From analogue-request@magnus.acs.ohio-state.edu Sat May 22 07:19:19 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA01089; Sat, 22 May 93 07:19:02 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from ohstvma.acs.ohio-state.edu by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA01083; Sat, 22 May 93 07:19:00 -0400
Date: Sat, 22 May 93 07:19:00 -0400
From: <RICK@SARA.NL>
Message-Id: <9305221119.AA01083@quark.magnus.acs.ohio-state.edu>
Received: from OHSTVMA.ACS.OHIO-STATE.EDU by OHSTVMA.ACS.OHIO-STATE.EDU (IBM VM SMTP R1.2.1MX) with BSMTP id 5889; Sat, 22 May 93 07:17:16 EDT
Received: from HASARA5.BITNET (NJE origin MAILGATE@HASARA5) by OHSTVMA.ACS.OHIO-STATE.EDU (LMail V1.1d/1.7f) with BSMTP id 5888; Sat, 22 May 1993 07:17:15 -0400
Received: from diamond.sara.nl by SARA.NL
          for analogue@magnus.acs.ohio-state.edu; 22 May 93 13:14 MET
Received: by diamond.sara.nl (5.61-AIX-1.2/1.0), id AA148622,
          (for analogue@magnus.acs.ohio-state.edu, from sscprick@diamond.sara.nl
Apparently-To: <analogue@MAGNUS.ACS.OHIO-STATE.EDU>
Status: OR

);
Message-Id: <9305221212.AA148622@diamond.sara.nl>
To: analogue@magnus.acs.ohio-state.edu
Subject: Elektor Vocoder
Date: Sat, 22 May 93 13:12:50 N
From: Rick Jansen <RICK%SARA.NL@OHSTVMA.ACS.OHIO-STATE.EDU>


On the subject of Vocoders: besides the Formant Elektor has published
quite a few electronic music projects to build yourself, also a
vocoder. Elektor designed it in cooperation with the Dutch company
Synton, who in those days (1980) marketed (a) vocoder(s).

It is a 10-band vocoder, one (24dB) filter print for each section,
plus an input filter and an output filter. A voiced/unvoiced detector
was published later, if I'm not mistaken. Provisions for later addition
of one were included anyway.

Rick Jansen
--
rick@sara.nl

  She's a Module and she's looking good

From analogue-request@magnus.acs.ohio-state.edu Fri May 21 04:31:58 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA15993; Fri, 21 May 93 04:31:29 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from SILVER.LCS.MIT.EDU by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA15988; Fri, 21 May 93 04:31:27 -0400
Received: by silver.lcs.mit.edu 
	id AA16198; Fri, 21 May 93 04:31:24 -0400
Date: Fri, 21 May 93 04:31:24 -0400
From: jna@silver.lcs.mit.edu (we can make it break it shake it to the limit)
Message-Id: <9305210831.AA16198@silver.lcs.mit.edu>
To: analogue@magnus.acs.ohio-state.edu
Subject: Re:  Vocoder Question
Status: OR

Could someone give me some more information on the PAIA vocoder? Here Joe L's
saying is wonderful, and I've had people tell me it's a piece of shit, lots of
noise and lots of hiss and lots of problems with the kit in general;

Now, i've seen a PAIA kit that a friend of mine was assembling, and it looked
easy as hell, even someone with nearly zero electronics experience could 
put some of their kits together from what it looked like (this was their 
hexfuzz/gutiar pre-amp kit..) 

Lemme know, because I've wanted a vocoder for YEARS.`

-john



From analogue-request@magnus.acs.ohio-state.edu Thu May 20 15:54:27 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA28688; Thu, 20 May 93 15:52:45 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from netcom.netcom.com by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA28682; Thu, 20 May 93 15:52:43 -0400
Received: by netcom.netcom.com (5.65/SMI-4.1/Netcom)
	id AA25927; Thu, 20 May 93 12:52:58 -0700
From: metlay@netcom.com (metlay)
Message-Id: <9305201952.AA25927@netcom.netcom.com>
Subject: Re:  Vocoder Question
To: locicero@YALE.EDU (Joe LoCicero)
Date: Thu, 20 May 93 12:52:57 PDT
Cc: analogue@magnus.acs.ohio-state.edu (Analogue Synth Mailing List)
In-Reply-To: <199305201827.AA26572@MINERVA.CIS.YALE.EDU>; from "Joe LoCicero" at May 20, 93 2:27 pm
X-Mailer: ELM [version 2.3 PL11]
Status: OR

Joe LoCicero sent this to me instead of the Analogue list.

>Speaking of vocoders, I own that PAiA vocoder and love it!  Does anyone
>have or is willing to type in that sheet of ideas by Craig Anderton?
>I'd love to see other things to do with it.  My favorite right now is
>to pump a bass line into the line input, and run my drum machine
>into the "mic" input, vocoding the bass line so that they really sync up!
>
>l8r,
>-- 
>---------------------------------------------------------------------
>-Joe LoCicero                         locicero@minerva.cis.yale.edu -
>---------------------------------------------------------------------

I have it. It's not just a list, it's a SOUNDSHEET, a flexidisc, a
real honest-to Pete sound recording, and a lot of fun to hear. I dunno
if I can get to it to type it up in a timely manner soon, tho....


-- 
mike metlay * atomic city * box 81175 pgh pa 15217-0675 * metlay@netcom.com
---------------------------------------------------------------------------
If this is a Global Village, then I must be the Global Village Idiot.......

