Subject: MAX Digest - 21 Dec 1999 to 22 Dec 1999 (#1999-365)
Date: Thu, 23 Dec 1999 00:00:10 -0500
Automatic digest processor <LISTSERV@LISTS.MCGILL.CA>
Reply-To: MAX - Interactive Music/Multimedia Standard Environments <MAX@LISTS.MCGILL.CA>
To: Recipients of MAX digests <MAX@LISTS.MCGILL.CA>

There are 12 messages totalling 432 lines in this issue.

Topics of the day:

  1. MIDI Pitchbend Controller info
  2. Question for Max Guru's (3)
  3. Knock around the clock
  4. floating among the points
  5. MAX installation
  6. step sequencer
  7. Max/Msp for New Year's Eve? (2)
  8. Multi-Touch Controller News
  9. MAX Digest - 20 Dec 1999 to 21 Dec 1999 (#1999-364)


Date:Wed, 22 Dec 1999 01:23:29 EST
Subject: MIDI Pitchbend Controller info

In a message dated 12/22/99 5:02:03 AM, LISTSERV@LISTS.MCGILL.CA writes:

>Date:Tue, 21 Dec 1999 11:34:22 +0000
david stevens <david@RESONANT.DEMON.CO.UK>
>Subject: strip controller
>i have a vague memory (well, loads of them actually) that someone once
>marketed a midi pitchbend strip that could be fixed to whatever
>controller keyboard you used. this would solve a problem for me right
>now - does anyone know of such a thing????

Check out Kurzweil's ExpressionMate -- recently released after several years
of vaporware. It has programmable zones, an advanced arpegiator, and
additional features. I believe there is a description of it in the Kurzweil
section of YoungChang's website.



to start -- contact me directly if unsuccessful and I'll try to locate better

R B Meng


Date:Wed, 22 Dec 1999 02:30:29 -0500
From:Stephen Kay <sk@COMPUSERVE.COM>
Subject: Question for Max Guru's

> My thought was that I could have every trigger within the installation
> send a bang whenever it is modified to the IF object. Then, hopefully
> the IF object could somehow bang the loadbang if it recieves no bangs
> for X seconds.

All you need is the 'clocker' object.

[bang]-->[clocker]-->[sel 300000]

Have every controller/trigger movement send something through the receive=
bang the clocker, which will restart its count at 0. It then counts
ms. If it reaches 300,000 (5 minutes) or whatever desired value, have it=

reset all your controls.

Stephen Kay=


Date:Wed, 22 Dec 1999 09:37:37 +0100
From:Jeffrey Burns <jeff@BERLIN.SNAFU.DE>
Subject: Knock around the clock

Dear Nick Rothwell,

Thanks for your reply. Sorry if I mixed two separate issues. In fact,
my question about sync machines had nothing to do with the one about
midi files.

What I still need is a way to do algorithmic composition, for example
with Detonate, write the result as a midi file to disc and open it in
score form with bar lines and meter changes. I don't think a
sequencer program will help, since the composition is not being done
in real time. Of course, I can already do this without meter changes,
but I don't know how to get rid of the default values of 120MM and
2/4 that Max puts into midi files.

Jeff Burns


Date:Wed, 22 Dec 1999 01:25:08 -0800
From:jhno <ear@SIRIUS.COM>
Subject: floating among the points

sorry for the length of this message - it is a little technical and
esoteric, but possibly relevant to subtle issues of digital sound quality
and character.

msp uses the 32-bit IEEE 754 floating point representation provided by the
power pc processor. the interesting thing about this format, from an audio
perspective, is that precision varies depending on the magnitude of the
number you are encoding. this is because a floating point number is
specified by a "fraction", which is raised to an "exponent" to produce the
number you are trying to describe.

the practical upshot of this is that precision is greatest near zero, and
diminishes as you count upward. for example, you can precisely specify over
8 million real numbers between 0.0000001 and 0.0000002 - but only two
between 8,388,607 and 8,388,608. larger than that, the precision is so
coarse that the difference between consecutive real numbers is greater than

of course, this is not usually relevant in msp, which utilizes the range
between 0. and 1. to represent audio signals (it can become relevant in
certain situations, like sample counting, however). fortunately, half the
real numbers you can represent in floating point lie in this range - 30
bits' worth (31 if you go from -1 to 1). however, the precision within this

range is not constant.

the interesting thing is that the increase in precision as you approach
zero does not occur smoothly... it occurs in steps, at each negative power
of 2. so, floating point real numbers are equally spaced between 1 and 0.5,
but then the precision changes and they are twice as dense between 0.5 and
0.25. floating point can specify exactly the same number of unique numbers
between 0.5 and 0.25 as it can between 0.5 and 1.

