Conference napalm::commusic_v1

    
    1. The only thing I see you confusing is cutoff and EG depth.
    	You only get cutoff changed by the EG (and the keyboard
    	tracking.)
    
    	The phase distortion technique differs from VCF's in one
    	important way: with PD, if the EG is "0", the wave is a
    	sine wave of 'full' amplitude, whereas decreasing the
    	cutoff of a VCF will result first a sine, but then results
    	in decreasing amplitude to 0 cutoff well below the input
    	frequency.
    
    	I find it easier to think of the wave produced as 'somewhere'
    	between a sine, and the one you selected based on the EG level.
    
    	Back with reply to 2. in a second
    
    	ron    
    

    
    yup, some rates/levels arent as useful as others. Since in your
    example is the first 'segment', it doesnt do much. 
    
    You can get trills, vibrato, chiffs, and other wierdness of
    transient pitch modulation with that EG with more meaningful
    envelopes.  For experimentation it's better to keep the
    "filter" UN-modulated to focus your ear on pitch change effects. 

    Ron

    
>    1. The only thing I see you confusing is cutoff and EG depth.
>    	You only get cutoff changed by the EG (and the keyboard
>    	tracking.)
    
    	I thought I said I saw cutoff as a function of the 'level' in
    the DEG.  Is that different from what you're saying?
    
>    	The phase distortion technique differs from VCF's in one
>    	important way: with PD, if the EG is "0", the wave is a
>    	sine wave of 'full' amplitude, whereas decreasing the
    
    	When you say 'EG is "0"' are you refering to the 'rate' or the
    'level'?
    
    
>    yup, some rates/levels arent as useful as others. Since in your
>    example is the first 'segment', it doesnt do much. 
    
	My example was a case in which it was the ONLY step in the DCO
    EG.  Does it still perform any function or is it simply a case where
    some programmer may have been editing a sound and forgot to take
    out a step that made no effect on the sound?
    
    	Thanks for the help so far.
    
    	Dan

    Letsee:
    
    1: Youve got it. Level of EG is "cutoff". Rate is how long it takes
       to get there.
    
    2: The example starts at 0 level, right? Then it goes at whatever
       rate to 0 level. Exciting or meaningful it aint.
	You want to at least go "up" to some level and then back down.
    	Some voices have a little quick blip on the attack which would
    	be more like R1:90,L1=50 (gets you "up" there) R2=75,L2=0.
        Experiment with the values. Its just important that L1 is non-0.
        
    Try:      

    more:
    
    	The "difference" that I mentioned of DCW and VCF you 
    	probably dont need to worry about now. It only means
    	that the VCA EG must be used to control volume, where
    	in some voices a VCF can be used without a VCA following
    	it. It must be my analog background....
    
    	The strange 'segment' were talking about may actually
    	"do" something on the 101 (I have a 5000). Especially
    	if you got it as a purchased preset. The way to find
    	out again is to get the OTHER parameters (DCW,VCA,vibrato,etc)
       out of the patch manually so that you can twiddle with
       the one in question-  pitch modulation. *You* go see/hear
    	what it does or doesnt do. Then change it till it does/doesnt.
    	Thats how to learn.
    
    	enjoy. 

    The function that is baffling me at the moment is the "Key follow".
    I thought that this would open the DCW or DCA as you go up the
    keyboard. But either I don't understand this correctly or the effect
    is very subtle. Even with a value of 9 here, I can't hear any noticable
    difference. Are my ears shot or have I just not hit the right
    combinations yet?
    
    Rob
    

    Key Follow changes the rates of the envelopes.  It does not work
    like key follow on an analog synth.  To make the effect obvious,
    set a very slow rate, then hit two keys simultaneously, one at the
    low end and one at the high end.  At key follow 9 you should hear
    an obvious difference in the rates.
    
    Going from 0 to 0 at any rate is a no-op.  The default rates set
    by the initialize button are 50.  That's 'cause 50 is halfway between
    0 and 100.  I think the initial envelope segment rate defaults to
    99.  A lot of patches do little or no pitch modulation.  An 8 segment
    EG for pitch is mild overkill, but it didn't cost them anything.
    You can make really bizarre SF noises using heavy pitch (DCO) EG
    and ring modulation.  Put different envelopes on the two oscillators.
    
    I wish they had put an EG on LFO rate instead.

    len.
    

    Len,
    
    Thanks for the info on Key follow. Being the hep analog synth cat
    that I am, I figured I knew what key follow meant and didn't read
    the manual closely enough.
    
    I agree 100% on your comments about the lack of LFOs. If only they
    had included some sort of "looping" action on the EGs instead of
    making them one shot only, you could create some real fancy LFO
    mod. Oh well, just wishful thinking.

    BTW, there hasn't been any action in the CZ patch exchange note
    lately. I've finally found some time to play with programming my
    beast and I'll try to put a couple of patches in soon. Anybody else
    have any interesting sounds to share?   
    