From analogue-request@magnus.acs.ohio-state.edu Thu May 20 11:49:41 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA14908; Thu, 20 May 93 11:47:35 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from netcom.netcom.com by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA14903; Thu, 20 May 93 11:47:32 -0400
Received: by netcom.netcom.com (5.65/SMI-4.1/Netcom)
	id AA16208; Thu, 20 May 93 08:47:55 -0700
From: metlay@netcom.com (metlay)
Message-Id: <9305201547.AA16208@netcom.netcom.com>
Subject: Re:  Vocoder Question
To: dacc@cmp-rt.music.uiuc.edu (Andrew C. Crowell)
Date: Thu, 20 May 93 8:47:54 PDT
Cc: analogue@magnus.acs.ohio-state.edu
In-Reply-To: <9305191817.AA15884@cmp-rt.music.uiuc.edu>; from "Andrew C. Crowell" at May 19, 93 1:17 pm
X-Mailer: ELM [version 2.3 PL11]
Status: OR

Slightly less previously, DACCrowell wrote;
>Previously, Chris Meyer wrote:
>> 
>> In very simple terms (since I'm typing this on-line, and I'm sure DAC
>> will come up with a very complete answer himself), a vocoder splits the
>> incoming signal (usually a voice) into several frequency bands - kind of
>> like each slider on a graphic equalizer, if you want a way to visualize
>> it - by passing it through multiple band-pass filters tuned to different
>> frequencies. The output of each of these filters is fed to an envelope
>> follower, to create a control voltage that represents the relative
>> strength of the input signal inside each frequency band at any given
>> moment. A second set of band-pass filters, tuned to the same frequencies,
>> then divide up a second input signal - typically an instrument of some
>> sort. The CVs culled from the first bank of filters then controls the
>> amplitude of the outputs of the second bank, which are then mixed back
>> together. The result is the harmonic spectrum of the first signal being
>> imposed upon the second.
>
>BING! Correcto! Actually, a much better explanation than the one I put down,
>but one thing here: in the keyboard vocoders, such as the Korg MS-series
>one and the Roland device made so famous by Laurie Anderson, that second
>bank of filters is missing, because you're imposing the control signal
>directly on the output of the synth that co-exists in the cabinet with
>the vocoder device itself. In some cases, you can use the filtering in
>the synth section to work like this (such as in that Korg unit), but it's
>a pale imitation of the true, no-foolin' multiband effect.

True, alas, but for many years it was the only game in town. No manufacturer
was widely marketing a true carrier-and-modulator vocoder. Even in the MIDI
era, the only vocoder handy for a long time was Korg's DVP-1, whose internal
synth was barely deserving of the name. The big turnaround in attitudes came
in 1986 (?) when PAiA marketed a true vocoder (eight bands? ten?) and Craig
Anderton put out a soundsheet of all the cool things you could do with one.
Suddenly Bode and EMS vocoders were rare and costly, and soon after that
came the SE-50, followed by the WS A/D. An interesting example of a poke in
the market with a sharp stick. 

-- 
mike metlay * atomic city * box 81175 pgh pa 15217-0675 * metlay@netcom.com
---------------------------------------------------------------------------
If this is a Global Village, then I must be the Global Village Idiot.......

From analogue-request@magnus.acs.ohio-state.edu Thu May 20 03:12:53 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA00773; Thu, 20 May 93 03:12:33 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from DB3.SPEECH.CS.CMU.EDU by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA00768; Thu, 20 May 93 03:12:31 -0400
Received: from DB3.SPEECH.CS.CMU.EDU by DB3.SPEECH.CS.CMU.EDU id aa07577;
          20 May 93 3:11:45 EDT
To: Steven Collins <steve@crl.hitachi.co.jp>
Cc: analogue@magnus.acs.ohio-state.edu
Subject: dsp (was Re: Vocoder stuff), modulation, simmons tmi
In-Reply-To: Your message of "Thu, 20 May 93 11:45:16 +0200."
             <9305200245.AA28220@hcrlgw.crl.hitachi.co.jp> 
Date: Thu, 20 May 93 03:11:38 -0400
Message-Id: <7575.737881898@DB3.SPEECH.CS.CMU.EDU>
From: Yoshiaki_Ohshima@DB3.SPEECH.CS.CMU.EDU
Status: OR


aha, that's wild. see, usually time-varying property of the convolution
filter would better be smooth or at least slower by the magnitude than
the signal to be processed. you mentioned the use of EG or LFO, which makes
sense. but still, chopping the amplitude and the phase with variable
filter coefficients is wild, probably not pleasant--you are gonna introduce
lots of epoch discontinuities, that's seems a no-no thing :-) one big
difference between image and sound is some wacky effects on sound are
intolerable, even though they look perfectly reasonable when plotted.
anyway, this will open up some area of exploitations...

in a more conservative context, i have been tossing around an idea of 
modifying my favorite modulation effects such as the ring modulator, phase 
shifter, flanger, etc by giving more freedom to the modulator block. not a 
new idea, but i think it's not been fully explored yet. the inspiration
comes from the filter matrix switch of those electro-harmonix pedals,
--yes, i'm also a string instrument player, and love EH toys--, old
roland stereo phasor--remember that ugly 2u box?--, dimension-d, 
mutron bi-phase, roland phase-five, funny cat, and of course vocoder...

lastly, someone please help! i'm looking for the manual for simmons tmi.
i asked a friend of mine get it for me at "high tech consignment" in mn.
it cost me only $99. i've lost the original message, but to the person
who posted about that store, that's so much!! a great place to check out!

