Re: [fluid-dev] Proposal for a new feature: lazy-loading of SoundFonts

[Tom M.]

I keep thinking about this preloading feature and I'm not yet fully convinced 
of it. I see quite high obstacles for it being beneficial i.e.:

- We need a user who plays MIDI files via command line and listens to them in 
real-time.
- The user must actively enable on demand sample loading
- The MIDI file(s) must make use of on-the-fly program changes (because only 
then hearable sound interrupts will occur)

I can fully understand to enable dynamic sample loading for non-realtime 
szenarios (like file rendering), or for Marcus special resource limited 
environment. But the only reason for a normal user to enable dynamic sample 
loading I see is to save memory when playing back a MIDI file with a very big 
(>1GiB) soundfont. But the low latency of the playback requires to have the 
samples in memory. After all this preloading feature would only mask the high 
latency of dynamic sample loading. And in worst case it would minimize the 
memory gain of the dynamic sample loading to zero, i.e. when it preloads 
(nearly) all samples or only the biggest ones.

So, do we need sample preloading?

Tom