Emax

Hi

Of course I wasn't sending these 90000 or 2000 commands to the 
Emulator II manually :-)
My program is indeed running in a loop and launching the commands 
with configurable delay times. Even with optimal delay settings the 
total communication time is between 7 and 20 minutes, depending on 
the data packet size.

Anyway, I did some experiments with the *Emax* now and these are my 
conclusions so far:
1/ I am able to communicate between the PC and the Emax via RS422, no 
problem.
2/ Re-handshaking is sometimes required but the communication seems 
to be more reliable compared with the Emulator II.
3/ I can't get higher speeds than Alchemy can. For a fully loaded 
Emax bank of 512K sample data, the total communication time will be 
between 2.5 minutes and 3 minutes (the same as on Alchemy). This is 
not caused by the actual baudrate or by the 512K size of the bank. It 
is simply caused by the fact that the Emax SYSEX protocol requires 
more than 6500 commands/handshakes to load or unload a full 512K bank.

Mainly due to the third conclusion I won't further investigate the 
possibility to load/unload banks by means of the 'normal' SYSEX and 
MMA commands. The total response time is simply not acceptable for 
the average musician.
Note that the response times are acceptable however for editing/VST 
purposes, but these are not on my current agenda.

The only thing I'm still considering is trying to find out if the 
Emax contains a special command which allows unloads/loads without 
handshaking all the time... so in fact I will try to find out the 
commands which have probably been used for communicating with the OMI-
3/CDS3 cdrom drives.

///E-Synthesist


--- In emax@yahoogroups.com, Michael Wisbech <michael@...> wrote:
>
> Hi
> I only have programming knowledge from flash so i may just be 
rambling around.
> and dont know if the script language you are using can do 'for' 
loops
> but in flash I made a solution to a problem like yours
> by making  'for loop' which was executed at some intervals, each 
time 
> 'sending'/ calculating some datas,
> and having a counter addding up everytime, to use only inside 
the 'for loop'.
> 
> 2000 commands? but each one would be the same commands-  in the 
script, right?
> so write one command,  put it in a for loop, and make it execute 
2000 times 
> with a interval of 'x' millisecons or what ever.
> 
> 
> But if you are trying to make  a rs422 interphase to dump banks, i 
would 
> think that SCSI is faster,
> I belive that rs422 was made for editing. but don't know.
> 
> 
> At 01-11-2008 12:50, you wrote:
> 
> >I thought about using small packages of bytes. This is indeed more
> >reliable but... for loading/unloading about 486000 bytes to/from 
the
> >EII, the EII would require more than 90000 commands. The Max is
> >sending only 2000 commands... Just initiating these 90000 commands
> >takes an awful lot of time (request/reply with ack/nack). I tried 
it
> >once and the result of this experiment is that the whole process
> >takes between 10 and 20 minutes :-)
> >I ordered a PCMCIA RS422 port now, which bypasses the USB port. 
Let's
> >see if this device will give better results...
> >
> >Right now I'm checking out the Emax communication.
> >The first thing I'll do is monitor the communication between 
Alchemy
> >3 and Emax, just to be sure I understand the RS422 sysex/mma specs
> >provided by Emu correctly. But I'm already fearing for the end-to-
end
> >response time when loading/unloading full banks with these 
commands.
> >Because when I select "Get All" in Alchemy, it takes minutes before
> >Alchemy has received all voices from the Emax via RS422; Alchemy is
> >probably doing the exact same thing as I was planning to do, i.e.
> >using this heap of sysex commands. So to be honest, that wouldn't 
be
> >an alternative for SCSI access, would it ?
> >
> 
> 
> -- 
> No virus found in this outgoing message.
> Checked by AVG. 
> Version: 7.5.549 / Virus Database: 270.8.4/1754 - Release Date: 29-
10-2008
>