audio -- some MIDI / Quicktime resources in Revolution

[Klaus Major]

Hi Mark,

i will try to answer some of you questions, too, if you don't mind...

> Xavier,
>
> Here's a set of wonderings I posted a couple of days ago directed to 
> Dan Shafer.   My current specific project involves a narrated eBook 
> where text hiliting (line level) would essentially follow the recorded 
> voice.  There would be multiple fields on any given card that would 
> require real audio to be read in support of them.
>
> Thanks,
> Mark
>
>> Questions that I have would involve:
>>
>> Audio formats:
>> What's the best for working with Rev under different circumstances?  
>> MP3, MP4, AIFF (huge files), M4B, etc.   What are the trade offs in 
>> terms of fidelity, file size, ease of manipulation, compatibility.

I guess that QuickTime may NOT be possible?
Think so...

>> Benefits/drawbacks of the different formats:
>> X-platform issues, standards that can succeed across them?  Example: 
>> Books on tape use the m4b format which apparently includes striping 
>> of some sort which allows a user to resume listening at a place he 
>> left off.  It may not to be compatible with Rev; at least when I 
>> tried to import such a file into a stack it was howlie garbage.

For spoken words the AU fileformat may be a good choice.
Quality is acceptable, so is the filesize AND it is crossplatform 
WITHOUT QuickTime...

>> Audio/text synchronization:
>> How can I synchronize a longer audio file with real-time text 
>> hiliting features?  Example: I want to have the narration/reading 
>> (real recorded voice, not text-to-speech) of a book playing while 
>> corresponding text is hilited on screen.  Not word by word, 
>> necessarily, but at least paragraph by paragraph. How might the audio 
>> drive the hilite feature in a text field (and vice versa)?  Can audio 
>> files somehow be tagged so as to trigger corresponding text events, 
>> call handlers that would import new text, scroll fields, etc.  Or 
>> would this have to occur in reverse.  It would be nice if audio could 
>> have markers that would be linked to text lines and trigger handlers. 
>>  But that may be fanstasy.

I would suggest to set and use "callbacks"...
"Callbacks" can trigger actions/handler and will do what you want.

Short explanation of "callbacks":
They are special properties of player objects where you define a time 
and the player will
trigger a handler at that time...

I have a littel stack at Rev-online "Fun with callbacks" that will get 
you started...

>> Playback of multiple streams:  Can more than one audio event play 
>> back at once?
>> Music synched with voiceover located in different files?

Don't know about Unix/Linux, but with players you can have differnt 
sounds playing at the same time
on windows without QuickTime...

No problem on a Mac anyway ;-)

>> Audio File Storage/retrieval/loading:
>> How would one anticipate and pre-load audio audio files so that there 
>> is seamless playback from file to file?  Where are the files kept, 
>> and how can they be queued so that the user hears no gaps or clicks, 
>> etc.?

Hmmm, any logical folder management will do.

There is "prepare" command, that will "pre-load" audioclips , but that 
will only work with the
"play ac xyz" command...

With QuickTime you could create a SMIL-file (XML-like text file that 
tells QT to play files in a row...)
and play that single file (that may contains the reference to many 
other files. I am about to release
a little SMIL-lib in the next time :-)

>> Memory requirements:
>> How much memory needs to be allocated to achieve a seamless 
>> integration of on screen visuals and supplemental audio?  How would 
>> this happen?

Sorry, no idea...


Hope that helps a bit :-)


Regards

Klaus Major
klaus at major-k.de
http://www.major-k.de