This came today in the net.music.synth...I assume no responsibility
    for it's accuracy but it seemd to me to be appropriate to enter
    as a reply to this note. Enjoy.....   
    
    
Newsgroups: net.music.synth
Path: decwrl!hplabs!hao!nbires!seismo!mcvax!ukc!reading!ellis
Subject: Casio CZ MIDI manual
Posted: 17 Nov 86 09:25:29 GMT
Organization: Comp. Sci. Dept., Reading Univ. UK.
Keywords: casio cz
 
 
[ Discerning line eaters like Larry Niven ]
 
-------------------------------------------------------------------------------
----------------- CASIO CZ MIDI GUIDE  condensed version ----------------------
-------------------------------------------------------------------------------
---- or: Everything you Never Wanted to Know about MIDI but are going to ------
------------------- Find Out Anyway -------------------------------------------
-------------------------------------------------------------------------------
 
        THIS IS AIMED SPECIFICALLY AT CZ101,CZ1000 and CZ5000 OWNERS.
 
NOTE: I beleive the data given to me by Casio to be a) correct, and b) public
	domain ( since they just give it to you if you are persistent enough.
	Please forgive spelling, syntactic, or grammatical errors since I am
	trying to condense the manual as I go along...
 
-------------------------------------------------------------------------------
 
Right, at long last, I have been able to get a reply out of CASIO UK about the
MIDI standards of the CZ series of synthesizers. For those of you who know
how to play your keyboard by remote control from a computer, this will be of
some interest, since it covers transmission of programming information ( both
to and from the CZ ), setting of the controllers you previously couldn't access
like tone mix level, and other bits besides.
 
First off, let's recap on the simple stuff. The MIDI is a digital interface to
musical instruments, and relies on serial transmission of data. These data are
usually talked of in terms of bytes, and I shall be using hexadecimal numbers
in this posting.
 
There are basically two types of bytes sent over MIDI - control bytes and data
bytes. Control bytes are distinguished by having values over 0x80 ( 80 hex, 
128 decimal ), and these have valious meanings:
 
1) NOTE ON
----------
 
A note on message consists of sending a "NOTE ON" control byte, the note number
you want to turn on, and the velocity at which you want to play the note. The
note on control is 90 plus the channel number. So, for example, if you want to
play note 32 ( = hex 20) at speed 64 (= hex 40), on the midi instrument receiv-
ing on channel two, then you would send:
 
	92		20		40
NOTE ON, channel 2	Note #32	Velocity 64
 
If you wish to turn two or more notes on at the same time, the control byte
need not be retransmitted. Eg to turn note 35 on as well, you could send
 
	92	20	40	23	40
NOTE ON, ch2	---32 on--	---35 on--
 
These codes can be transmitted both ways on all the CZ 101,1000,5000, but since
they do not detect note velocity, it is always transmitted and recognized as
64 (= 40 hex).
 
2) NOTE OFF
-----------
 
Just send a note on message with velocity 0. Eg to turn note 35 off, send
 
	92	23	00
NOTE ON,Ch 2	--35 off--
 
 
3) CONTROL CHANGE
-----------------
There are several controls that can be set from MIDI. Just send a "CONTROL CH"
byte , which is B0 plus the channel number, the number of the control that you
wish to change, then the value you wish to set it to. Eg for CZ101 portamento
time, send
 
	B0	05	10
CONTROL, ch 0	--ctrl 5=16--
 
The controls are:
 
CZ101/1000
 
01	Vibrato on/off		Send 0 for off, 7F for ON
05	Portamento time		Send number 00..63 (0..99)
06	Master tune		Send number 00..7F
41	Portamento on/off	Send 0 for OFF, 7F for ON
 
CZ5000
 
01	Modulation wheel	Send number 00..7F
05	Portamento time		Send number 00..63
06	Master tune		Send number 00..7F
40	Sustain pedal		Send 0 for OFF, 7F for ON
41	Portamento on/off	Send 0 for OFF, 7F for ON
 
4) PROGRAM CHANGE
-----------------
This allows you to change between the preset sounds ( and your internal
sounds and cartridges ). Just send C0 plus the channel number, then the
program number. Eg to set CZ101 on channel 1 to Synth Bass:
 
	C1		07
PROGRAM ch 1		Program 7
 
Note that the preset tones are given numbers :
 
CZ101/1000
 
0..0F	Preset sounds 1..16
20..2F	Internal sounds 1..16
40..4F	Cartridge sounds 1..16
 
CZ5000
 
00..1F	Preset sounds A1,A2,A3....D6,D7,D8
20..3F	Internal sounds A1,A2.....D6,D7,D8
 
 
5) PITCH BEND CHANGE
--------------------
 
This is acheived by sending E0 plus the channel number, then two bytes 
denoting the new value of pitch bend. The first byte is the most significant,
and the second the least significant. Note also that the lower 6 bits of the
lower byte are not used, and that the central position of the wheel corres-
ponds to the byte sequence 40 00.
 
