challism wrote: ↑14 May 2022
I might have something going on with my computer. I'll try to do some maintenance and see if I can get some better performance. I haven't seen much complaining about the CPU load, which is why I'm bringing it up to see if anybody else has had the same issues. Could be a local problem. I'll also try it out on my other (much faster, more powerful) desktop computer.
Hi challism, that it's showing up on your DSP meter indeed suggests a laptop and it's fair to say that's generally not a strong point for Reason anyway, which in my experience struggles on any non-gaming level Windows laptop in the RE era (long gone are my nights spent jamming with Reason 2.5 on an Athlon-era AMD Compaq
). And then the HS2 is likely a little tougher on the underwhelming amount of laptop DSP cycles than most RE synths, but one needs to be careful not to overstate it: in practice I'd argue one instance of HS2 in a moderately-sized rack probably isn't going to be any more of an issue than any other single DSP-heavy synth, where polyphony/voices assignments are being used appropriately for the device in question.
On a modern, post-2016 desktop (e..g. my old 1800X or currently got a 3800X), it wouldn't be running more than about 3-4% idle and that wouldn't be enough to flicker the first DSP bar. And if that number
sounds high, remember it's a 32-osc synthesizer, with osc sync, and per-voice filters. Then you have to consider polyphony when it's active, so you can end up having a lot of voices being generated. You can't compare it to, say, even a simple non-osc-syncable 3-osc poly synth like Europa, which itself runs up to ~1.5% idle on its own. Honestly, for the number of oscillators and voices, the HS2 is really bloody efficient. (
scuzzyeye)
Devs don't know what device the end-user is using the instrument on, so there are cases where it's reasonable to just allow it to scale up as far as the user can go push it. That equally means that users do need scale their instruments back to what their hardware is actively capable of.
And honestly, there's nothing shameful about having to bounce tracks if one has to. That said, I did plenty of demos using only multiple HS2's and I've never bounced an HS track, and I remember I did have to bounce when putting together a track for The Legend competition. That was much harder on DSP. And the less said about ReSpire, the better. [I should note I do recall keeping the HS's Over-sampling amount down (and possibly even turned off in one or two cases), until needing to do the export render - that can be pretty crucial when DSP is tight to avoid bouncing. 2x oversampling is OKk for most preview/track development situations I ever had. But obviously, everyone's DSP mileage is going to vary somewhat. That's the nature of all these beasts.]