVS TSK To generate and playback 1 A map file binary version contains information on where data will be stored in Sound RAM 2 One or more tone file s binary version contains the sample data together with information on how to play them back Structure of tone banks They are divided into Voices and Layers A Voice contains one or more Layers each Layer contains a sample which is playable over a certain pitch range has its own volume pan and dsp effect Therefore a Voice can contain several samples Layers Voices are switched between by sending Program Change Commands You can have if you want only one voice and assign a seperate sample to a individual note value within that therefore you do not have to send any Program Change commands e Why you want a sound box You cannot preview what your samples are going to sound like until you have incorporated the data into your code There are two systems the old large Sega Sound Box quite rare
9. ducts Ltd or Sega Technical Support in your area for more details on the MIDI interface module The CartDev compatible tools e g SoundSimulator Tone Editor and DSP Linker are known to crash whenever high volume MIDI data input and CartDev lt gt Macintosh SCSI access occur at the same time This problem also occurs with the Sophia Sound Box to a lesser extent To solve this problem avoid MIDI data input to the target while SCSI access occurs between the Macintosh and CartDev AIFF to RED Book MAC This tool coverts AIFF sound files to Red book audio by stripping the header It takes a file that contains audio data in 16bit stereo at 44 1kHz and creates a file whose data fork contains that audio data in raw Red Book format ADPCM Encoder Version 1 00E The ADPCM Encoder is a tool for producing ADPCM data that is compatible with the Sega Basic Library s SBL ADPCM Playback Library The encoder is capable of converting 8 16 or 24 bit mono stereo AIFF or Sound Designer II format files You can quickly test your Saturn Sound Box by using this Tool Demo Please refer to the Getting Started ReadMe Also included is a demo that is compatible with the MIDI Cartdev This is a collection of example code demos for the programming box The four directories contain SEQ directory starts a MIDI sequence ONESHOT directory fires sound effects using MIDI direct command PCM directory plays a PCM file CDDAPLAY directory Plays and audio tra
10. ed at 44Khz ie 22Khz we want to adjust this so that a note assigned to note 60 for example will be played back at 22Khz and not 44Khz then we change this to 72 84 for 11Khz and 96 for 8Khz NB gt gt gt gt The reason for this is due to sound theory ie in order to raise the pitch of a sound by one octave 12 tones we need to double the playback frequency Ie from 44Khz to 88Khz we need to raise the pitch by 1 octave Therefore to compensate for a 22Khz sample we also need to raise the base note 12 tones 1 octave 18 Close the window after adjusting the base note 19 Double click on Total Level box But DON T RUSH AND TOUCH ANYTHING YET in this window yet an important note the Total slider is there to fool you Total Level actually Attentuation Ie Total level of 0 means max volume Total level 127 no sound Also of interest here is the SDIR again don t touch But if you are in trouble checking this box will send all sample sound direct to output bypassing all envelope generators and effects therefore this is useful for debugging your banks if you cannot hear anything Velocity controls the ratio of velocity value to actual volume We will leave this at untitled for now This is a flat velocity curve ie it makes no difference if you want you can edit one to have a linear proportional one but you need to learn to edit these read the manual 20 Close the window you haven t changed anything have you 21
11. er Version 2 0 DSP Assembler Version 2 0 DSP Parameter Editor Version 2 0 es FF FF HF All applications marked with an asterisk support the CartDev the 68000 Assembler Environment 1 14 was modified to support CartDev since it is a target independent tool These applications check the SCSI bus at startup to determine whether a Sound Box or a CartDev is connected to the Macintosh The program debugger for the 68000 nor the Wave Editor support CartDev due to hardware limitations of the DevSaturn modified Saturn While editing functions such as cut and paste of waveforms are possible on the Wave Editor digital input based sampling or playback is not possible Important Notes Two files sound coff and satmon bin need to be placed in the same folder as the SoundSimulator for CartDev support Allow up to 20 seconds for SoundSimulator to start up NOTE The sound RAM size on the DevSaturn is 512KB 080000H While the Sound Box is capable of handling sound data up to 1024KB 0100000H it is not possible to download or play back sound data over 512KB on the DevSaturn In the SOUND TOOL DEMOS CARTDEV folder is a demo that is compatible with the MIDI Cartdev It is necessary to have a DevSaturn equipped with a special MIDI interface module for CartDev and DevSaturn MIDI based sound development Only the SoundSimulator can be used for the playback of project files if the interface module is not connected to the DevSaturn Contact Cross Pro