	HIGHEST		7F 40
 
	HIGHER
 
	CENTER		40 00
 
	LOWER
 
	LOWEST		00 00
 
 
So, to bend the instrument on channel two UP by about half its maximum amount,
send
 
	E2		60	00
BEND channel two	---1/2 up--
 
 
6) AFTER TOUCH
--------------
Is not supported on the CZ101/1000/5000. Sorry!
 
7) MODE CHANGE
--------------
This is very similar to the CONTROL CHANGE message, and can be regarded as a
special case. 
 
OMNI ON		send E0 + channel, 7D, 00
OMNI OFF	send B0 + channel, 7C, 00
POLY ON		send B0 + channel, 7F, 00
POLY OFF	send B0 + channel, 7E, 00
 
OMNI mode plays any MIDI data received at the MIDI IN plug on the back of
the machine, regardless of channel. POLY mode is equivalent to the SOLO button
on the front panel. With the CZ101, for instance, POLY OFF ( =SOLO ) allows
the synth to be used as four monophonic synthesizers under remote control.
 
LOCAL ON	send B0 + channel, 7A, 7F
LOCAL OFF	send B0 + channel, 7A, 00
 
Local mode means that the keyboard is "connected" to the sound producing 
part of the CZ within the machine itself. With LOCAL ON ( the default setting ),
playing the keyboard both sends note messages out of the MIDI port, and also
makes sounds at the same time. If you want to do wierd things like keyboard
splitting, LOCAL OFF will allow you to see what the keyboard is doing without
the CZ making any sound at all. You could then act on that information and send
the keyboard a command depending on the keys that has nothing to do with them,
eg program change or pitch bend. The possibilities are endless !
 
 
SEQUENCER MESSAGES
------------------
The CZ5000 has its own internal sequencer, which can be controlled by:
 
F8	Clock byte: transmitted 24 times per quarter note ( crotchet )
FA	Start: same as pressing the PLAY button on the front panel
FB	Continue: continue song where last stopped
FC	Stop: stops song play at current position
FD	Active sense: basically, a cry of "Is there anybody out there". If no
	reply is received within about 1/3 second, it shuts the voice off.
 
 
SYSTEM EXCLUSIVE MESSAGES
-------------------------
 
At last, the really meaty stuff. :-)
 
These all have the basic form:
 
	F0	machine ID	some bytes	F7
SYS EX MESSG	YES, YOU	DO THIS		END OF SYS EX
 
Ok, so not very specific, bu that was deliberate to allow manufacturers to use
all the lovely bells and whistles they put on their machine over the MIDI !
 
Since these are usually controlled by computer, I have set them out as a
computer/synthesizer dialogue. Note that the computer MUST wait for replies
before proceding, or all will fail !
 
Here we go, then 
 
1) SET BEND RANGE
-----------------
 
Computer:	F0 44 00 00 70+channel 40 data F7
 
Eg to set bend range to 8 on channel 4, send
		F0 44 00 00 74 40 08 F7
 
2) KEY TRANSPOSE
----------------
 
Computer:	F0 44 00 00 70+channel 41 data F7
 
Data is as follows:
Key:	G  A  A# B  B# C  C# D  E  E# F  F#
Data:	45 44 43 41 41 00 01 02 03 04 05 06
 
Eg to set key on channel 0 to C#, send
		F0 44 00 00 70 41 01 F7
 
3) TONE MIX
-----------
 
Computer:	F0 44 00 00 70+channel 42 data F7
 
The data is 00 to turn tone mix off, or 41..49 for mix level 1..9
 
Eg to set tone mix on channel 0 to 7, send
		F0 44 00 00 70 42 47 F7
 
4) ASK ABOUT PROGRAMMER ( Send request 2 )
-----------------------
 
Computer:	F0 44 00 00 70+channel 19 00
CZ101/1000:	F0 44 00 00 70+channel 30
Computer:	70+channel 31
CZ101/1000:	data1 data2 F7
Computer:	F7
 
data1 is the program selected ( see PROGRAM CHANGE )
data2 returns the vibrato/portamento on/off setting:
 
data2	00	10	20	30
Vibrato	OFF	OFF	ON	ON
Port'o	OFF	ON	OFF	ON
 
Eg an exchange such as
Computer:	F0 44 00 00 70 19 00		"Want data on channel 0"
CZ101:		F0 44 00 00 70 30		"Gotcha.. data ready"
Computer:	70 31				"Ok, give it to me"
CZ101:		27 30 F7			"Internal 8, v on, p on"
Computer:	F7
 
REMOTE PROGRAMMING
------------------
The send request 1 and receive request 1 messages.
 
