Subject: MAX Digest - 2 Jun 1998 to 3 Jun 1998
Date: Thu, 4 Jun 1998 00:00:16 -0400
From: Automatic digest processor 
Reply-To: MAX - interactive music/multimedia standard environments
     
To: Recipients of MAX digests 

There are 4 messages totalling 158 lines in this issue.

Topics of the day:

  1. Numerical properties of random and clocker objects.
  2. Max on Powerbook G3
  3. MAX Digest - 1 Jun 1998 to 2 Jun 1998
  4. Max networx / OpenSoundControl

----------------------------------------------------------------------

Date:    Wed, 3 Jun 1998 09:05:30 +0100
From:    lossius 
Subject: Numerical properties of random and clocker objects.

I'm writing music installations that's meant to be able to run forever
without ever repeating themselves, and this have raised a few questions
concerning numerical behavior of some Max objects:

How "random" is the random object? This might sound as a weird question,
but I'll explain: Some years ago I took a university course on numerical
methods, and one of the items covered was random processes. I don't have
the textbook at the moment, but as far as I remember, random processors
use some quite complex function to calculat the output, most often using
the previous output as input. This means that at least some of them are
cyclic, repeating themselves, althought with an extrem long period.

Next question: What happens to the output from the clocker object when
the upper limit of integer representation is reached? I've figured that
after about 24 to 25 days counter object hits the highest integer value
that the object is able to represent numerical (assuming it to be the
same as for integer Number Box: 2147483647). What happens next: return to
zero, maximun negative value, or a potential crash?

Thanks!

Trond Lossius

------------------------------

Date:    Wed, 3 Jun 1998 10:30:12 +0100
From:    Lawrence Casserley 
Subject: Re: Max on Powerbook G3

In message , Fred Collopy  writes
>Amy Knoles wrote:
>>
>>Anyone try running MAX on the new G3 powerbooks?  I'm thinking of the
>>233 MHz vers....Amy,
>
>I'm using MAX on one of the early G3 Powerbooks (before they got all the
>numbers) and love it. It's pretty nearly perfect.
>.....

I am an old ISPW hand now also running msp on a 266MHz G3. I am doing
live signal processing in performance. I would love to use a Powerbook
to make my travelling lighter (especially those gigs where I have to
fly). My problem is that I need more I/O channels - 4in/4out is a basic
requirement, but 8 would be ideal. Is anyone out there
designing/building a multi-channel audio interface for Powerbooks? Then
it _would_ be perfect!

Lawrence Casserley

--
Lawrence Electronic Operations -Tel +44 1494 481381 -FAX +44 1494 481454
Signal Processing for Contemporary Music -email leo@chiltern.demon.co.uk
http://www.chiltern.demon.co.uk

------------------------------

Date:    Wed, 3 Jun 1998 10:19:06 MST
From:    Mike Metlay 
Subject: Re: MAX Digest - 1 Jun 1998 to 2 Jun 1998

From:    Rex Brian Coggins 
>I'll bet anyone a quarter that Gibson will never get Opcode to roll a
Galaxy
>Plus Editor for the OB-Mx, Xpander or Matrix 12.

Why should this year be any different from any other year that the Xpander,
M12, and OBMx were in existence? Just be grateful there's librarian
support. I haven't missed an editor, since I've always felt that the Xp
lent itself to actual front-panel programming, but then I've had mine for
nearly thirteen years so maybe I'm just used to the interface.

>Oh, I forgot, that's one the reasons I picked up MAX!

I always felt that writing an ed/lib with MAX was kind of like swatting
flies with a brick, but what do I know? If the tools work, they work. I
can't wait to see what Amanda comes up with for the Xpander she just bought
(I have infected her, mua ha ha ha ha)

mike

ps. Thanks to all list members for advice and chatter--I'm taking a break
from the MAX list for a while, as I just got started on a long project
that's going to keep me from doing any MAX work at all for something like a
year. Kudos to David Z for a great product, and good luck to you all.
Seeyaz!

--
  "Please don't reply to this post until you've wiped the vomit from your
   keyboard!"                                                  (m. simon)
===========================================================================
Mike Metlay * ATOMIC CITY * P.O. Box 17083 * Boulder, CO * 80308-0083 * USA
  + atomic@tesser.com  *  1-800-924-ATOM  *  http://www.atomiccity.com  +
CD orders via LOFTY PURSUITS: 800.548.6724 & 904.385.6463, FAX 904.668.5825

------------------------------

Date:    Wed, 3 Jun 1998 10:45:57 -0700
From:    Matt Wright 
Subject: Re: Max networx / OpenSoundControl

>Michael Sweet said:

>I finally got the OTUDP object to work on my PPC, and it seems to work fine
>although I'm a little miffed by OpenSoundControl.  Whenever I try to send a
>capitol letter across the network it comes out on the other end as a large
>float.  Can anyone clue me in as to how OpenSoundControl works?

I wrote both of those objects.  I'm curious about the "finally" about
OTUDP---please let me know what you encountered along the way.

OpenSoundControl is documented in copious detail on the WWW site (and the
help
patch for the OpenSoundControl Max object gives the URL):

        http://www.cnmat.berkeley.edu/OpenSoundControl

Here's the explanation of the problem you're having.  In Max, data objects
(i.e., atoms) have a manifest type tag, meaning that they are represented by
a
struct that consists of the actual data and also another field that says
"these bytes are a float" or "these bytes are a pointer to a symbol" or
"these
bytes are an int".

OpenSoundControl doesn't work that way.  For efficiency reasons, the OSC
protocol does not include these manifest type tags; the assumption is that
the
application receiving the OSC messages knows that, for example, its "gain"
message expects a float argument, and its "load-a-new-soundfile" message
expects a string.

So what happens when you have a Mac receiving OSC messages and trying to
convert them back into Max data objects?  It doesn't know what the original
data types were; so it has to look at 4 bytes like 0x41000000 and decide
whether they're the float 8.0, the string "A" (with some extra null bytes),
or
the integer 1090519040.  So it guesses.  The rules it uses to make this
guess
are detailed in

        http://cnmat.CNMAT.Berkeley.EDU/OpenSoundControl/dumpOSC.html

in the section "Message Argument Type-guessing Heuristics".

Yes, it's a kludge.  The really right thing would be to be able to tell the
OpenSoundControl object which types to expect with which particular OSC
messages.  But trying to build this kind of association table in Max seemed
like a lot of work for not much payoff.

Remember that the address of each particular OSC message is definitely a
string, so if all you want to send is strings, you could take advantage of
that.

This is getting pretty esoteric; perhaps we should continue this thread in
private email.

-Matt

------------------------------

End of MAX Digest - 2 Jun 1998 to 3 Jun 1998
********************************************