Good question. The majority of the algorithm used by the VCV Rack tape module is identical to the algorithm used by the plugin, but there are three differences.Michey wrote: Sat Aug 01, 2020 12:37 pm Is there a difference sound-wise to the regular ChowDSP TAPE ?
1. The VCV module currently does not support oversampling, I'm still trying to work out how possible this is within the VCV format. Running your session at a higher sampling rate will help reduce aliasing artifacts, but the algorithm also relies on internal oversampling to help with stability. With that in mind I had to make a couple tweaks to help improve stability in other ways...
2. The latest Chow tape plugin allows users to pick which type of ODE solver to use for running the tape hysteresis simulation. In the VCV module, I've fixed this to use a Newton-Raphson solver, since it is most robust to potential instabilities.
3. Finally, in both versions, I use an Alpha Transform for digitizing the hysteresis equation. This transform has a parameter alpha, that is typically chosen between 0 and 1. In general, lower values introduce high frequency damping, but also make the system more robust to potential instabilities. In the plugin algorithm, I have alpha fixed to 0.85, while for VCV I've set this to 0.25. While I do notice a slight audible difference at lower sample rates, I think this trade-off is worth it for the improved stability.
Hope this answers your question!
Thanks,
Jatin