These dump a lot of data across the MIDI, which is the same for both messages,
except that the data go the other way. The exchanges are:
 
Send request
Computer:	F0 44 00 00 70+channel 10 program
CZ101/1000:	F0 44 00 00 70+channel 30
Computer:	70+channel 31
CZ101/1000:	<tone data> F7
Computer:	F7
 
Receive request
Computer:	F0 44 00 00 70+channel 20 program
CZ101/1000:	F0 44 00 00 70+channel 30
Computer:	<tone data> F7
Cz101/1000:	F7
 
The program byte is the same as that set by the PROGRAM CHANGE function, with
the addition that you can request the temporary sound area as well ( number is
60 ). This is the area that is used if you have altered a preset and not saved
it into internal memory.
 
<tone data> is a sequence of 256 bytes containing a LOT of info. Now Casio have
done a lot of funny things with these, like splitting bytes in half and encoding
things in wierd ways so please bear with me.
 
To keep everything this side of infinite length, I shall adopt the same strategy
as the manual, which is to write data in bytes, although they are transmitted
in half- bytes. For example, me writing a byte as 5F requires you to transmit or
receive as 0F 05 ( wierd, huh ? ). This will obvoiusly save a lot of space.
 
So, here goes again :-)
 
There are 25 distinct sections within <tone data>
 
Sec#	Length	Symbol		Contents
	(bytes)
 
1	1	pflag		line select data, octave range
2	1	pds		detune up or down
3	2	pdl,pdh		detune range
4	1	pvk		vibrato wave number
5	3	pvdld,pvdlv	vibrato delay time
6	3	pvsd,pvsv	vibrato rate
7	3	pvdd,pvdv	vibrato depth
8	2	mfw		dco1 waveform
9	2	mamd,mamv	dca1 key follow
10	2	mwmd,mwmv	dcw1 key follow
11	1	pmal		end step number of dca1 envelope
12	16	pma		dca1 envelope rate/level
13	1	pmwl		end step number of dcw1 envelope
14	16	pmw		dcw1 envelope rate/level
15	1	pmpl		end step number of dco1 envelope
16	16	pmp		dco1 envelope rate/level
17	2	sfw		dco2 waveform
18	2	samd,samv	dca2 key follow
19	2	swmd,swmv	dcw2 key follow
20	1	psal		end step number of dca2 envelope
21	16	psa		dca2 rate/level
22	1	pswl		end step number of dcw2 envelope
23	16	psw		dcw2 rate/level
24	1	pspl		end step number of dco2 envelope
25 	16	psp		dco2 rate/level
 
1) PFLAG
Looking at bits,
	0000  00   00
 Not used^   OCTV  LS
 
OCTV controls octave range: 00=octave 0, 01=+1, 10=-1
 
LS is the line select: 00=1, 01=2, 10=1+1', 11=1+2'
 
So, fo Octave +1, line select 1+1', PFLAG=00000110 = 06
 
2)PDS
For detune +, PDS is 0, for detune - it is 01
 
3)PDETL,PDETH
 
Two bytes controlling the depth of the detune.
The first byte is the FINE data.
 
FINE:	0..15	16..30	31..45	46..60
Byte:	00..0F	11..1F	21..2F	31..3F
 
The second contains both the octave and note data:
 
OCT:	0	1	2	3
NOTE:	0..11	0..11	0..11	0..11
Byte:	00..0B	0C..17	18..23	24..2F
 
 
4) PVK
 
This is the vibrato wave number, encoded as follows
 
WAVE NUMBER:	1	2	3	4
Byte:		08	04	20	02
 
5) PVDLD,PVDLV
 
This is the vibrato delay time, transmitted in three bytes.
 
Delay	Bytes		Delay	Bytes		Delay 	Bytes
25	19 00 19	50	32 00 4B	75	4B 00 DF
26	1A 00 1A	51	33 00 4F	76	4C 00 E7
27	1B 00 1B	52	34 00 53	77	4D 00 EF
28	1C 00 1C	53	35 00 57	78	4E 00 F7
29	1D 00 1D	54	36 00 5B	79	4F 00 FF
30	1E 00 1E	55	37 00 5F	80	50 01 0F
31	1F 00 1F	56	38 00 63	81	51 01 1F
32	20 00 21	57	39 00 67	82	52 01 2F
33	21 00 23	58 	3A 00 6B 	83	53 01 3F
34	22 00 25	59	3B 00 6F	84	54 01 4F
35	23 00 27	60	3C 00 73	85	55 01 5F
36	24 00 29	61	3D 00 77	86	56 01 6F
37	25 00 2B	62	3E 00 7B	87	67 01 7F
38	26 00 2D	63	3F 00 7F	88	58 01 8F
39	27 00 2F	64	40 00 87	89	59 01 9F
40	28 00 31	65	41 00 8F	90	5A 01 AF
41	29 00 33	66	42 00 97	91	5B 01 BF
42	2A 00 35	67	43 00 9F	92	5C 01 CF
43	2B 00 37 	68	44 00 A7	93	5D 01 DF
44	2C 00 39	69	45 00 AF	94	5E 01 EF
45 	2D 00 3B	70	46 00 B7	95	5F 01 FF
46	2E 00 3D	71	47 00 BF	96	60 02 1F
47	2F 00 3F	72	48 00 C7	97	61 02 3F
48	30 00 43	73	49 00 CF	98	62 02 5F
49 	31 00 47	74	4A 00 D7	99 	63 02 7F
 