							--aki


From analogue-request@magnus.acs.ohio-state.edu Wed May 19 22:45:56 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA25864; Wed, 19 May 93 22:45:32 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from hitwide.hitachi.co.jp by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA25859; Wed, 19 May 93 22:45:29 -0400
Received: from [133.144.31.130] by hitwide.hitachi.co.jp (5.65/2.7W-HINOC)
	id AA21644; Thu, 20 May 93 11:45:19 +0900
Received: from hcrlgw.crl.hitachi.co.jp by hcrlgw92.crl.hitachi.co.jp (4.1/6.4J.6)
	id AA10871; Thu, 20 May 93 11:45:18 JST
Received: by hcrlgw.crl.hitachi.co.jp (4.0/6.4J.6)
	id AA28220; Thu, 20 May 93 11:45:16 JST
Date: Thu, 20 May 93 11:45:16 JST
From: steve@crl.hitachi.co.jp (Steven Collins)
Return-Path: <steve@crl.hitachi.co.jp>
Message-Id: <9305200245.AA28220@hcrlgw.crl.hitachi.co.jp>
To: analogue@magnus.acs.ohio-state.edu
Subject: Vocoder stuff
Status: O


>would you elaborate on your ideas a bit more, as it's not making
>clear sense to me so far. i'd like to hear more about it.

True, it did sound a little muddled.  

The idea is, when performing DSP, quite often you simply apply a convolution
matrix to the input digital waveform.  In video this is 2d so you'd use a
4x4 or so matrix as the convolution filter kernel.  This you know.  OK, in
sound, you generally have a 1D convolution vector, so how about a real-time
tweakable vector where the scalars in the vector are controlled by some
potentiometers.  Alternatively allow the possibility of using an LFO or EG
as input to one of the scalars.  Therefore you end up with a good deal of
control over the convolution that's applied to the input waveform.  I've no
idea whether this would be useful, it would really only serve as a very 
unusual filter.  Do an FFT on the waveform first and then filter using the
same technique to get other results.

I confess to not knowing very much about DSP, but intuitively these things
sound interesting to me.  I've played with vision algorithms, doind vision
stuff like edge detection/enhancement and smoothing.  I just wondered what
this would sound like in the sound domain.  Altering convultion kernels
applied to video images produces some wacky effects (in real-time using banks
of transputers), so I'd imagine that doing the same to an audio input should
be interesting also.  

Does this clear up the mess coming out of a 'heat oppressed brain'

steve

PS: dig the shakespeare quote eh?

---
+-------------------------------------+----------------------------------------+
| Steven Collins, Visiting Researcher | email: steve@crl.hitachi.co.jp         |
| Hitachi Central Research Lab. Tokyo | ph: (0423)-23-1111 fax: (0423) 27-7742 |
+-------------------------------------+----------------------------------------+

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 22:30:29 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA25548; Wed, 19 May 93 22:29:36 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from SPEECH1.CS.CMU.EDU by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA25542; Wed, 19 May 93 22:29:35 -0400
Received: from SPEECH1.CS.CMU.EDU by SPEECH1.CS.CMU.EDU id aa11016;
          19 May 93 22:28:38 EDT
To: Steven Collins <steve@crl.hitachi.co.jp>
Cc: analogue@magnus.acs.ohio-state.edu, Yoshiaki_Ohshima@SPEECH1.CS.CMU.EDU
Subject: Re: Vocoder & stuff 
In-Reply-To: Your message of "Thu, 20 May 93 10:14:35 +0200."
             <9305200114.AA25519@hcrlgw.crl.hitachi.co.jp> 
Date: Wed, 19 May 93 22:28:28 -0400
Message-Id: <11014.737864908@SPEECH1.CS.CMU.EDU>
From: Yoshiaki_Ohshima@SPEECH1.CS.CMU.EDU
Status: O


>LFOs and VCO etc.  It would be the ultimate filter, would require a reasonably
>fast DSP chip with DAC and ADC etc.  Anybody know of anything like this.

as i confessed in the previous note, dsp is my specialty.
would you elaborate on your ideas a bit more, as it's not making
clear sense to me so far. i'd like to hear more about it.

							--aki

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 21:15:24 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA23562; Wed, 19 May 93 21:15:05 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from hitwide.hitachi.co.jp by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA23547; Wed, 19 May 93 21:14:55 -0400
Received: from [133.144.31.130] by hitwide.hitachi.co.jp (5.65/2.7W-HINOC)
	id AA19487; Thu, 20 May 93 10:14:40 +0900
Received: from hcrlgw.crl.hitachi.co.jp by hcrlgw92.crl.hitachi.co.jp (4.1/6.4J.6)
	id AA04372; Thu, 20 May 93 10:14:37 JST
Received: by hcrlgw.crl.hitachi.co.jp (4.0/6.4J.6)
	id AA25519; Thu, 20 May 93 10:14:35 JST
Date: Thu, 20 May 93 10:14:35 JST
From: steve@crl.hitachi.co.jp (Steven Collins)
Return-Path: <steve@crl.hitachi.co.jp>
Message-Id: <9305200114.AA25519@hcrlgw.crl.hitachi.co.jp>
To: analogue@magnus.acs.ohio-state.edu
Subject: Vocoder & stuff
Status: O


