OK, sounds strange. The difference is that convolution gain compensation is applied at source rate when the convolution is performed, whilebibo01 ha scritto: 15/07/2020, 14:03Ho già ottenuto risposta da Jussi!!
...
Gli ho fatto presente, però, che con -20dB di attenuazione sul convolver si aveva ancora clipping.
Mentre bastavano -9dB sul volume di sistema per non avere clipping.
system volume is later after rate conversions and such. But I don't have direct answer out of my head why in this case it would be different.
You could also compare doing the convolution through the convolution dialog vs doing it through the pipeline matrix (and applying the gain
compensation as mix gain). Note that you can use either one, but not both, so either convolution engine or matrix (where convolution is
optional part of the process).
Convolution gain compensation is handled such way that you should get similar volume when switching the convolution on/off on the fly. Matrix
convolution cannot be switched on the fly, because it is commonly used for cross-over filters and it would be bad to switch your cross-over off
on the fly!
____________________________________________
Carlo,
se per favore riesci a fare quest'ulteriore verifica (come da paragrafo 2), così poi riesco a convincere Jussi a fare il test lui stesso.
Trovi una spiegazione di Matrix Pipeline a pg.27 del manuale.