Re: [linux-audio-dev] [ann] guitar preamp plugin, v2
On Wed, Nov 05, 2003 at 10:43:17AM +, Steve Harris wrote: Nope, polyphases only filter at fs/(2^n), they are used for resampling and reconstruction. You seem to use a very odd definition of 'polyphase filter'. In fact any FIR you like can be turned into a polyphase filter - just interpolate the coefficients in the same way you'd interpolate a signal. -- Fons
Re: [linux-audio-dev] [ann] guitar preamp plugin, v2
On Wed, Nov 05, 2003 at 08:06:38PM +0100, Fons Adriaensen wrote: On Wed, Nov 05, 2003 at 10:43:17AM +, Steve Harris wrote: Nope, polyphases only filter at fs/(2^n), they are used for resampling and reconstruction. You seem to use a very odd definition of 'polyphase filter'. In fact any FIR you like can be turned into a polyphase filter - just interpolate the coefficients in the same way you'd interpolate a signal. OK, I misunderstood what makes it polyphase then - thanks. - Steve
[linux-audio-dev] Microsoft gets a patent on partitioned convolution
This might be interesting to some of you: The newly granted patent US6625629 System and method for signal processing using an improved convolution technique owned by Microsoft, covers partitioned convolution. The patent was filed in May 2000, and granted in September 2003. This is in most claims the same convolution algorithm previously patented by Lake Technology with priority date back in July 7 1992, US5502747. And that is in most claims the same algorithm as invented and published by Soo and Pang and others prior to 1992, which is the same algorithm used in my software BruteFIR. Isn't the patent system wonderful? The requirements new and non-obvious were in practice scrapped a long time ago... Anyway, Microsoft does as far as I know not have the tradition to harass free software developers as Lake has, so I don't think the fact that another one owns the partitioned convolution algorithm will cause any more trouble. It would be interesting to know what Lake thinks about this. If Microsoft has patented it, they probably use it, and then Lake might want them to pay royalty, or else /Anders Torger
[linux-audio-dev] swh-plugins for jackEQ?
I'm trying to run jackEQ. If I use the current release of swh-plugins, it fails because dj_eq_1901 is missing. If I use the latest snapshot I can find on Steve's web site, swh-plugins-2003-10-21.tar.gz, this plugin exists. But using that, jackEQ and JAMin both segfault on startup. GDB shows a segfault in lookahead_limiter_1435.xml, line 110. So, what version of swh-plugins works for jackEQ? -- Jack O'Quin Austin, Texas
[linux-audio-dev] [ANN] GNUsound 0.6.1 released
Hi, GNUsound 0.6.1 was released. This is a maintenance release to fix a few critical bugs. I really wanted to skip this release and go straight for 0.7, but there's too much work to be done on that version still, and the 0.6 bugs are quite serious. Changes from version 0.6: Fixed bug where channels past 4 wouldn't play (by removing unrolled mixer_mux()'s, they hardly added any speed anyway). Fixed possible crash drawing markers (Jens Mauer). Fixed crash canceling File Save dialog. Fixed crash on machines with MMX but without MMX2 (original Pentium). Download at the usual place: http://awacs.dhs.org/software/gnusound Thanks, Pascal.