12. ge MIDI interface MIDI Controller Saturn Sound Box Supplied by SEGA SCSI Cable Supplied by SEGA InLine SCSI terminator Tools Supplied by SEGA Sound Simulator MAP Editor Tone Editor Wave Editor Linker DSP Editor eLinker by request includes 3 D sound modules QSound Ysound SDSS 68000 Assembler optional DSP programming tool SSBug the debugging tool that works with SDSS Other Tools needed Midi Sequencer MIDI file format 1 It is suggested to work with a more sophisticated digital audio editor than the one that is provided Specs For a more detail description of the specs see ST 081 32 note polyphony 8sound control numbers 1 sequence or 127 direct midi commands per control number up to 32 op FM note each carrier or modulator takes up one note of polyphony 4 channels of PCM although it hogs a lot of bus time Functional Overview Two CPU interfaces 68000 data transfer can be operated in parallel Built in MIDI interface 2 x 16 channels 128 step DSP Built in ring filter buffer for filtering Memory control functions allow program in sound memory to operate the CPU 16 channel Digital Mixer DMA transfer between SCSP Control Registers Sound Memory 3812 bytes Sound Development Documentation Sound Development Manual ver 1 1 Saturn Sound Simulator Manual Wave Editor User s Manual SCSP Waveform Editor Technical Specifications Tone
13. ick on End Box here is the highest note value that our layer will extend to We will enter 60 again ie this sample will only sound be valid when note value 60 C4 is sent on our voice 12 Drag our Direct Level slider to 7 max This is the amount of sample signal that is routed straight to the output Effect Send is the amount of sample signal that is routed via the DSP more specifically the via the DSP input specified by Effect Select however this we will leave at 0 Direct Pan we leave 13 Double Click on No 0 box Up appears a diagram this is how our sample is processed 14 Double Click on the empty rectangle box in the center of the window A file requestor should appear 15 Select the AIF sample you want Information on it will appear take note of the sampling rate Click OK 16 The sample name will now appear in the rectangle Yeah 17 Double Click on the Upside Triangle the Inverter Symbol Here we can change the LOOP SELECT either off for no loops or forward for looping others We will leave this OFF We can also change the Base Note This affects the playback rate that our sample is played at A Base Note of 60 means that if a 44Khz Sample is triggered on note value 60 it is played back at 44Khz Therefore if we trigger a 44Khz on note value 61 cause another sample is assigned already to note 60 we want to change this to base note 61 in order to still playback at 44Khz Also if our sample is not record
14. ing OK until they go away The reason this happens is because a Sound Simulator map stores the paths of the files it contains Whenever the files are moved they must be re loaded individually Then the map can be saved again along with the new paths At this point the blank map should be replaced by the new map Each line in the table indicates an area in memory used by a file or DSP work RAM The area is shown by the start end addresses in the third column These are in hex The type of data is shown in the fifth column and the file name if any is shown in the sixth At this point memory has been allocated for these files but they haven t been loaded Double click on the line for the file CHOICE COLLECT You will get a dialog box with the name of the currently selected file along with the start and end addresses currently allocated Click once on the oval containing the name of the file A file selection box will appear open to the folder that the SndSim application is in Find the file CHOICE COLLECT and double click on it s name The file selection box will go away leaving you with the dialog box Nothing has changed in the box but by re loading the file you have established it s new path Close the dialog box highlight the same line in the map go to the File menu and choose DOWNLOAD After a few seconds an X should appear in the first column This means the file is now loaded Repeat the steps in this paragraph for the other files in
15. nd driver you have very little control over this Procedure 1 Load sound data into HI RAM sound driver 2 Initialize sound driver with sound map see below 3 Change Map see below 4 Transfer sound data items to sound ram using SND library see below 5 For each midi channel you use you need to see below Send a Tone Bank Select this selects the tone bank that will be active on that channel Send a Program Change this selects which voice within that bank will be active on that channel Send a Note On Triggers the desired note on that voice Send a Note Off switches off that note 6 More advanced stuff like Pitch Bend midi command 6 data 1 pitch up data2 pitch down Volume Change midi command 3 data 1 10 data 2 volume Use the following code as a template based on oneshot c on DTS CD void sound void SndIniDt snd_init External Reference to our data files in memory main extern char sddrvs extern char sddrvs_end extern char scsp_tone_bank extern char scsp_tone_bank_end extern char scsp_map extern char scsp_map_end Le process 1 initialize the sound driver version 2 04 here g SND_INI_PRG_ADR snd_init Uint16 sddrvs 68k snd drv program SND_INI_PRG_SZ snd_init Uint16 sddrvs_end sddrvs sound driver size SND_INI_ARA_ADR snd_init Uint16 scsp_map sound area map 7 SND_INI_ARA_SZ snd_init Uint16 sc
