See http://www.johnson-amp.com/pdf/CabinetImaging.pdf for info on speaker simulation.
Crossover distortion has nothing to do with x-overs, it's the distortion effect of a push-pull power amp at phase crossover point where positive and negative phases don't meet at where they should. Best way (but not effective) to simulate speaker would be to sample the impulse response and use convolution (like AcousticMirror does). I guess Johnson and Line6 use other regular IIR and FIR filters. I thin there are some filter design apps available for the job. -Mikko > -----Original Message----- > From: ext Erik de Castro Lopo [mailto:[EMAIL PROTECTED]] > Sent: 15. January 2002 9:16 > To: [EMAIL PROTECTED] > Subject: Re: [linux-audio-dev] New plugins, fixes etc. > > > On Mon, 14 Jan 2002 23:43:18 +0000 > Steve Harris <[EMAIL PROTECTED]> wrote: > > > http://plugin.org.uk/releases/0.2.3/ > > > > * Spellunk fixes (thanks to Frank Neumann) > > * Almost all plugins now have working activate methods, so > they won't have > > spillover if you stop and restart them (on propertly wrtten hosts) > > * Added crossover distortion > > * Added tube rectifier simulation > > > > The additions mean that it should be possible to simulate a > tube (valve) > > amp (but not cabinet yet) with a chain of: > > > > tube (or harmonic gen) -> tube rectifier -> crossover dist -> EQ > > I have never seen a real tube guitar amplifier with a crossover so the > crossover should probably be ditched. Because guitar amps usually have > 10, 12 or 15 inch speakers (from 1 up to 8) and rather large > cabinets the > correct EQ should probably be a high order (8 to 12) > butterworth lowpass > at somewhere between 4Khz and 10kHz to accurately model the frequency > response of the speakers and cabinet. > > > I haven't forgotten about the compressor, I just haven't > worked up the > > energy to tackle it yet. > > I have a rough hack of a compressor (but nice sounding) I'm > getting ready > for a tutorial I'm presenting at Linux.conf.au in a couple of > weeks. I > should probably downlaod the LADSPA stuff and see if I can > create a plugin. > > My compressor currently has: > > - switchable feed-forward / feed-back architecture > - attack time > - hold time > - release time > > The attack and release timers currently do exponential ramps, > but I should > also look into doing linear ones. > > Cheers, > Erik > -- > +-----------------------------------------------------------+ > Erik de Castro Lopo [EMAIL PROTECTED] (Yes it's valid) > +-----------------------------------------------------------+ > Linux: Because rebooting is for adding new hardware >
