Rs-232 vs MIDI vs SCSI?

Thu Apr 24 18:11:27 CEST 1997

3)
	id sma005876; Thu Apr 24 16:22:32 1997
Received: from horus.sara.nl (horus.sara.nl [192.16.188.40]) by ns1.sara.nl (8.8.
3/8.7.1) with ESMTP id RAA06322; Thu, 24 Apr 1997 17:16:31 +0200 (MET DST)
Received: by horus.sara.nl (8.8.3/4.04)
          id PAA103301; Thu, 24 Apr 1997 15:14:31 GMT
Received: from cdale3.midwest.net by horus.sara.nl (8.8.3/4.04)
          id PAA35709; Thu, 24 Apr 1997 15:14:29 GMT
Received: from cyborg0.midwest.net (cdale-7.midwest.net [204.248.41.82]) by cdale
3.midwest.net (8.7.3/8.6.9) with ESMTP id KAA21299; Thu, 24 Apr 1997 10:14:00 -05
00 (CDT)
Message-Id: <199704241514.KAA21299 at cdale3.midwest.net>
Reply-To: <cyborg0 at midwest.net>
From: "Rob01" <cyborg0 at midwest.net>
To: "Paul Schreiber" <synth1 at airmail.net>,
        "'synth diy general post'" <synth-diy at horus.sara.nl>
Subject:  Rs-232 vs MIDI vs SCSI?
Date: Thu, 24 Apr 1997 10:20:34 -0500
X-Msmail-Priority: Normal
X-Priority: 3
X-Mailer: Microsoft Internet Mail 4.70.1157
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Sender: owner-synth-diy at horus.sara.nl
Precedence: bulk
-------------- next part --------------
     Alternatively, why not use something like RS-422 (unidirectional, 
     twisted pair, 10Mbps) as a drop-in replacement for MIDI. The protocol 
     would be exactly the same, only faster, so retrofitting to existing 
     synths would simply (ha!, "simply" he said, oh the innocence of 
     youth...) mean replacing the serial port system to handle the higher 
     data rate.

     This has been an idea mulling around my mind for some time now, 
     especially while I wait for samples to cross the MIDI void, and wait, 
     and wait, and wait....

     Certainly 10Mbps would be _much_ faster than 31,250baud (by a factor 
     of 320) which would be well worth the effort for sample download time 
     and message latency reduction - just imagine, a sample download that 
     takes 5 minutes at MIDI rates would take less than ONE second at 
     RS-422 rates!  Also, being serial the hardware is somewhat cheaper 
     than implementing full-blown SCSI.

     Going out on a limb now, using the extra bandwidth one could implement 
     a more comprehensive network protocol, something like ethernet, where 
     messages are passed around in larger packets and thus able to break 
     the 16 device address limit.  Ummmmm, maybe.

     Neil Johnson
     ============

|I know it would prolly be a little difficult, but it would be really neat 
|if we start a standard that uses SCSI to transfer the data instead of 
|MIDI.. I would like to do a few experiments along these lines. THEN you 
|wouldnt have the problems of things like the big midi log-jam..BUT you 
|would be confined to 8 units per scsi bus, but honestly, how many units 
|needing this bandwidth are you gonna have? UM, on second thought, dont ANY 
|of you answer, cuz I dread the consequences.. :)) Anyway, with SCSI you 
|could do controller info out the YINYANG and never have to worry a whole 
|lot about delay.. I would think its possible, but anyone have any opinions 
|on it.. ??? Maybe well just bog the CPU down.. Heck, i can do that now 
|with my Drumstation..:) I need a SGI for music..

                  Rob01
             Zeros & Ones
The Mind Controls the Machine 
The Machine Controls the Body