Casio CZ/ VZ/ FZ - Pro Series

On 22 Aug, 2013, at 10:44 PM, bernard.escaillas wrote:

> The midi protocol is just a subset of the serial protocol from "the  
> gold old days".
> The speed has been defined once for all @ 31250 bauds.
> ( The Midi association has been working on new "fast Midi" for two  
> decades without beeing able to decide "what next"...So this won't  
> change anytime soon :( )

SCI used faster MIDI in some of their machines.
>
> You can always tweak this speed by a few percents in hardware  
> because receiving device tend to sync to signal. So you may be fine  
> using 32000 bauds for exemple if it is more convenient.

Yamaha, Korg and maybe other firms have used also MIDI transfer  
through common Serial cable (with Mini DIN connector), and for Mac  
there's a speed 38400 bps.

> Now concerning the loss of long messages between a modern computer  
> and a synth from the 80's, the problem lies in the CPU speed of the  
> synth and its input buffer:
> when the synth receive data on its midi port it has to read and  
> analyse data present in the input buffer. For sysex, the analysis  
> can take some time before the synth has consumed the data and freed  
> the input buffer. If during this analysis the PC keeps pushing data  
> (even at 31250 bauds) into the synth's input buffer, the later can  
> become full and the serial chip refuses (or ignores) incomming data...
> Then the sysex is incomplete.
>
> To give the synth enough time to digest the new sysex information,  
> the sending PC should do some pauses. This "pause" setting is  
> available in a tool like MidiOx. There you can set the size of  
> "packets" and the pause to do between packets.
> It may take some testing to find the correct values for a given  
> synth...

Exactly so, this is the same what I have written. Thanks for  
confirmation.
>
> Old synth don't have more than 64, 128 or 256 bytes at best of input  
> buffer ! Flash memory was very expensive at that time.
> I don't know the CZ5000 specs...
> Try to send packets of 128 bytes with 50ms in between.
> If messages are still lost, either decrease packet size of increase  
> pause duration.
> If communication is fine you can try to increase packet and reduce  
> pauses.


Daniel Forro



[Non-text portions of this message have been removed]