For delays in the range 0..31, just transmit 00..1F, 00, 00..1F eg for delay
of 12, send 0C 00 0C. This is convenient since it saves me typing in another
column of boring numbers ;-)
 
6) PVSD,PVSV
 
Again, here comes another table for conversions. The first column (0..24) is
omitted since the only difficult thing needed is to add 01 00 20 to each entry
( The first few go 00 00 20, 01 00 40, 02 00 60, ... 06 00 E0, 07 01 00, ..)
 
Rate	Bytes		Rate	Bytes		Rate	Bytes
25	19 03 40	50	32 09 E0	75	4B 1C E0
26	1A 03 60	51	33 0A 60	76	4C 1D E0
27	1B 03 80	52	34 0A E0	77	4D 1E E0
28	1C 03 A0	53	35 0B 60	78	4E 1F E0
29	1D 03 C0	54	36 0B E0	79	4F 20 E0
30	1E 03 E0	55	37 0C 60	80	50 23 E0
31	1F 04 00	56 	38 0C E0	81	51 25 E0
32	20 04 60	57	39 0D 60	82	52 27 E0
33	21 04 A0	58	3A 0D E0	83	53 29 E0
34	22 04 E0	59	3B 0E 60	84	54 2B E0
35	23 05 20	60	3C 0E E0	85	55 2D E0
36	24 05 60	61	3D 0F 60	86	56 2F E0
37	25 05 A0	62	3E 0F E0	87	57 31 E0
38	26 05 E0	63	3F 10 60	88	58 33 E0
39	27 06 20	64	40 11 E0	89	59 35 E0
40	28 06 60	65	41 12 E0	90	5A 37 E0
41	29 06 A0	66	42 13 E0	91	5B 39 E0
42	2A 06 E0	67	43 14 E0	92	5C 3B E0
43	2B 07 20	68	44 15 E0	93	5D 3D E0
44	2C 07 60	69 	45 16 E0	94	5E 3F E0
45	2D 07 A0	70	46 17 E0	95	5F 41 E0
46	2E 07 E0	71	47 18 E0	96	60 47 E0
47	2F 08 20	72	48 19 E0	97	61 4B E0
48	30 08 E0	73	49 1A E0	98	62 4F E0
49	31 09 60	74	4A 1B E0	99	63 53 E0
 
7) PVDD,PVDV
 
These are again encoded as three bytes in a most obscure way. Below 32, the
encoding is 00..1F, 00, 01..20 eg for depth 13, send 0D 00 0E.
 
Depth	Bytes		Depth	Bytes		Depth	Bytes
25	19 00 1A	50	32 00 4F	75	4B 00 E7
26	1A 00 1B	51	33 00 53	76	4C 00 EF
27	1B 00 1C	52	34 00 57	77	4D 00 F7
28	1C 00 1D	53	35 00 5B	78	4E 00 FF
29	1D 00 1E	54	36 00 5F	79	4F 01 07
30	1E 00 1F	55	37 00 63	80	50 01 1F
31	1F 00 20	56	38 00 67	81	51 01 2F
32	20 00 23	57	39 00 6B	82	52 01 3F
33	21 00 25	58	3A 00 6F	83	53 01 4F
34	22 00 27	59	3B 00 73	84	54 01 5F
35	23 00 29	60	3C 00 77	85	55 01 6F
36	24 00 2B	61	3D 00 7B	86	56 01 7F
37	25 00 2D	62	3E 00 7F	87	57 01 8F
38	26 00 2F	63	3F 00 83	88	58 01 9F
39	27 00 31	64	40 00 8F	89	59 01 AF
40	28 00 33	65	41 00 97	90	5A 01 BF
41	29 00 35	66	42 00 9F	91	5B 01 CF
42	2A 00 37	67	43 00 A7	92	5C 01 DF
43	2B 00 39	68	44 00 AF	93	5D 01 EF
44	2C 00 3B	69	45 00 B7	94	5E 01 FF
45	2D 00 3D	70	46 00 BF	95	5F 02 0F
46	2E 00 3F	71	47 00 C7	96	60 02 3F
47	2F 00 41	72	48 00 CF	97	61 02 5F
48	30 00 47	73	49 00 D7	98	62 02 7F
49	31 00 4B	74	4A 00 DF	99	63 03 00
 
8) MFW
 
These two bytes transmit the waveform for DCO1, and also the modulation ie
ring, noise or none.
 