I've always wanted to play around with marrying DSP
and analogue technology a lot more.  I'm sure there's loads more stuff can be
done, like what about other DSP operations akin to vision's Edge Detectors,
Sobel Operators and other convulations.  In fact it would be fun to setup a
matrix panel, say 4x4, have each represented by a pot with a DAC, use this to
real-time alter the convolution matrix, and allow this to operate on a carrier
signal and see what you get.  In fact, the matrix panel could be fed from
LFOs and VCO etc.  It would be the ultimate filter, would require a reasonably
fast DSP chip with DAC and ADC etc.  Anybody know of anything like this.

steve
---
+-------------------------------------+----------------------------------------+
| Steven Collins, Visiting Researcher | email: steve@crl.hitachi.co.jp         |
| Hitachi Central Research Lab. Tokyo | ph: (0423)-23-1111 fax: (0423) 27-7742 |
+-------------------------------------+----------------------------------------+

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 20:55:58 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA23039; Wed, 19 May 93 20:55:29 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from netcom3.netcom.com by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA23031; Wed, 19 May 93 20:55:27 -0400
Received: by netcom3.netcom.com (5.65/SMI-4.1/Netcom)
	id AA26305; Wed, 19 May 93 17:56:02 -0700
Date: Wed, 19 May 93 17:56:02 -0700
From: cetacean@netcom.com (Michael O'Hara)
Message-Id: <9305200056.AA26305@netcom3.netcom.com>
To: analogue@magnus.acs.ohio-state.edu, steve@crl.hitachi.co.jp
Subject: Re:  Re vocoder
Status: O

>..wave-shapes.  It would be a form of re-synthesis real-time of course and
>would be analog.  These could then be passed through some VCFs and stuff.
>I'd love to build something like this to see what it would sound like.

>steve

Exactly.. when I finish up my video project, I may just attempt to implement
one.  I, too, also really liked that clockwork orange sound.  I'll name
the company "ultra vlolent sound, inc" (just joking.. how about ultra violet)?

:)

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 20:31:44 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA22372; Wed, 19 May 93 20:31:17 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from hitwide.hitachi.co.jp by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA22367; Wed, 19 May 93 20:31:13 -0400
Received: from [133.144.31.130] by hitwide.hitachi.co.jp (5.65/2.7W-HINOC)
	id AA18912; Thu, 20 May 93 09:31:03 +0900
Received: from hcrlgw.crl.hitachi.co.jp by hcrlgw92.crl.hitachi.co.jp (4.1/6.4J.6)
	id AA01069; Thu, 20 May 93 09:31:01 JST
Received: by hcrlgw.crl.hitachi.co.jp (4.0/6.4J.6)
	id AA24426; Thu, 20 May 93 09:30:59 JST
Date: Thu, 20 May 93 09:30:59 JST
From: steve@crl.hitachi.co.jp (Steven Collins)
Return-Path: <steve@crl.hitachi.co.jp>
Message-Id: <9305200030.AA24426@hcrlgw.crl.hitachi.co.jp>
To: analogue@magnus.acs.ohio-state.edu
Subject: Re vocoder
Status: O


>I've always wondered what a vocoder would sound like if it had a lot more
>bands than a typical design.  This sounds like fertile territory for a DSP
>chip.  I can imagine doing an FFT on the voice input and feeding the levels
>into some kind of equalizer.

I often wondered myself, what the effect would be if you took a signal, used
an FFT to split into regular frequency bands, say about 10-15, and fed the
amplitude and phase info. to 10-15 separate oscillators, each with various
wave-shapes.  It would be a form of re-synthesis real-time of course and
would be analog.  These could then be passed through some VCFs and stuff.
I'd love to build something like this to see what it would sound like.

steve

---
+-------------------------------------+----------------------------------------+
| Steven Collins, Visiting Researcher | email: steve@crl.hitachi.co.jp         |
| Hitachi Central Research Lab. Tokyo | ph: (0423)-23-1111 fax: (0423) 27-7742 |
+-------------------------------------+----------------------------------------+

From cook@stout.atd.ucar.EDU Wed May 19 20:19:49 1993
Received: from ncar.ucar.edu by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA21943; Wed, 19 May 93 20:19:48 -0400
Received: from stout.atd.ucar.edu by ncar.ucar.EDU (5.65/ NCAR Central Post Office 03/11/93)
	id AA07401; Wed, 19 May 93 18:19:45 MDT
Message-Id: <9305200019.AA29980@stout.atd.ucar.EDU>
Received: from bitter.atd.ucar.edu by stout.atd.ucar.EDU (5.65/ NCAR Mail Server 04/10/90)
	id AA29980; Wed, 19 May 93 18:19:44 MDT
To: analogue-request@magnus.acs.ohio-state.edu
Subject: Re: Vocoder Question
Date: Wed, 19 May 1993 18:19:43 -0600
From: Forrest Cook <cook@stout.atd.ucar.EDU>
Status: O

Yoshiaki_Ohshima says:
>real time implementation of arbitrary filter shapes:
>real time implementation is possible only if the transfer function is 
>causal. the answer is no.

How about building an "n" band graphic equalizer like a lot of DSP
effect boxes have in them and modulating them with fft info.
Not quite as nice as the smooth arbitrary filter shape idea but
probably possible without any heavy magic.  That could be done with
fixed band pass filters where the output level is modulated, essentially
the same model as the analog vocoder.

