Re: [music-dsp] 20k

On 2015-08-31, Theo Verelst wrote: [...] because for the second sample of your own input to the linear system represented by the convolution of the measured input with your own input signal, the air mass isn't anymore "at rest", or in other words there are different initial conditions, so line

[music-dsp] 20k

On 2015-08-31, bzk0...@aol.com wrote: Real-time auralization of dynamic sound sources in Virtual Environments would be one application. Myself coming more from the graphics/interaction side, our fellow acoustician colleagues do this in order to couple arbitrarily moving sound sources with a po

Re: [music-dsp] music-dsp Digest, Vol 2, Issue 47

On 2015-08-31, andy butler wrote: I can see how FFT based processing can be equivalent to a large FIR in terms of frequency response, but not how it would be useful for a convolution reverb. The thing about FFT based processing is that it does pretty much the same for convolutions as logarit

Re: [music-dsp] [ot] 20k

On 2015-08-30, robert bristow-johnson wrote: And you know, you can do kernels hundreds of millions of samples long that way in real time, on what machine? does it breathe fire? Basically, yes. The example I've seen came from radar hardware, once again. It was liquid cooled custom ECL logic

Re: [music-dsp] 20k

i think the patent is to use direct convolution for the first partition to get zero latency > On Aug 30, 2015, at 11:52 PM, Victor Lazzarini > wrote: > > What does the partitioned convolution patent cover? > > Victor Lazzarini > Dean of Arts, Celtic Studies, and Philosophy > Maynooth Univers

Re: [music-dsp] 20k

Scott Gravenhorst wrote: music-dsp@music.columbia.edu wrote: >On 2015-08-30, Scott Gravenhorst wrote: > >> This amounted to using a microphone to sample the effects of an >> impulse (starter's pistol or some such) on some audio environment like >> a church or concert hall, -- ScottG I

Re: [music-dsp] music-dsp Digest, Vol 2, Issue 47

Have a look for this project, it includes a wrapper for the vDSP FFT, Ooura's FFT and FFTW, with uniform partitioned convolution algorithms https://github.com/HiFi-LoFi/FFTConvolver Ivan MusicalEntropy.com Blog | Twitter

Re: [music-dsp] music-dsp Digest, Vol 2, Issue 47

Searching online for "fast convolution" will give you plenty of sites showing how it works. There is a special relationship between "time domain" (stream of samples) and "frequency domain" (the world of frequency responses, FFT etc), stated briefly as: * multiplication in the time domain (e.g.

Re: [music-dsp] music-dsp Digest, Vol 2, Issue 47

This article explains the relation between convolution and FFT + spectral multiplying and has some pointers to further literature: https://en.wikipedia.org/wiki/Overlap–add_method Today, you have no specialized processors for FFTs. The

Re: [music-dsp] 20k

On 31/08/2015 5:05 PM, Richard Dobson wrote: The Lake patent never covered plain equal-size partitioned convolution (there were indeed extensive debates about this at the time), an example of which was even included in the "Numerical Recipes" book, but specified mixed-size convolution with the ve

Re: [music-dsp] music-dsp Digest, Vol 2, Issue 47

Can someone clarify/explain. Got a notion there's something I'm missing here. I can see how FFT based processing can be equivalent to a large FIR in terms of frequency response, but not how it would be useful for a convolution reverb. I'd always thought that Cool Edit/ Audition used an efficien

Re: [music-dsp] [ot] 20k

On Sun, 30 Aug 2015 20:37:58 +0300 (EEST) Sampo Syreeni wrote: > On 2015-08-30, gwenhwyfaer wrote: > > > Cool, someone else who's as loath to join the 21st century as I am... > > This kind of makes you wonder about when you'd actually try to brute > force tens of thousands of sample long FIR's

Re: [music-dsp] 20k

Not every reverb company. There was one little company, which was smart enough to use the vDSP framework from Apple. And from there on, there was no need to change a single line of code. Steffan > On 31.08.2015|KW36, at 04:15, Tom Duffy wrote: > > When Macs switched to x86, all the reverb

Re: [music-dsp] 20k

You would need to define what you mean by "real time" in this context. For music, the primary issue for "real-time streaming" is latency. Ostensibly, if you need to process 1M samples as a block, you can't generate the output (however fast you manage the actual computation) until you have recei

Re: [music-dsp] 20k

The Lake patent never covered plain equal-size partitioned convolution (there were indeed extensive debates about this at the time), an example of which was even included in the "Numerical Recipes" book, but specified mixed-size convolution with the very first (short) block using time-domain FI

Re: [music-dsp] 20k

> Am 31.08.2015 um 01:43 schrieb Scott Gravenhorst : > > Or is it possible that these sites say "convolution reverb" but internally, > it is faked using FFT? In my understanding "Convolution Reverb" doesn't say anything about the implementation method, only that the result should be the convoluti

Re: [music-dsp] 20k

What does the partitioned convolution patent cover? Victor Lazzarini Dean of Arts, Celtic Studies, and Philosophy Maynooth University Ireland > On 31 Aug 2015, at 03:15, Tom Duffy wrote: > > History review: > > When convolution reverbs first became a computer real-time > solvable solution, th