		First byte	Second byte
		000 000 0  0  00 000 000
 
First=1		000        0  00
Fisrt=2		001        0  00
First=3	        010        0  00
First=4		100        0  00
First=5         101        0  00
First=6         110        0  01
First=7		110        0  10
First=8		110        0  11
Second=1	    000 1  0  00
Second=2	    001 1  0  00
Second=3	    010 1  0  00
Second=4            100 1  0  00
Second=5	    101 1  0  00
Second=6	    110 1  0  01
Second=7	    110 1  0  10
Second=8	    110 1  0  11
NO MODULATION			 000
RING MODULATION			 100
NOISE MODULATION		 011
 
So, for instance, to set first = 4, second= 2, ring modulation, we have
 
	100 001 1 0 00 100 000 = 1000 0110 0010 0000 = 86 20
 
9) MAMD,MAMV
 
These two bytes set the DCA1 key follow:
 
Key follow:	0   1   2   3   4   5   6   7   8   9
1st byte:	00  01  02  03  04  05  06  07  08  09
2nd byte:	00  08  11  1A  24  2F  3A  45  52  5F
 
 
10) MWMD, MWMV
 
These two bytes set the DCW1 key follow
 
Key follow:	0   1   2   3   4   5   6   7   8   9
1st byte:	00  01  02  03  04  05  06  07  08  09
2nd byte:	00  1F  2C  39  46  53  60  6E  92  FF
 
 
11) PMAL
 
This sets the position of the end step on DCA1. Step 1..8 gives bytes 00..07.
 
12) PMA
 
This consists of 8 repetitions of Rate,Level.
 
Given that you wish to set rate r, the data you need to send is
 
	Byte= 119 x r
	      -------
		99
Conversely, if byte=0, rate=0, if byte=7F, rate=99, otherwise
 
	r=99 x byte
	  --------- + 1
	     119
 
Add 80 hex if the level will be coming down on this step.
 
The level goes up linearly, with 0 being 00, up to 99 is 7F, so that
 
	Byte= 127 x level
	      -----------
	         99
 
and	Level= 99 x byte
	       --------- + 1
	          127
 
except at byte=0 where level=0, and byte=127, where level=99
 
In all these conversions, fractional parts are ignored, so a result of
byte=24.6987 would be taken as byte=24.
 
13) PMWL
 
End step number for DCW1. Same as PMAL
 
 
14) PMW
 
This sets the steps in DCW1, and consists of 8 repetitions of Rate,Level.
The format is similar to PMA, so that you add 80 ( 128 dec ) to the rate if
the level is coming down this step, and that you add 80 to the level if you
wish to set a sustain point. The level data is handled the same as PMA, but 
for some strange reason the rate data is encoded differently.
 
So	byte= 119 x level
	      ----------- + 8
	          99
 
and	level= 99 x (byte-8)
	       ------------- +1
	            119
 
except where byte=8, level=0, and where byte=77, level=99
 
15) PMPL
 
Another end step setting, this time for DCO1. Same as PMAL and PMWL
 
16) PMP
 
Another envelope setting, this time for the DCO1 rates and levels. Again uses a
completely different encoding scheme.
 
	byte= 127 x rate
	      ----------
	          99
and
	rate= 99 x byte
	      --------- + 1
	         127
 
except where byte=00, rate=0, where byte=7F, rate=99
 
For the level, level data 0..63 translate as bytes 00..3F, and level data
64..99 translate as bytes 44..67.
 
17) SFW
 
These two bytes set the waveform for DCO2. They use the same format as MFW
does for DCO1, except that the modulation bits are ignored ( it is best to set
these bits to zero , just in case ).
 
18) SAMD,SAMV
19) SWMD,SWMV
20) PSAL
21) PSA
22) PSWL
23) PSW
24) PSPL
25) PSP
 
All the above use the same formats as their counterparts for the first set of
DCO,DCW,DCA, and perform exactly the same functions on the DCA2,DCW2,and DCO2.
 
 
 
This concludes what I hope has been an informative article ( if rather a long
one :-)  ). 
 
Thong
 
 
 
 
[ The views above are my own (except any quotes !) and not anyone elses.. so   ]
[ flame me personally, not them                                                ]
+------------------------------------------------------------------------------+
|                                                                              |
|   "Thong" Ellis, Reading University Computer Science VAX "Sage", England     |
|                                                                              |
|     "But there aren't any REAL people here at all..." : Roosta               |
|                                                                              |
+------------------------------------------------------------------------------+
    

    OK Guys.  I have had my CZ5000 for about six months and I am 
    trying to master the method of creating sounds, and  getting 
    there  slowly.    I heard recently on a DX21 the "Sample and 
    Hold" preset, which I would dearly like  to  create  on  the 
    Casio.   I know the method for the creation of sounds on the 
    Casio is different than the DX, but I had hoped I understood 
    enough  about  the  "sound"  from  the  DX and the method of 
    creating on the CZ to accomplish it.
    
