Here's why I won't be doing this anytime soon:Ch00rD wrote:I would absolutely *love* to see a "no compromises, make things sound as good as possible" hq interpolation mode in Vee3 (and although I don't have VRoom - yet - I guess the same goes for that). Since Vee3's CPU consumption per instance is pretty low as it is (almost negligible even on my 'old' C2D MBP), I would be more than happy to sacrifice at least 16x as much CPU slurping for improved audio slurping. Probably more.valhallasound wrote:[…] ValhallaRoom and ValhallaVintageVerb only interpolate some of the delays in the structure, to save CPU. If every delay was interpolated, the same slurping sounds could be achieved as in ÜberMod and Shimmer, but the CPU would be at least 2X as high as the current CPU consumption. Probably more.
Sean Costello
- The interpolation wouldn't improve the sound of VintageVerb in its static state, just in the transition between states. VintageVerb uses high quality or low quality interpolation, depending on the Color mode. In NOW mode, the interpolation used has a flat frequency response with low noise.
- Not everyone likes the tape echo slewing sound for transitions. Some people would prefer an instant transition without any artifacts whatsoever. This is hard to do without causing clicks, but I have some thoughts about eliminating these in the future.
- In order to get a slewed reverb, the architecture needs to allow this from the get go. This isn't about a particular topology, so much as the programming style used - but certain topologies lend themselves to particular programming styles. An algorithm like Shimmer works great with heavy slewing, as it is made up of discrete chunks of processing, each of which can be individually slewed. VintageVerb doesn't work this way, and I would have to recreate every algorithm from scratch in order to properly slew the sounds.
- CPU consumption is still very important for a lot of users. Putting a high CPU option in VintageVerb incurs the risk that people would turn that on by default. If I was going to add higher CPU options to VintageVerb, I'd rather do it in such a way as to improve the base sound of an algorithm, rather than trying to obtain a particular effect during Size transitions.
Sean Costello

