Standard LV2 Plug-In Categories
- Spectral Processor.Pitch Shifter
- Dynamics Processor.Compressor
- Dynamics Processor.Expander
- Dynamics Processor.Limiter
- Dynamics Processor.Gate
Possible New Categories (Tentative )
- Timeline Changer
- Maybe as a subcategory of Utility? Or maybe better at top level? Larsl
- Noise Removal
- This should probably go under Utility. Wouldn't be unreasonable to get this into a new revision of the LV2 schema since some sorts of noise reduction can be done online (even if some of Audacity's built-in noise removers aren't). Larsl
Discussion of New Categories
- Do fade-in and Fade out fit in under 'Amplifier?'
- Something like timeline-changer is needed to cater for certain offline effects (change-speed, reverse...).
- Can we call it "Temporal" or something that isn't hyphenated? It just seems a very cumbersome name for it. --Richard
- Noise removal doesn't just consist of Fourier domain filters. In Audacity it may involve a 'compound' effect consisting of an analysis effect and an active effect combined together.
- I don't think this matters. We can put our built-in effects in any category we want without affecting the rest of the LV2 world. --Larsl
- James My first thought was that we should use 'Filter.Noise~Removal', however 'Filter' as a category might be too narrow. It all depends how LV2 committee define 'Filter'.
- Larsl Talked a bit to the LV2 people, looks like 'Filter' is for filters in the DSP sense - lowpass, highpass, notch etc.
- Richard Can anyone explain what "utility" is supposed to mean? To me it's no more helpful than having a "misc" category, and I'm loath to hide anything useful in there.
Analogies / Future Scaling
At the moment the category system is being viewed as a way to create hierarchical menus. However, it should be clear that it is a general 'tagging' mechanism, analogous to how e-mail messages and photos can be 'tagged' with arbitrary labels.
Photo sorters and finders may deal with 100,000's of photos, so the tagging and searching mechanisms are well developed there. Although it's too early for it, eventually equally sophisticated searching mechanisms may be needed for sound effects too. As the boundaries between effects, instruments and compositions starts to blur, we could reasonably anticipate a future with 100,000's of effects. Consider composite complex data-driven effects, particularly effects that change the character of a voice. Like musical compositions it's likely that in future it will be possible to create such effects without programming - and they'll be available in large numbers. We can see how the LV2 categories would scale well - the same kind of hierarchical tagging is used with photos.