The latency of VST3-plugins in Wavelab or Audacity is far better than in Spectral Layers.
It is indeed ! But WaveLab and Audacity don’t apply VST3 plugins to spectral selections - hence why it has more latency in SpectraLayers.
3 Likes
@Sunnyman a new username and profile? oh my
Yes. It does not change the facts, right?
Sadly, it got my attention again when using FabFilter Pro-Q 4.
The kind of “stutter-step”-latency renders the plugins as unusable (for me) in SL.
1 Like
I don’t mean to derail this topic but (as far as my understanding goes), A.I. has improved drastically especially over the last 10 years. I say all of that to make the point that optimizing something like this 10 years ago would have been challenging, however A.I. has improved so much that now AMD started implementing A.I. hardware acceleration within their chips.
So yeah, I sat all of that to make the point that I am optimal and optimistic that optimization would improve drastically for this issue soon.
1 Like
I do find this still not acceptable for todays standards and computing power.
Even if the data are all in the spectrum, todays computers are fast enough to transform these data back and move them to another channel. (FFT should suffice)
I still believe, that the actual datamodels of Spectral Layers might need some algorithmic improvements.
Numerical efficiency and fast datahandling (parallel computing?) are always a challenge.
→ https://www.youtube.com/watch?v=c33AZBnRHks
(The concept of Spectral Layers is ingenious, no question about that.)
1 Like