I hope this thread isn't straying too far away from the analog synth
domain, it is interesting.

Forrest Cook
cook@stout.atd.ucar.edu WB0RIO
{husc6|rutgers|ames|gatech}!ncar!stout!cook

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 19:24:12 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA20100; Wed, 19 May 93 19:20:00 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from SPEECH1.CS.CMU.EDU by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA20095; Wed, 19 May 93 19:19:59 -0400
Received: from SPEECH1.CS.CMU.EDU by SPEECH1.CS.CMU.EDU id aa10478;
          19 May 93 19:19:40 EDT
To: analogue@magnus.acs.ohio-state.edu
Subject: Re: Vocoder Question 
In-Reply-To: Your message of "Wed, 19 May 93 12:32:40 CDT."
             <9305191732.AA15753@cmp-rt.music.uiuc.edu> 
Date: Wed, 19 May 93 19:19:36 -0400
Message-Id: <10476.737853576@SPEECH1.CS.CMU.EDU>
From: Yoshiaki_Ohshima@SPEECH1.CS.CMU.EDU
Status: O


>        Vocoders were originally developed for use in the communications
>industry, but in 1967, Alvin Lucier made the first use of a vocoder (albeit
                                              ^^^^^ ^^^

right, what was ingeneous about was to "misuse" the vocoder for the
musical purposes, as there was nothing new about the vocoder itself.

this happens to be the area i work for living, and i feel we should bring 
it into perspective. the vocoder (or to be more correct, the "channel
vocoder") is the oldest form of speech coding device (hence vo-coder), 
and was invented by dudley.

h. dudley, "the vocoder," bell labs record, vol 17, pp122-126, 1939

i think this is important to recognize, as his design and principle
were used with no fundamental modifications. at least, this is a norm
among researchers in communications and speech technology.


now in the attempt to answer some of previously made questions:

number of filters:

as a codec, how capable it's of restoring the original signal depends
solely upon whether it satisfies the principle of filter bank synthesis, 
having mathematically nothing to do with the number of channels. 
   as a wacky signal processor for music making, it's a different story
and now a matter of timbre perception. research says human ears are 
capable of differentiating timbre changes from about the seventh harmonics 
of the findamentals of complex tones. in general, densely spaced BPFs 
would be prefered, but mind you are convolving the spectrum of an musical 
instrument with the transfer function of a vocal tract that is slowly varying.

real time implementation of arbitrary filter shapes:

real time implementation is possible only if the transfer function is 
causal. the answer is no.


							--aki

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 18:19:07 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA17887; Wed, 19 May 93 18:17:34 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from academ07.mty.itesm.mx by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA17881; Wed, 19 May 93 18:17:32 -0400
Received: by academ07.mty.itesm.mx id AA37854
  (5.65c/IDA-1.4.4 for analogue@magnus.acs.ohio-state.edu); Wed, 19 May 1993 16:14:43 -0600
From: Max Enrique Sequeira Garza <al160184@academ07.mty.itesm.mx>
Message-Id: <199305192214.AA37854@academ07.mty.itesm.mx>
Subject: SH-09 manual? anyone?
To: rogoff@teradyne.com (David Rogoff x4627)
Date: Wed, 19 May 1993 16:14:42 +22335737 (CST)
Action: b
In-Reply-To: <9305181555.AA19543@midas.teradyne.com> from "David Rogoff x4627" at May 18, 93 08:55:56 am
X-Mailer: ELM [version 2.4 PL21]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 152       
Status: O

    I was wondering if anyone new how i could get hold of a SH-09 manual. I'm
gonna buy one used and would defenitely want it!
    Thanks!

        MAX

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 15:10:57 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA07586; Wed, 19 May 93 15:07:28 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from att-out.att.com by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA07570; Wed, 19 May 93 15:07:21 -0400
From: wbf@aluxpo.att.com
Received: from alux1.cnet.att.com by aluxpo (4.1/DCS-aluxpo-040293)
	id AA19049; Wed, 19 May 93 15:05:51 EDT
Received: by alux1.cnet.att.com (4.1/DCS-aluxpo_client-022293)
	id AA02074; Wed, 19 May 93 15:05:51 EDT
Date: Wed, 19 May 93 15:05:51 EDT
Original-From: aluxpo!wbf (William Fox)
Message-Id: <9305191905.AA02074@alux1.cnet.att.com>
To: analogue@magnus.acs.ohio-state.edu
Subject: Re: Vocoder Correction
Status: O

David Rogoff on my vocoder ASCIImatic:
> Minor note (unless you tried to build one): all of your
> LPFs should be BPFs.

	Thanks for saying that it is a MINOR note.  (I thought only
chords could be minor!)  Of course, you are right, I did mean band pass
filter not low pass.  I was merely typing one thing and thinking 12
other things.  (A thought row?)

