On 1/9/03 6:24 PM, "Aaron Eppolito" <synthesis77@...> wrote:
> --- drK <drk@...> wrote:
>> Aaron I was surprised by the statement about pattern chaining in an
>> SMF, not because I know different (I don't) but because I was under
>> the assumption that the whole reason Dave Oppenheim pushed for the
>> Type-2 SMF format was to allow subsequences, which in my mind always
>> included the simpler case of a bunch of chained sequences. Are we
>> both right because type-2 was not an option to use because it was
>> not ever widely implemented (was it ever implemented outside of
>> Opcode's Studio Vision?)?
>
> Type 0 files are single-sequence, single-track files.
> Type 1 files are single-sequence, MULTI -track files.
> Type 2 files are MULTI -sequence, SINGLE-track files.
>
> If there were a type 3 file (I'm envisioning mutli-seq, multi-track),
> that would be almost perfect, but they don't exist. Type 2 files don't
> cut it primarily because they don't support multitrack (a requirement
> for XL-7).
>
> Also, there's the issue of data ownership. Sometimes it's nice to have
> all of the data bundled together (like multiple sequences in one file),
> however, that leads to much duplication and editing issues. On the
> other hand, files using "links" to other separate files are pleasantly
> small, but can break when the links fail.
>
> The branching/looping working group (headed by Chris Grigg if I
> remember correctly) was proposing a few additions to the SMF tags that
> would allow some form of file link, some conditionals (like "if",
> "else" "for", "while", etc), and anchor/jump points. I kinda stopped
> paying attention about a year ago when I ran out of time and it was
> clear that the problem wouldn't be solved in time for XL-7.
>
>> Sorry for the spurious interrupt. The subject is probably not of
>> much interest to anyone else. Just my curiosity.
>
> Nice use of "spurious interrupt". =) Trust me, file format issues
> like this were the source of MUCH agony when designing the sequencer!
>
> -Aaron