IEEE 754 specifies a 23-bit fraction and an 8-bit exponent. this results in
the following comparison to integer systems: at signal values between 0.5
and 1, you are working with precision analogous to a 24-bit integer
representation. at lower signal levels, precision increases dramatically,
until near zero it would take a 152-bit integer system to provide the same

so, i think of integer space as a linear progression of equally spaced
values, and floating point space as a continuum with much greater span, but
a density of values that varies logarithmically in comparison to integer
numbers. and, this variation occurs in steps, or tiers, not smoothly.

(then again, numbers are arbitrary - you could imagine floating point
specified numbers to be the real reference points, from which perspective
integer space would look kinked...)

anyway, i'd be interested to know if anyone has any thoughts on practical
implications of these concepts. aphex twin told me he could hear the
difference between integer and floating point math. i doubt this was a
double-blind test, but i can imagine that these differences would be

note that i am not questioning the use of floating point numbers in msp -
the sound quality is excellent, and the convenience of floating point math
is part of the program's unmatched elegance. i am just curious about the
way our ears and brains work.

i learned a little about IEEE 754 from this little tutorial:

it has a typo in one of the examples, which i will leave as an exercise to
the reader.

i wrote some little max patches to translate among floating point, two's
complement, and binary representations... pretty interesting to check out
the numbers and investigate the limits of these systems. it is tweaky


() ))(((( ))) ))))) ( )((()) (( ))( )) (((( )(()( (()
san francisco, ca


Date:Wed, 22 Dec 1999 09:00:35 EST
Subject: MAX installation

I am trying to install MAX from the 3.5 CD on to a Powerbook G3 400 mhz with
OS9. When I try to open MAX the computer freezes. Does anyone know what is
going on? I've paired down to only the necessary extensions.



Date:Wed, 22 Dec 1999 11:38:10 +0000
From:Robert Henke <robert@MONOLAKE.DE>
Subject: step sequencer

... a little xxxmas present for the friends of repetitive music:


the secret life of digital music


Date:Wed, 22 Dec 1999 08:45:52 -0600
From:"Timothy A. Place" <tplace@CCTR.UMKC.EDU>
Subject: Re: Max/Msp for New Year's Eve?


These folks are from the Peabody Conservatory, where I studied last year.
They are indeed using Max. Forrest has developed some amazing patches to
use in conjunction with the Buchla Lightning (listed in the article as
infrared batons). In the past this group has always used a Kurzweil K2000
for the actual sound, not MSP. Though I wouldn't rule out that they may
have gone this direction in the past six months - my suspicion is that they
will still be using the K2000 controlled by Max.

- Tim

>From:Nathan Chase <zampa_@HOTMAIL.COM>
>Subject: Max/Msp for New Year's Eve?
>Hey Maxers. Happy Holidays.
>Any clue as to whether or not this is Max/Msp based? Uses a "virtual baton"
>and is "all done on the computer" according to the article.

./`./`Timothy A. Place


Date:Wed, 22 Dec 1999 10:05:59 -0500
From:Neal Farwell <nfarwell@FAS.HARVARD.EDU>
Subject: Re: Question for Max Guru's

Hi DR Studen

The delay object might serve. I don't know what it's maximum length is, but
illustrations in the Max manual show 300000ms ie 6 minutes.

If you give it a bang you'll get one out after that delay; but if another
bang comes along before then, the first is forgotten. Ie. nothing comes out
until there has been nothing in for >= the delay time.



Date:Wed, 22 Dec 1999 12:56:19 -0500
From:Eric Singer <eric@ERICSINGER.COM>
Subject: Re: Question for Max Guru's

At 7:58 PM -0500 12/21/99, D. Studen wrote:
> .... I am thinking that using an "if" object which contains an
>expression similar to this: IF no activity within the exhibit for
>Xseconds then bang loadbang to thispatcher. The problem at this point
>is twofold.
#1 How can I have max monitor activity within a continuoulsly running
#2 Is there any way to write an IF oject expression as depicted above?

Use a 'delay' object. Add a 'send activity' box to the output of any
activity-generating object you wish to monitor (and make it the leftmost
box to give it low priority). Then, create an object chain as follows to
your reset logic (here I'm using || xxx || for an object box and | xxx |
for a message box):

|| receive activity ||
|| b1 ||
| stop, bang |
|| delay 60000 ||
|| send reset ||

Any message (i.e. activity) received by the top box is converted into a
bang. This generates a stop message (clearing any previous bang sent to
the delay), followed by a bang message into the delay. If no activity is
received within 60 seconds (in this case - set your own desired value), the
delay outputs the bang, which is sent to your reset circuitry.