Bill

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 14:21:41 1993
Received: by magnusug.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA07826; Wed, 19 May 93 14:16:45 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from cmp-rt.music.uiuc.edu by magnusug.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA07821; Wed, 19 May 93 14:16:42 -0400
Received: by cmp-rt.music.uiuc.edu (AIX  2.1 2/4.03)
          id AA15884; Wed, 19 May 93 13:17:34 CDT
From: dacc@cmp-rt.music.uiuc.edu (Andrew C. Crowell)
Message-Id: <9305191817.AA15884@cmp-rt.music.uiuc.edu>
Subject: Re:  Vocoder Question
To: analogue@magnus.acs.ohio-state.edu
Date: Wed, 19 May 93 13:17:33 CDT
In-Reply-To: <199305191657.AA15608@well.sf.ca.us>; from "Chris Meyer" at May 19, 93 9:57 am
X-Mailer: ELM [version 2.3 PL5]
Status: O

Previously, Chris Meyer wrote:
> 
> In very simple terms (since I'm typing this on-line, and I'm sure DAC
> will come up with a very complete answer himself), a vocoder splits the
> incoming signal (usually a voice) into several frequency bands - kind of
> like each slider on a graphic equalizer, if you want a way to visualize
> it - by passing it through multiple band-pass filters tuned to different
> frequencies. The output of each of these filters is fed to an envelope
> follower, to create a control voltage that represents the relative
> strength of the input signal inside each frequency band at any given
> moment. A second set of band-pass filters, tuned to the same frequencies,
> then divide up a second input signal - typically an instrument of some
> sort. The CVs culled from the first bank of filters then controls the
> amplitude of the outputs of the second bank, which are then mixed back
> together. The result is the harmonic spectrum of the first signal being
> imposed upon the second.

BING! Correcto! Actually, a much better explanation than the one I put down,
but one thing here: in the keyboard vocoders, such as the Korg MS-series
one and the Roland device made so famous by Laurie Anderson, that second
bank of filters is missing, because you're imposing the control signal
directly on the output of the synth that co-exists in the cabinet with
the vocoder device itself. In some cases, you can use the filtering in
the synth section to work like this (such as in that Korg unit), but it's
a pale imitation of the true, no-foolin' multiband effect.

D.A.C. Crowell
Computer Music Project/School of Music
University of Illinois at Urbana/Champaign
(dacc@cmp-rt.music.uiuc.edu)


From analogue-request@magnus.acs.ohio-state.edu Wed May 19 14:10:16 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA04044; Wed, 19 May 93 14:07:15 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from att-out.att.com by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA04036; Wed, 19 May 93 14:07:11 -0400
From: wbf@aluxpo.att.com
Received: from alux1.cnet.att.com by aluxpo (4.1/DCS-aluxpo-040293)
	id AA09153; Wed, 19 May 93 14:05:03 EDT
Received: by alux1.cnet.att.com (4.1/DCS-aluxpo_client-022293)
	id AA24733; Wed, 19 May 93 14:05:02 EDT
Date: Wed, 19 May 93 14:05:02 EDT
Original-From: aluxpo!wbf (William Fox)
Message-Id: <9305191805.AA24733@alux1.cnet.att.com>
To: analogue@magnus.acs.ohio-state.edu
Subject: Re: Vocoder Question
Status: O

>Date: Wed, 19 May 93 09:06:16 PDT
>From: att!magnus.acs.ohio-state.edu!analogue-request
>(Kin Blas) (kin@guam.island.COM)
>To: analogue@magnus.acs.ohio-state.edu
>Subject: Vocoder Question
>
>Hi,
>
>	I hate to sound stupid, but, what is a vocoder?
>--== Kin Blas ==--
>kin@island.com

Hi Kin,

	You don't sound stupid.  There's something you don't know and
you're smart enough to ask a question.  Here's an ASCIImatic block
diagram that I hope will help.  The first n filters are fed audio
signal 1.  The second n filters are fed audio signal 2.  The envelope of
signal 2 is superimposed onto signal 1 at each of the center frequencies
1 through n.
                                                     |-----|
LPF(1)   ----------------> VCA --------------------->|     |
                            ^                        |  M  |
                            |                        |     |
LPF(2)   ----------------------> VCA --------------->|  I  |
  .                         |     ^                  |     |
  .                         |     |                  |  X  |---> OUT
LPF(n-1) ----------------------------> VCA --------->|     |
                            |     |     ^            |  E  |
                            |     |     |            |     |
LPF(n)   ----------------------------------> VCA --->|  R  |
                            |     |     |     ^      |-----|
                            |     |     |     |
LPF(1)   ---> EnvFol -------|     |     |     |
                                  |     |     |
                                  |     |     |
LPF(2)   ---> EnvFol -------------|     |     |
  .                                     |     |
  .                                     |     |
LPF(n-1) ---> EnvFol -------------------|     |
                                              |
                                              |
LPF(n)   ---> EnvFol -------------------------|

Bill Fox

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 13:35:01 1993
Received: by bottom.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA15661; Wed, 19 May 93 13:31:53 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from cmp-rt.music.uiuc.edu by bottom.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA15655; Wed, 19 May 93 13:31:50 -0400
Received: by cmp-rt.music.uiuc.edu (AIX  2.1 2/4.03)
          id AA15753; Wed, 19 May 93 12:32:42 CDT
From: dacc@cmp-rt.music.uiuc.edu (Andrew C. Crowell)
Message-Id: <9305191732.AA15753@cmp-rt.music.uiuc.edu>
Subject: Re: Vocoder Question
To: analogue@magnus.acs.ohio-state.edu
Date: Wed, 19 May 93 12:32:40 CDT
In-Reply-To: <9305191606.AA07326@guam.island.com>; from "Kin Blas" at May 19, 93 9:06 am
X-Mailer: ELM [version 2.3 PL5]
Status: O

Previously, Kin Blas wrote:
> 
> Hi,
> 
>   I hate to sound stupid, but, what is a vocoder?

