Abe Pazos wrote:
In my case there is a strong interactive part, and there is not
enough time to send user interaction data to the server and
then render the sound. And we have thousands of simultaneous
users, so it's too much load for any cpu.
It should not take long to render and thousands of users simultaneously
wanting to compose a song sounds like a rather interesting application
with a huge market and lots of ways to get funds.
I doubt if it will get bigger than Google or MSN so I would not worry
about the server side stuff very much. If you actually do get to 1000
simultaneous users (implies several tens of thousands of clients) you
will likely store the MP3 notes in memory and build the composition
right into the HTTP server's response stream on a number of load
balanced parallel servers.
Anyways I know there is client side sound generation, like the
found in http://code.google.com/p/popforge/ or
http://8bitboy.popforge.de/ or all examples from Andre Michelle.
Looks interesting and from the short description, pretty much what you
are looking for.
Maybe I try that in the future. Then the sound will probably
have no glitches. I wonder about how well would be graphics
synchronized to the sound though...
If you are playing a sound file, the trick is to synchronize the
graphics to the sound. Watch the sound file progress and orchestrate the
graphics to match the sound's progress.
We do this all the time in our eLearning delivery application that
synchronizes on-screen actions with the narration. If something
graphical is supposed to happen at 10.25 seconds into the sound file,
then trigger it at that time.
If you are doing timeline graphics then you need to watch the timeline
and keep it in synch.
It is less obvious to the user if you miss the timing on a visual event
than if you screw up the playing of music. That is why I originally
suggested the construction of a single sound file rather than trying to
deliver a note on schedule.
Ron
Ron Wheeler wrote:
I would do it on the server side.
Since you are using a stable set of notes, you could have a duplicate
set on the server and have the Flash Player send the list that you
want assembled and the server could return an MP3.
You could also have the server save the sequence for later
examination or replay.
Looking more like MIDI by the day. Put a MIDI player on the server.
There are probably ways to do it on the client as well.
Ron
Kerry Thompson wrote:
Ron Wheeler wrote:
If you have a playback button that plays the composed work, you can
concatenate the sounds/notes "selected" and play the composed segment.
That's interesting--can that be done at runtime? How can you
concatenate the sounds?
Cordially,
Kerry Thompson
_______________________________________________
Flashcoders mailing list
Flashcoders@chattyfig.figleaf.com
http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
_______________________________________________
Flashcoders mailing list
Flashcoders@chattyfig.figleaf.com
http://chattyfig.figleaf.com/mailman/listinfo/flashcoders
