Hive 2.1: new sonic territory – Page 11 – u-he Forum

recursive one wrote:

Sun Nov 07, 2021 4:14 pm

Urs wrote:

Fri Nov 05, 2021 2:57 pm

Hehehe, but then we would have to increase the modulation control rate for the ModMatrix and lose all the CPU advantages that we currently have.

Why not going the synapse audio route with different options for modulation speed? So that the audio rate modulation could be just switched off when it’s not needed.

I do second the Pdxindy suggestion!

It would largely mean a rewrite. The code is written to do certain things in blocks of samples, e.g. parameter interpolation. We can probably compile the whole thing to use a smaller block size. But we can’t easily change the block size on the fly. So, by our design, which has pros and cons, one of the cons is that increasing control rate and thus lowering block size doesn’t lend itself to dynamic switching, such that faster modulations using the ModMatrix inherently increase CPU load – even for patches where you wouldn’t need it. There are pros, too. One of them is a low CPU hit despite the modular nature of our plug-in foundation.

I have spent a lot of time experimenting with various control rate settings in Hive and I’m convinced that for the current set of features, a higher control rate – and the CPU impact this has – is not justified. I’m also precious about the very low CPU foot print of Hive in real world situations, and it would mean a lot of pain to me to give that up.

Hive simply isn’t meant to be an FM/CrossMod synth. Not every synth can excel at everything.

Read more here: Source link