[For context, Giorgio is an MPEG delegate who's making proposals
to help improve the tempo handling in SAOL. I asked him to
have the discussion here because it's a good snapshot of where
we are in reference software implementation now.]
Giorgio ZOIA wrote:
> Anyway, last week I had time to go through saolc to better investigate
> how it exactly works. If I well understood the tool as it is now, it should be
> feasible for me to implement and test the modifications. I think that
> at the moment all the input (either file or bitstream) is processed at
> the very beginning, all the events are scheduled before starting the
> performance.
Yes, this is correct. The reason everything is handled up front is
because
only the bitstream header is handled currently; all the score, orch,
and samples are packed up there. I want to finish the rest of the
bitstream handling before Tokyo (March 17).
> The points to work on should be a) the event scheduling
> (even if not necessary at the time) and b) the tempo execution. It also
> seems to me that no hooks for the outside are implemented so far (3-d,
> etc.) and probably the bifs part concerning synchronization and interaction
> can be delayed.
Yes, there's no 'systems' part attached to 'saolc' right now.
Mike Coleman, myself, and an ad-hoc group are doing that now for
all the audio tools, but you shouldn't worry about it
for this. Your assessment of the right points to work on is correct:
as each event gets scheduled, and when a new 'tempo' event comes in.
I think the semantics of interactive tempo change should be exactly
the same as the semantics of a new 'tempo' event in the score, yes?
> Before beginning practically the work I'd like to know if what I have (post
> Fribourg) is more or less a stable version of the code. In such a way once
> the modifications are tested, I can add them to the last version in a few
> time.
It's as stable as anything is right now, but I'm working very hard on
the code base between now and Tokyo. However, if you make changes
to the reference software as of Fribourg, I'm happy to integrate them
into the latest working version myself.
> Moreover I'd like to understand better if there are already some official
> and approved modifications (and which ones) to the CD, or if any new version
> will be delayed after the Tokyo meeting (regarding SA but also Systems). In
> such
> latter case which kind of activity is required in the next weeks.
No official modifications were allowed at the last meeting, because
some nations have not yet conducted their ballot and comments on
the CD. We'll make changes in Tokyo. In San Jose, we produced a list
of unofficial changes which will be officially made in Tokyo. You can
check the San Jose output documents for documents like "Study on CD",
which has some of these. Most of the changes to Structured Audio
are still private, though.
> A small technical detail: besides the bug I reported you before San Jose, I
> have
> an annoying distortion problem with some waveforms, i.e. I can't obtain at the
> output the same sound level of the wavetable without distortion. To your
> experience, this comes from a precise reason or there is something not working
> to look at ? ( I even tried a very direct path read-loscil-write, NT and
> Solaris, and
> the problem is there).
>
I haven't seen this bug, I don't think. Can you post an orchestra/score
and the command line you're using so I can try to replicate it?
Thanks!
-- Eric
-- +-----------------+ | Eric Scheirer |A-7b5 D7b9|G-7 C7|Cb C-7b5 F7#9|Bb |B-7 E7| |eds@media.mit.edu| < http://sound.media.mit.edu/~eds > | 617 253 0112 |A A/G# F#-7 F#-/E|Eb-7b5 D7b5|Db|C7b5 B7b5|Bb| +-----------------+
This archive was generated by hypermail 2b29 : Wed May 10 2000 - 12:14:09 EDT