Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[6233638]

Well this is something that I have noticed for quite some time now; for some unknown reason, DSP seemed to respond quickly when playing to some zones/devices, and slowly in others.
 
When testing another VST plug-in today, I noticed that it seemed to be updating about four times every second or so... and it just so happened that I had recently changed that device to use a 250ms buffer size.
Even MC's own analyzer is affected by this, and rather than just being slow to update, the displays actually break entirely in some plug-ins once you go over 100ms.

Now I'm not sure why the update rate is tied into the buffer size - perhaps there's a good reason for it.
I understand that, of course, any DSP changes would take that long before they have any effect (e.g. half a second for a 500ms buffer) but I'm not sure why the displays are only updating at that rate as well.
 
Would it be possible to decouple them so that we get smooth updating (ideally 60 FPS+) without having to reduce the buffer size to 10ms or lower? (10ms = 100 FPS max, 25ms = 40 FPS max)

[Matt]

Now I'm not sure why the update rate is tied into the buffer size - perhaps there's a good reason for it.
I understand that, of course, any DSP changes would take that long before they have any effect (e.g. half a second for a 500ms buffer) but I'm not sure why the displays are only updating at that rate as well.

DSP happens right before giving the data to the sound card.  This is the buffer size.  So it makes sense that the delay varies with the buffer size selected.

[6233638]

I suspect that it's probably one of those things that is probably a big architectural change, but is there not any way to decouple the two?
I suppose there is no way to do it without adding more latency, but something like processing the DSP first and then feeding it into a separate buffer?
 
Anything above 10ms starts to show the effects of this reduced framerate, but 10ms is not the most stable of buffer sizes. (100-250ms is a better choice when latency is not an issue)