Audio Triggering + Mouse Event latency

Bob Sneidar bobs at twft.com
Tue Nov 16 11:58:46 EST 2010


On that note, I remember reporting that in certain situations a fast mouse click was not seen by (then) RunRev. They are probably related. BTW, do you have any kind of virtual machine software running? It occurs to me that if something like Parallels is running, they install special drivers in the Mac to interface the keyboard with the guest os'es. Nuther shot in the dark. 

You may be dealing with interrupt priorities. I would poke around in Google and see if there is any way to up the ante on mouse event priorities. Of course, that would mean making modifications to the OS which is not ideal for a commercial application. 

Bob


On Nov 15, 2010, at 8:59 PM, Anthony Howe wrote:

> Hiya Folks,
> 
> Further to our recent emails on this...
> 
> I've now just implemented the same functionality using keyboard on keyDown events for each sound instead of the mouseDown event with startling (positive) results. (thanks Jacqueline for suggestion!)
> 
> It seems that the keyDown event arrives into LC much faster than a mouseDown event does for the purpose of triggering audio on a real time basis. Slightly less responsive on WinXP, but fantastic on OSX.
> 
> I've even been able to nest the keyDown handler at the card level and run a conditional before triggering the sound in order to get the right one, without affecting playback too much.
> 
> Whilst this is still not an ideal UI position to be in, and will limit the features we can provide onscreen, at least it's clearer that the issue probably revolves around considerable internal latency with capture of the mouseDown event, in comparison to its 'keyDown' counterpart.
> 
> If there's any way of tweaking the mouseDown event speed into LC, (assuming that IS actually the problem) it'd be great to provide this dual benefit to our users.
> 
> Looks like we may be able to avoid the flash/swf implementation by using keyboard events directly into LC though.... so a solution seems close.
> 
> Any thoughts based on that discovery folks?
> 
> Cheers,
> 
> A._______________________________________________
> use-revolution mailing list
> use-revolution at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-revolution




More information about the use-livecode mailing list