16. ry map Each of these slots contains a certain type of data The types are BANK Data binary files for tone banks SEQUENCE Data converted MIDI sequences DSP Program program data for the DSP Linker DSP Work RAM buffer area to hold data to be processed by DSP algorithms The Sound Simulator must be run first and a memory map loaded or created before any of the other applications can be used It is also where you convert MIDI files into sequence data which the Saturn can read and arrange it into sequence banks After this has been done you can play the sequences back with any DSP to hear how it will sound in the Saturn e Loading a map and running a demo sequence The folder containing the SndSim application should also have the following files CHOICE MAP CHOICE MAP bin NEW CHOICE TONE NEW CHOICE TONE bin CHOICE TUNE more EXB CHOICE TUNE more EXL CHOICE TUNE more LRI CHOICE TUNE more YLI CHOICE COLLECT Make sure these files are in the SndSim s folder before proceeding because the above files are located in the SOUND TOOL DEMOS CHOICE directory e System Startup 1 Double click on the Saturn SndSim E1 28 icon A blank map should appear 2 Select OPEN from the File menu 3 Choose DEMO MAP click on OPEN If this is the first time you are opening the file read the following Otherwise go on to 4 You will get a series of dialog boxes indicating that certain files can t be found Don t worry just keep click
17. s Getting Started When you launch the Tone Application the only thing that appears is the Master Volume window This sets the overall output level for the tone bank To open up a tone bank select OPEN from the File menu and click on a file name to open it or select NEW from the file menu to create a tone bank from scratch If you select a different file from the one loaded in the Sound Simulator it will become the active tone bank Voice Architecture A tone bank may contain up to 128 different voices numbered 0 127 which correspond to standard MIDI program change numbers Each of these voices is made up of up to 255 layers A layer is an AIFF sample along with key mapping output and panning information as well as velocity mapping and modulation data Voice Parameters Voices contain the following parameters Number Standard MIDI program change number 0 127 Name Double click on to edit Bend Range Pitch bend range in semitones from center to either extreme 0 12 0 12 semitones 13 2 octaves Portamento As of version 1 13 this does not work Vol Bias Determines how the output level of the voice responds to key velocity and volume ctrl 7 information A value of 0 can be thought of as neutral producing roughly a linear response As Volume Bias is set higher positive lower velocities and volumes are output at higher levels In addition the maximum output level is reached with lower velocities and volumes The v
18. sp_map_end scsp_map snd area map size Initialize this transfers map data from main to sound ram SND_Init amp snd_init SND_ChgMap 0 change sound area map start with map 0 For each data item in your map transfer it to sound ram SND_MoveData Uint16 scsp_tone_bank source data address in HI ram Uint32 scsp_tone_bank_end scsp_tone_bank source data size SND_KD_TONE type of data 0 identifier ie as specified in map SND_ChgMix 0 0 change mixer default SND_SetTI1V1 15 max volume Change tone bank 0 select SND_CtrlDirMidi 0x02 0x00 sound control and priority number not important now 0x03 midi command control change 0x01 midi channel 1 0x20 data 1 tone bank change 0x00 data 2 tone bank number Program change 1 SND_CtrlDirMidi 0x02 0x00 as above 0x04 midi command program change 0x01 midi channel 1 0x01 data 1 program number ie voice number 1 0x00 data 2 not used 0 Play a simple note SND_CtrlDirMidi 0x02 0x00 as above 0x01 midi command note on 0x01 midi channel 1 60 data 1 note value pitch 60 127 data 2 velocity 127 while 1 General Description The Sound Simulator is an application which allows you to allocate sound RAM into individual slots in the form of a memo
19. the Sound Simulator 3 02 Tone Editor 2 07 and Sound Driver 2 10 are currently being translated and will be included on the DTS Website shortly Sound Simulator This tool allows the sound designer to map out the sound memory and to simulate performance in game code Refer to document ST 81 for more detailed instructions The Sound drivers must be included in the same directory as the Sound Simulator For more detailed information please refer to ST 168 To read more about MIDI conversion please refer to ST 66 The SndSim301 folder contains version 3 01 of SoundSimulator along with version 2 10 of the sound driver PLEASE NOTE Sound Driver 2 10 is NOT backwards compatible if your tone bank is not created using at least 2 0 of the Tone Editor then you will NOT get any sound out of it A testbed for this will be included on the DTS Website shortly SoundSimulator version 3 00 and above offers a more user friendly English interface along with improvements in the Song Collector and Function Key features It is also capable of supporting both Japanese and English Macintosh Sytem environments The differences between Sound Driver 2 04 and 2 10 are currently being translated For your reference the Sound Driver Readme details differences between previous versions of the sound driver 2 04 and below Tone Editor This tool is a voice patch editor It is used to organize both sampled sounds and complex FM for sequential control and sound effect
Download Pdf Manuals
Related Search