Synthesizing impulse responses

Everything and anything, but nothing about the LinuxSampler project.
User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Synthesizing impulse responses

Post by dahnielson » Sat Mar 22, 2008 12:52 am

lowkey wrote:Could light refraction formale be used? Although light is faster than sound they are still waveforms.
Shure, Snell's law apply to acoustics although only total internal reflection is mostly of interest. But since the velocity of the sound is lower in a wall than in air the angle of refraction would be less than the angle of incidence, and going from a fast to a slow (denser) medium will decrease the wavelength. But I assume because it has to go back into air and the spatial shift caused by the refraction doesn't matter that much. I guess transmission loss is more important to sound than refraction: "When sound hits a wall there is a certain proportion of the sound reflected back into the room, some is lost in the absorption of the wall and the rest travels through the wall and is called the transmission loss."
Last edited by dahnielson on Wed Mar 26, 2008 2:40 am, edited 2 times in total.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Synthesizing impulse responses

Post by dahnielson » Sat Mar 22, 2008 2:02 am

Extending my previous description these nice impulse response plots will now make sense without being blinded by the equations that goes along with them:

* http://www.dspguide.com/ch6/2.htm

* http://www.dspguide.com/ch7/1.htm

BTW, I've been aware of this book for a long time (think I've even downloaded it completeley as PDFs a couple years ago) but haven't really read it until now. Here's some words from the introduction that I couldn't agree more with:
This recent history is more than a curiosity; it has a tremendous impact on your ability to learn and use DSP. Suppose you encounter a DSP problem, and turn to textbooks or other publications to find a solution. What you will typically find is page after page of equations, obscure mathematical symbols, and unfamiliar terminology. It's a nightmare! Much of the DSP literature is baffling even to those experienced in the field. It's not that there is anything wrong with this material, it is just intended for a very specialized audience. State-of-the-art researchers need this kind of detailed mathematics to understand the theoretical implications of the work.

A basic premise of this book is that most practical DSP techniques can be learned and used without the traditional barriers of detailed mathematics and theory. The Scientist and Engineers Guide to Digital Signal Processing is written for those who want to use DSP as a tool, not a new career.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Synthesizing impulse responses

Post by dahnielson » Mon Mar 24, 2008 7:19 pm

Here's a product I'm currently fascinated by simply because I don't fully understand how it works, yet.

http://www.prosoniq.com/main/rayverb/

It's basically the opposite of my IR synthesizing reverb.
Inverse raytracing looks at a sampled room impulse response and adapts a general room model to reproduce your particular sonic space. Not only does this capture your acoustic space with ultimate clarity - it also allows for slight as well as dramatic changes!
My guesstimate is that the general room model would be the sum of a rectilinear room with parallel floor, ceiling and walls.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Synthesizing impulse responses

Post by dahnielson » Tue Mar 25, 2008 4:07 pm

Just a reminder to myself: After developing an acoustic engine capable synthesizing impulse responses by ray-tracing (which sounds fancy but is nothing more than a bit of point reflection, trigonometry, absorption coefficients and inverse square laws) I'll look into "the concept of reverberation fields" (a term I just made up) for sound radiation, propagation and reflection as a generalized method.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Synthesizing impulse responses

Post by dahnielson » Fri Apr 04, 2008 11:53 pm

Some added research from the Aella-slides that can come in handy for massaging ("re-synthesizing") existing IRs:

Early Reflections
  • ER occur in the first 5...80 ms after the direct sound.
  • They are not heard as a separate sound but merge with the DS.
  • The pattern and directions of ER provides information about the size and shape of the acoustic environent.
  • ER patterns depend on source location, and must be correct in order not to contradict the direct sound.
Reverb Tail
  • The RT is heard as a distinct separate sound.
  • Typical decay is exponential, with exceptions.
  • The RT depends again on source location, but we can’t hear the difference in most cases.
  • One RT concolution can often be shared for all sound sources that appear from the same general direction.
(My emphasis.)
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

Post Reply