Date:Wed, 22 Dec 1999 09:57:12 -0800
From:"Brian K. Slater" <brian@TACTEX.COM>
Subject: Multi-Touch Controller News

Season's Greetings fromall of us at Tactex Controls! And welcome to Volume
2 of the Tactex Multi-Touch Controller News.
This quarterly e-mail newsletter is intended to provide you with the inside
scoop on current developments here at Tactex Controls. You are on this list
if you have received our first newsletter or expressed an interest in
receiving an update on our activities - if you wish to be removed from this
list please let us know and we will remove your name. In this issue we
report on the success of our Beta test program, and announce the production
release of the MTC Express.

In mid-July we began shipping Beta units to select users in the US and
Canada. Many thanks go out to our Beta testers (many of whom are Opcode/MAX
users) who put the MTC Express through its paces, and provided us with the
dynamic and insightful feedback we were looking for. We are pleased to see
such an enthusiastic and creative community of MTC users emerge from the
program and provide such encouraging feedback. Here's what they had to say:
Make it faster!
The Beta units provide data 40 times a second, which is about the speed of a
mouse. The unanimous call for more speed has inspired us to speed up the MTC
to 200 data samples per second!
Make it more sensitive!
This was the second most important need identified by our Beta team as being
critical to really intuitive performance applications. The original pad
required a Minimum Activation Pressure of .4psi. It is less than half that
now, if you can feel your finger touch the pad - the pad can feel you
touching it!
How about that control surface?
This was a bit of a surprise for us - initial feedback indicated that the
leather feel of the MTC Express was very attractive, but in retrospect this
may have been more of a novelty than a practical surface. On the suggestion
of one of our testers we mocked up a pad with a matte Lexan surface. When
others saw and felt this surface their comments indicated that, while the
leather surface was really unique and interesting, the Lexan provided a much
more positive feel and a more durable surface.
It's too expensive!
OK! The good news here is that the Beta program, in conjunction with our
marketing research, has clearly shown we are on the right track and can make
considerably more of the production MTC Express, thus reducing the price...
Now $495 US.
When is the MTC Express production release?
We ship the first production units on February 28, 2000.
The controller is in the same anodized aluminum case as the Beta units, but
with a Lexan control surface.
Speed is 200 Hz and sensitivity is finger-touch light.
In addition, we are including some MAX patches and other demonstration
examples developed by our Beta testers along with the API, our Tactex Max
object and a few other goodies.
Full details and specifications will be available on our web site.
In other news:
Growing along with our Beta program, our OEM product development proceeds
apace with 3 manufacturers currently planning for Q3 2000 delivery of
control pads in volume. You will see some cool new products in retail for
Christmas next year.
NAMM: We'll be there! Will you? We don't have a booth - but drop us a line,
we'd love to meet you for coffee and show you the new MTC Express!
Well that's about it for now, thanks for taking a moment to catch up on our
news. If there is anything of particular interest that you would like to see
in this report - or on our website, such as OEM profiles, MAX user profiles,
a place to post and share MAX patches, please let us know!
Team Tactex
#3-203 Harbour RoadPhone:250-480-1132
Victoria, British ColumbiaFax: 250-480-1142
NOTE: If you do not wish to receive the Tactex Multi-Touch Controller News,
featuring updates and new developments, details on pending product releases,
and other related information, simply reply to this e-mail and, in the body
of the message, type UNSUBSCRIBE along with the e-mail address to which this
message was sent and we will remove you from our list. Our apologies if this
message has caused any inconvenience.


Date:Wed, 22 Dec 1999 16:13:53 -0500
From:David Gamper <dgamper@EARTHLINK.NET>

Subject: Re: Max/Msp for New Year's Eve?

My first impression is it's someone with a Buchla Lightning
controller and a synth module - period. Wouldn't even need Max, let
alone MSP.

>Any clue as to whether or not this is Max/Msp based? Uses a "virtual baton"
>and is "all done on the computer" according to the article.


Date:Wed, 22 Dec 1999 16:15:45 -0500
From:David Gamper <dgamper@EARTHLINK.NET>
Subject: Re: MAX Digest - 20 Dec 1999 to 21 Dec 1999 (#1999-364)

Is this the expensive Kurzweil ExpressionMate @ $549?

>i have a vague memory (well, loads of them actually) that someone once
>marketed a midi pitchbend strip that could be fixed to whatever
controller keyboard you used.


End of MAX Digest - 21 Dec 1999 to 22 Dec 1999 (#1999-365)