Emu XL-7 & MP-7 User's Group

On Saturday, September 8, 2001, at 08:45 PM, dubtribe@... 
wrote:

> --- In xl7@y..., Scott Ruda <scottr@e...> wrote:
>
>> There might be a problem if you insert the changes exactly at time
>>  zero, which might get overridden when the local MIDI map settings
>> are sent which is done when the sequence is first loaded. But I'm not
>> exactly sure how that works (I'll look into it).
>
> THIS explains a lot.
>
> let me ask you, with no scorn at all:
>
> where else would you be expected to put a program change???
>
> seems to me that this is a serious issue to overcome.
>
> nearly all program changes need to be located at 1.0.000 in order to
> operate prior to the note playing.

Some systems use a 'setup bar' so that any time consuming setup 
information can be sent and absorbed before the first beat. This 
technique is problematic within the definition of an SMF. It doesn't 
easily support the definition of a loop point, and there would need to 
be a way to define a setup bar plus a loop start point in a looping 
pattern sequence.

**-7 sends any setup information found at time 0 on pattern select in an 
attempt to get that stuff out of the way before you have pressed Play.

Have you tried Aaron's suggestion of putting 2 bank select/program 
changes at the beginning of the sequence - one to force the system off 
the desired preset in case it is already there on the local system and 
another to bring the internal and external sounds back to the right 
program?

My suggestion was that you could possibly put the program changes at 
tick 1, and then you might have to move all of your first note events to 
tick 2. That may or may not be practical and it might not loop very 
accurately. Aaron's suggestion should not have this problem, so it is a 
better workaround. Since the setup information is sent only once at 
pattern select time it will not cause any loop timing errors.

I'm sorry that this is causing troubles for you right now. We are 
looking into a clean way to resolve this.

Scott