    I was wrong.
    
    I am getting nowhere. 
    
    Can anybody tell me:-
    
    	a) Can it be done on a CZ
    
    	b) If so, where do I start 
                               
    Bear with me if I am asking the obvious!!
    
    				Peter.
    

    First of all, the CZs and DXs produce their sounds by *totally*
    different mechanisms.  People who argue that the CZs are a "poor
    man's DX" simply don't know what they're talking about.  This
    is not directed at you, but just a general comment.  There is
    no "real" FM capability in the CZ, and while the DXs may use a timevarying
    readout rate from a basic wavetable, they do so in a way completely
    different from the CZs.
    
    OK.  It's possible that the CZ can reproduce or closely approximate
    some DX patches, but it depends on the patch in question.  Without
    knowing anything about this "Sample and Hold" patch, I couldn't
    tell you where to start, and from what you say you don't understand
    enough about either the patch's sound or the CZ architecture to
    put the two together.  Note that even if you did, the answer might
    still be "can't be done".
    
    Given that, I'd suggest you start by finding a CZ sound that's at
    least vaguely similar to the DX patch you covet, look at its
    parameters,and try to alter them to get it closer.
    
    Good luck.  Sorry to be so "negative", but you've asked a *very*
    difficult question that only a joint CZ/DX expert could answer.
    
    len.
    

    re: .11--The DXs don't use a timevarying readout rate from a basic
    wavetable.
        John Sauter

    Good point: what is the patch?
    
    I have a CZ5000 too, and sorry, it doesnt have a sample and hold
    function. The typical use is to sample noise (or other signal) to
    produce a random values at a given (slowish) rate. Then this value
    is used to modulate timbre,pitch, or what_have_you.
    
    The only way I see to emulate this (if its a sample/hold into filter)
    is to use the sequencer to cycle through different user-defined presets.
    You'd dedicate all the definable banks to one voice, but each
    would have a different filter cutoff value (tone). The seq. rate
    
    And of course the big problem is that you'd have to sequence the
    the entire 'random' phrase instead of it happening while you play.
    Hmmm, that means that maybe you could get away with fewer than
    all 64 slots having different voices.
        
    Still, it may be a good patch, even tho it takes most of the machine!
    
    If the CZ were velocity sensitive, you could also
    put a computer between the kbd midi out and midi in, and have it
    modify your note messages with random (velocity) values at a 
    certain rate...
    
    Get the idea?
    ron
    
    

    The main characteristic of that patch is it's dependency on the
    LFO using the S/H waveform. (As opposed to triangle, square or saw)
    
    The patch itself just kinda bounces around 'randomly'.
    
    Edd

    John - the book you loaned me strongly implied that the DX family
    read out of a sine wavetable at a timevarying rate.  Maybe it was
    just an analogy to explain FM to the masses.
    
    The CZ has no random/noise waveform on the LFO.  Just sine, square
    (useful for trills) and saws (up and down).  You can get some FM-like
    effects by pitch modulating one DCO with its EG and using
    ring modulation to modulate the other DCO.  The LFO doesn't go high
    enough to get into real FM effects.

Let me reinforce some of the things that have been pointed out.

1.  The DX is a wavetable machine in the sense that it has a global sinewave
	stored in a wavetable that is used for each operator.

2.  The DX does have the Sample/Hold (random) LFO waveform.  The S/Hold
	patch on the DX uses this random LFO to change the output level of 
	a modulator in a random fashion, which alters the overall timbre
	of the sound generated in a random fashion.  It sounds kinda like
	an analog synth with a random voltage generator modulating the VCF.

Unless the CZ has some sort of random-somthing-or-other, you can't make it
do "Sample and Hold".  Hope this helps...

Todd.
	

  
    
    Thanks  for  your help.  You confrimed what I thought, if it 
    were possible it is still way beyond my  capabilities!  I'll 
    study  the  info you left, even if only to understand the CZ 
    better, but I'll leave option 2 open:-
     
    		Buy a DX ......     :-)
    
    	Peter

    BTW -
             The Sample and Hold patch uses the S/H LFO wave to effect
             1 operator only. Amplitude Modulation Sensitivity is the
             affected parameter...
    
                    LFO Wave    -     S/H
                    Speed             34
                    Sync              nah....
    
    Edd

    You CAN get a pseudo-sample/hold effect on the CZ
    if you *must* have it.
    
    Some restrictions, but it should be good enuf for
    an effect.
    
