6 ms·

"Only possible in the last 10 years" is not correct. Computer musicians have used the IRCAM ISPW board as far back as 1989 to implement live real time convolut

"Only possible in the last 10 years" is not correct. Computer musicians have used the IRCAM ISPW board as far back as 1989 to implement live real time convolution. Single purpose realtime convolution reverb hardware has been commercially available since Sony's 1999 DRE S777. Non-real time convolution has been done for decades as well, including by myself in 1991 and likely hundreds of others during that time period. Real time convolution on an off the shelf affordable consumer platform has been available for a long time as well, including eMagic's Space Designer plugin which was released in 2003. It's still part of Apple Logic.

James_Henry 7y ago
I think Abel was talking about his and his colleagues work in this paper when he said the 10 years comment: https://www.researchgate.net/profile/Bissera_Pentcheva/publication/277732009_Estimating_Room_Impulse_Responses_from_Recorded_Balloon_Pops/links/563a4b4408ae45b5d284a8ce.pdf https://www.researchgate.net/profile/Bissera_Pentcheva/publi... He's not just talking about being able to do convolution.
- red-indian 7y ago
  To be perfectly clear, some have been sampling reverb spaces for use in convolution reverbs using a starter pistol since the 1980s, and the procedure became widespread in the 1990s. It is not 10 years old at all in any imaginable interpretation. This person's "innovation" of popping balloons instead of using a starter pistol is completely and totally irrelevant and inconsequential.
repiret 7y ago
Right. I'm curious what the recent advance is. "Pop a balloon to get the impulse response and use that to build a linear filter" was well known when I was doing signal processing work in the very early 2000's.
8bitsrule 7y ago
My reaction to '10 years' also. I recall using some (forgotten, Mac) inpulse conv in the early 2000s. (Not real-time!) This SOS article links many reviews dating back to 2002-04 (the earliest for the Audio Ease 'Altiverb'). https://web.archive.org/web/20161209164745/http://www.soundonsound.com/techniques/convolution-processing-impulse-responses#para11 https://web.archive.org/web/20161209164745/http://www.soundo...

James_Henry 7y ago

I think Abel was talking about his and his colleagues work in this paper when he said the 10 years comment: https://www.researchgate.net/profile/Bissera_Pentcheva/publication/277732009_Estimating_Room_Impulse_Responses_from_Recorded_Balloon_Pops/links/563a4b4408ae45b5d284a8ce.pdf https://www.researchgate.net/profile/Bissera_Pentcheva/publi... He's not just talking about being able to do convolution.

red-indian 7y ago
To be perfectly clear, some have been sampling reverb spaces for use in convolution reverbs using a starter pistol since the 1980s, and the procedure became widespread in the 1990s. It is not 10 years old at all in any imaginable interpretation. This person's "innovation" of popping balloons instead of using a starter pistol is completely and totally irrelevant and inconsequential.

repiret 7y ago

Right. I'm curious what the recent advance is. "Pop a balloon to get the impulse response and use that to build a linear filter" was well known when I was doing signal processing work in the very early 2000's.

8bitsrule 7y ago

My reaction to '10 years' also. I recall using some (forgotten, Mac) inpulse conv in the early 2000s. (Not real-time!) This SOS article links many reviews dating back to 2002-04 (the earliest for the Audio Ease 'Altiverb'). https://web.archive.org/web/20161209164745/http://www.soundonsound.com/techniques/convolution-processing-impulse-responses#para11 https://web.archive.org/web/20161209164745/http://www.soundo...