Not a stupid question, actually, as the use of them is kind of unusual
these days. A vocoder is a device that uses a "control" signal (in many
cases, a voice, hence the name) to control a series of parameters that
processes and controls a "carrier" signal. The resulting sound is that of the
carrier modulated by the control.
	Vocoders were originally developed for use in the communications
industry, but in 1967, Alvin Lucier made the first use of a vocoder (albeit
a digital one, a unit manufactured by Sylvania for use in telecommunications)
in a musical work, his "North American Time Capsule 1967". The device
associated with the "vocoder effect" in music, however, traces its origins
back to the work of Walter Carlos and Bob Moog, and their work with an
analog equivalent of this device called a "spectrum follower" was first
heard on the "Clockwork Orange" soundtrack. The spectrum follower was the
device that took an inputted signal from an external source and, using an
array of envelope followers which had been filtered to track only very
specific narrow-frequency bandwidths, it sent voltage information that
"followed" an inputted "control" signal to the other modules in Carlos's
large modular Moog system, frequently the filters and the envelope
generators.
	It was later on that you finally began to see commercially-produced
vocoders. Originally, they were separate devices that took inputs of both
carrier and control signals and were strictly dedicated to the processing
task. The Electro-Harmonix multiband vocoder and the Sennheiser unit used
by Kraftwerk so often in the mid-70s were devices of this type. However,
thanks to artists such as Kraftwerk, one began to see companies making
units that were able to generate their own internal carrier signals, and
only accepted external control signals, as a rule. Roland made a couple
(I believe) such units, and there was a very excellent example of this in
Korg's semi-modular MS-series synths. However, these later devices would
only track a few, or sometimes one, bandwidth, so the quality of the
effect obtained by some of these later models was rather lacking in
comparison to the lushness of the units which dealt with a larger amount
of discrete bands.
	But at this point, we've sort of come full-circle. A goodly amount
of digital processors now are vocoder-capable devices (such as certain
Lexicon devices, certain Zoom devices, as well as some synth models like
the Korg Wavestation A/D) and that brings us back to where we started,
in the digital domain, but this time emulating an effect developed out in
the analog world.
	Hope that's helpful!

D.A.C. Crowell
Computer Music Project/School of Music
University of Illinois at Urbana/Champaign
(dacc@cmp-rt.music.uiuc.edu)

-- 

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 13:32:22 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA01986; Wed, 19 May 93 13:30:42 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from ncar.ucar.edu by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA01964; Wed, 19 May 93 13:30:38 -0400
Received: from stout.atd.ucar.edu by ncar.ucar.EDU (5.65/ NCAR Central Post Office 03/11/93)
	id AA00611; Wed, 19 May 93 11:30:32 MDT
Message-Id: <9305191730.AA18694@stout.atd.ucar.EDU>
Received: from bitter.atd.ucar.edu by stout.atd.ucar.EDU (5.65/ NCAR Mail Server 04/10/90)
	id AA18694; Wed, 19 May 93 11:30:31 MDT
To: analogue@magnus.acs.ohio-state.edu
Subject: vocoder
Date: Wed, 19 May 1993 11:30:30 -0600
From: Forrest Cook <cook@stout.atd.ucar.EDU>
Status: O

I've always wondered what a vocoder would sound like if it had a lot more
bands than a typical design.  This sounds like fertile territory for a DSP
chip.  I can imagine doing an FFT on the voice input and feeding the levels
into some kind of equalizer.   I don't know if an arbitrary equilization
curve can be generated and applied in real-time with the current state of
DSP software, but it sounds intriguing.  Has anyone worked on this problem?

Forrest Cook
cook@stout.atd.ucar.edu WB0RIO
{husc6|rutgers|ames|gatech}!ncar!stout!cook

From analogue-request@magnus.acs.ohio-state.edu Wed May 19 13:00:25 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA29929; Wed, 19 May 93 12:57:28 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from nkosi.well.sf.ca.us by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA29916; Wed, 19 May 93 12:57:25 -0400
Received: from well.sf.ca.us by nkosi.well.sf.ca.us with SMTP id AA04235
  (5.65c/IDA-1.5 for <analogue@magnus.acs.ohio-state.edu>); Wed, 19 May 1993 09:57:18 -0700
Received: by well.sf.ca.us (5.65c/SMI-4.1/well-930413-1)
	id AA15608; Wed, 19 May 1993 09:57:01 -0700
Date: Wed, 19 May 1993 09:57:01 -0700
From: Chris Meyer <cybpunk@well.sf.ca.us>
Message-Id: <199305191657.AA15608@well.sf.ca.us>
To: analogue@magnus.acs.ohio-state.edu, kin@guam.island.com
Subject: Re:  Vocoder Question
Status: OR

In very simple terms (since I'm typing this on-line, and I'm sure DAC
will come up with a very complete answer himself), a vocoder splits the
incoming signal (usually a voice) into several frequency bands - kind of
like each slider on a graphic equalizer, if you want a way to visualize
it - by passing it through multiple band-pass filters tuned to different
frequencies. The output of each of these filters is fed to an envelope
follower, to create a control voltage that represents the relative
strength of the input signal inside each frequency band at any given
moment. A second set of band-pass filters, tuned to the same frequencies,
then divide up a second input signal - typically an instrument of some
sort. The CVs culled from the first bank of filters then controls the
amplitude of the outputs of the second bank, which are then mixed back
together. The result is the harmonic spectrum of the first signal being
imposed upon the second.
 - CM

