Ronald Meij Maker · 1 d ago
Ten. The distinction sounds academic until it stops being academic. A linear time invariant system is completely described by its impulse response. Capture the impulse, convolve, and you have it, exactly. That is why reverb sampling works. Run the same trick on a ladder filter being driven hard and you get a polite, dead thing that sounds nothing like the original, because everything interesting about that filter is what it does differently at different levels. The impulse response cannot see it. So: the first two videos establish what linear means and what an impulse response really promises. Then the ladder filter as a circuit. Then four conference talks on modelling non-linearity properly, and finally Volterra and Wiener-Hammerstein, which are two ways of trying to describe a non-linear system with something more than one impulse. The last one is from a dynamical systems course, not an audio one. It was the only serious treatment of Wiener-Hammerstein found, and it is included on that understanding. 01 What is a Linear Time Invariant (LTI) System? Iain Explains Signals, Systems, and Digital Comms Linearity and time invariance defined separately, and what each one buys you. The foundation the rest of this list stands on. www.youtube.com/watch?v=5JCuqlExTvo 02 What is an Impulse Response? Iain Explains Signals, Systems, and Digital Comms Why one impulse tells you everything about a linear system. Which is exactly the claim that collapses the moment the system is not linear. www.youtube.com/watch?v=WTmelRV_Yyo 03 ECE4450 L22: Moog Ladder Filters Analyzed (Analog Circuits for Music Synthesis, Georgia Tech course) Lantertronics Circuit analysis of the transistor ladder itself. The original whose non-linearity every model is chasing. www.youtube.com/watch?v=IRxeIEeAAnU 04 A Comparison of Virtual Analog Modelling Techniques - Jatin Chowdhury - ADC20 ADC - Audio Developer Conference The families of virtual analogue modelling set against each other on the same circuits. www.youtube.com/watch?v=FeS6wQJ6mYI 05 A Comparison of Virtual Analog Modelling Techniques (Part 2) - Christopher Clarke & Jatin Chowdhury ADC - Audio Developer Conference The follow up, with Clarke, extending the comparison. www.youtube.com/watch?v=ITjMLxzhVpQ 06 Recent Trends in Virtual Analog Modeling Based on Nonlinear Wave Digital Filters - R Giampiccolo ADC ADC - Audio Developer Conference Wave digital filter structures carrying non-linear elements, which is the standard route to a circuit accurate model. www.youtube.com/watch?v=hF1yPIjr4c4 07 Ivan Cohen - Fifty shades of distortion (ADC'17) ADC - Audio Developer Conference A survey of non-linear processing in plug-ins: waveshaping, circuit derived distortion, and the aliasing that comes free with both. www.youtube.com/watch?v=oIChUOV_0w4 08 Matthieu Brucher - Modelling and optimizing a distortion circuit (ADC'17) ADC - Audio Developer Conference Take a real circuit, derive its non-linear equations, then make the solver fast enough to run inside a plug-in. www.youtube.com/watch?v=HHiMD_QGRo0 09 Nebula Audio and Volterra kernel based sampling - Giancarlo del Sordo Borg Soundlab Acustica's co-founder on Volterra kernels: the commercial attempt to capture gear that one impulse response cannot describe. www.youtube.com/watch?v=OgcqqwR9Nh8 10 Hammerstein-Wiener models in NODEs (DS4DS 4.09) Data Science for Dynamical Systems (DS4DS) Block oriented models: a static non-linearity between linear dynamics. Not an audio course, and the only decent treatment found. www.youtube.com/watch?v=Kz_4tqW3FCM Every link in this channel was checked against YouTube itself: the title and the channel name under each one are what YouTube returns for that video, not what anybody remembered. That check earned its keep more than once, because a search result is quite happy to hand you a title that belongs to a different video. What could not be done is watch them all. So the line under each link comes from the verified title and channel, not from having sat through it. If one turns out not to be what it says, reply and it comes off the list. Same if a link dies, which they do.