    See note 490.13
    
    Scheme, dont give up.
    
    ron
    

    
    I've got a CZ-1000 programming question.  One of the presets you get
    with the CZ-1000 is the familiar brass patch--on my keyboard it's
    Internal No.2.  The nicest thing about that patch is the reverb
    sound when you let off the keys.
    
    My problem is I can't for the life of me figure out what parameters
    are causing that reverb affect.  I'd like to add some reverb qualities
    to some other patches, but first I've got to identify which parameter
    creates that sound.  To make it a little easier, the brass patch
    I'm talking about uses only one line (1+1').  Obviously, the reverb
    sound is coming from one of the settings in line 1.
    
    But after analyzing and comparing those settings with other patches,
    I can't seem to find the which rates and level settings are causing
    the reverb when I lift off the keys.
    
    Any ideas?
    
    Peter

    Does the CZ use any permutation of the standard ADSR envelope?
    
    'Reverb' can be added by using a slower release rate...
    
    Edd

RE < Note 490.20 >

	Edd's got the idea.  Reverb can be simulated by causing the filter EG
to close earlier than the Amplifier EG.  This is explained in a programming
example in one of the books you get with the CZ (it is even talking about 
modifying a brass patch).

	Dan

RE: .22

   No swag, just fact.   8-)   That's a standard on most analog units
   that have seperate filter and amp envelopes.  Been doing the same
   thing on my Xa for years. 

   Another cute trick to make a non-velocity unit sound touch sensitive
   is to shut filter attack down to 0 for a given patch, then adjust amp
   attack (of ADSR, of course).  Try around 10-15% of wide open.  Should
   give you a pretty convincing percussive hit.

8^)

    How about posting the patch?  We can compare it with the CZ-101
    patch (assuming there's enough similarity and that the CZ-101 factory
    preset doesn't have the reverb...)
    
    Steve

    You want to look at the rate parameter for the envelope stage(s)
    after the SUStain stage.  Note that just closing the filter down
    faster than the amplitude (DCA) envelope will not result in "reverb" -
    the decay times need to be long enough to be perceived as reverb,
    probably greater than .1 second, out to as long as 2 seconds.  A
    2 second decay though will probably be useless, because neighboring
    (in time) tones will "overlap".  You can hold the filter envelope
    constant (on the CZs, that's the waveform envelope) after SUS
    and get a usable reverb, but you will get a more realistic effect
    by causing the high frequencies to decay faster (as Dan suggested,
    by having the filter envelope decay faster than the amplitude
    envelope).
    
    For those not familiar with CZ envelopes, they are 8 stage, with
    each stage having a rate and level parameter.  Rates and levels
    range over the values 0..99.  Two of the stages can be tagged
    as SUS(tain) and END.  On a NOTE-ON, the envelopes do their rate/
    level thing until they get to the SUS(tain) stage, at which level
    they remain until NOTE-OFF at which point they proceed through the
    rest of the stages until they get to END.  The level for END is
    always forced to be zero (you can't set it to anything else).  If
    a NOTE-OFF comes in before you reach the SUS(tain) level, the EG
    just jumps to the stage after SUS(tain).  These EGs allow you to
    define arbitrarily (well, within 8 stages) complex Attack, Decay
    and Release shapes (using traditional ADSR terminology).
    
    The low rate values correspond to very long times - I once clocked
    a CZ EG at over 90 seconds, and that wasn't the longest it was capable
    of!
    
    Also, the CZs allow you to apply envelopes to oscillator pitch (the
    level is pitch above the base pitch), oscillator waveform (0 ->
    sine wave, 99 -> no "smoothing" of basic waveform; you get to pick
    from 34 different waveforms for each oscillator, not all of which
    are usefully distinguishable) and amplitude.
    
    len.
     
    

    re:   CZ's and Atari-ST's
    	I contacted Casio (UK) last year, and they sent me a fairly
    detailed MIDI spec. From this, I was able to get some things going
    without much hassle. However, who wants to mess about, when you
    can buy cheap sequencers/notators ?
    	However, I did write a program which saves LOTS of hassle when
    manipulating a CZ from an Atari. As you may know, setting up the
    CZ to act as a multi-timbral expander (ie. MONO mode) is untidy
    (ie. cable-swapping, button-bashing, etc). So; from the MIDI spec
    (in particular, the CZ sys-exclusive formats), I put this program
    together with the function of assigning different (MONO) voices
    to particular channels. It means that you can set up your CZ's voices
    (8 on my CZ-3000) without getting up from your seat - mouse only
    required. It's fully GEM'd, so it's easy to use. If anybody wants
    a copy, then let me know. [It's available in some PD Librarys under
    the name - CZMONO; written in Modula-2].
    
    			Jim Burke (a Scotsman in Munich)
    
    so it's easy to use. 

    
    
    I'me afraid I gave up using my CZ in multi mode some time ago. I
    would however be interested in a PD patch librarian for Atari/Cz1000.
    
    Any clues anyone?
    
    					Paul.