From analogue-request@magnus.acs.ohio-state.edu Mon May 24 04:44:28 1993
Received: by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA02177; Mon, 24 May 93 04:42:13 -0400
Errors-To: analogue-request@magnus.acs.ohio-state.edu
Sender: analogue-request@magnus.acs.ohio-state.edu
Received: from sihp03.si.estec.esa.nl by quark.magnus.acs.ohio-state.edu (5.65/3.910213)
	id AA02172; Mon, 24 May 93 04:42:07 -0400
Message-Id: <9305240842.AA02172@quark.magnus.acs.ohio-state.edu>
Received: by sihp03.si.estec.esa.nl
	(16.7/16.2) id AA05783; Mon, 24 May 93 10:48:41 +0100
From: Andrea TONI <andrea@sihp03.si.estec.esa.nl>
Subject: -- vocoder -- 
To: analogue@magnus.acs.ohio-state.edu
Date: Mon, 24 May 93 10:48:40 MET
Organization:   ESA-ESTEC (European Space Research and Technology Center)
                Keplerlaan 1	2200 AG  Noordwijk (EUROPA)		
Phone: 	       	Int +31 1719 83606
Fax: 	       	Int +31 1719 84697
Machine: 	HP-UX sihp03 A.B8.05 A 9000/730 941406112	
Mailer: Elm [revision: 66.33]
Status: O

Hi, 
I found this in my archive .. hope it helps .. 


----------------------------------------------------------------------------
        A vocoder is a device that combines the frequency distribution of
one signal with the waveform of another to produce a single output signal.
        One way to think of it is as a spectrum analyzer that drives a
graphic equalizer.

                        +---------------------+
                        | . . . * . . . . * . |
Spectrum-control signal | . * * * . . . * * . |
----------------------->| * * * * * . * * * * |
                        | * * * * * * * * * * | Control
                        | * * * * * * * * * * |>=| signals
                        +---------------------+ ||
                           Spectrum analyzer    ||
                                                ||
                              Graphic equalizer vv
                            +---------------------+
                            | . . . = . . . . = . |
Waveform-control signal     | . = = . . . . = . . | Vocoded signal
--------------------------->| = . . . = . = . . = |--------------------->
                            | . . . . . = . . . . |
                            | . . . . . . . . . . |
                            +---------------------+

The frequency bands on the equalizer are slaved to the corresponding frequency
bands on the analyzer...so that if the high-frequency content of the
spectrum-control signal suddenly goes up, the high end of the equalizer is
instantly boosted a corresponding amount, and you end up hearing more of the
high end of the waveform-control signal.
        Possibly the most common use of a vocoder involves a microphone
plugged into the spectrum-control input and something else (synth, guitar,
or anything else with an appropriately broad frequency spectrum) into the
waveform-control input.  Since in a lot of (non-Oriental) languages,
spoken words depend more on dynamic filtering than on pitch, you can speak
into the microphone in such a system and, by imposing the spectral character-
istics of your voice on the output from a wildly-fuzzed electric guitar, make
the guitar seem to sing words.
        However, Walter/Wendy Carlos (and probably others) has employed vocoders
in a number of pieces that weren't used for anything remotely vocal.
        A more flexible arrangement than the one illustrated above would
allow you to move the center frequencies of the analyzer and equalizer
bands, so that, for instance, you could modulate the entire 20-20KHz
frequency range of the equalizer with only the lower half (20-10KHz) of the
spectrum of the input signal, with higher-frequency information being
discarded.  An even more flexible arrangement would allow you to change
the control connections between analyzer and equalizer frequency bands--
so that you could reverse them, and have high-frequency material control
low-frequency equalization, and vice versa.  (Wonder what that would sound
like.  Any of you transistor jockeys out there have the equipment (and the
motivation) to try it?)
        You can get almost-intelligible basic vocal sounds ("oo", "ah", "ow",
etc.) with a cheap five- or seven-band vocoder, but to get truly understandable
speech ("Fourscore and seven years ago..."), you need at least ten or so bands,
and if you're really whacked-out about control you can go as high as thirty
or more.  I've never built one, but I would assume that you could probably do
a fair job with an expensive spectrum analyzer, one of those mono rackmount
GEQs with scads of bands, and a whole raft of VCAs.
        I may not know what I'm talking about here, but I'd bet, from the
advertisements of the Roland Vocalist that claim non-Chipmunk harmonies and
the fact that it also offers a vocoder, that that's basically what the Vocalist
is--a many-band vocoder with an on-board five-voice (or is it four?) tone
generator that produces a sound with a nice fat spectrum of harmonics to feed
through the EQ.  At least, that's how *I'd* do it, given those specifications
and a couple of burly, musclebound DSPs.  In other words, there'd be no
traditional pitch-shifting done at all, although depending on the implemen-
tation there may be some pitch-extraction going on.

Shalom,
Dan Wiebe
------------------------------------------------------------------------------

... And now the question .. 
Has anybody try the vocoder on the Boss SE70 ??? 
Actually has anybody ever seen the BOss Se70 in the shops ?? 
Street price ?? 

Ciao, 
------------------------------------------------ENV---ENV---ENV----------------
 Andrea TONI (andrea@sihp03.si.estec.esa.nl)     |     |     |
 Planetary and Space                            VCO-->VCF-->VCA--> DUCATI 900SS
 Science Division (SI)                           |     |     |
------------------------------------------------LFO---LFO---LFO----------------

