No announcement yet.

Ambiophonics - An experiment and a How to using foobar2000.

  • Filter
  • Time
  • Show
Clear All
new posts

  • Ambiophonics - An experiment and a How to using foobar2000.

    To give you the basic idea, ambiophonics aims at "fixing" what is wrong with stereophonic sound-i.e. removing acoustic crosstalk and opening up the 'sweet spot' and giving you that sense of "you are there". It's based on research about human hearing, concert halls, and recording methods that combines the best aspects of binaural listening+ambiosonics+monophonic.

    It's hard to really sum it up, especially since I'm still researching it myself, but I wanted to share the opportunity with others on how you can try it, since some of us are concerned with realistic music reproduction.

    For more info:

    Now on to the experiment for you to try out.

    Part I

    A little experiment - First Try this. I'm not going to go into any details on what this is. I suggest you first just try the experiment, then go back and read about it. But if you must know what this is all about it's all here:

    What you need
    -Pair of stereo speakers
    -2 channel amp/receiver
    -computer connected to receiver/amplifier or burn the pre-processed files to CD
    -your ears

    Place your speakers at a distance of 50cm apart from each other. Your listening position should be about 2m back from them so that they form ~+/-20deg angle between the line straight in front of you and the speaker.

    Like this (but without the barrier)

    Then listen to the “left right width test” from this link and find the spot at which you notice the widest stage width. For more setup info:

    Setup & Calibration
    Ambiophonics is optimized for one or two listeners on the medial plane of a pair of front speakers separated only about 50~75cm positioned at ear level about 2~3m from the listener(s). Choueiri CXC is optimized for separation of 49.5 cm positioned 1.9 m from the listener(s). Speakers should be matched pairs, known to have good phase response through any crossover networks. Correct speaker positioning, avoiding acoustical interference, and calibrating levels and timing are essential for successful crosstalk cancellation.
    Your ears are the ultimate instruments of perceived quality. But first, calibrate speaker SPL using an SPL meter ($30 at Radio Shack) and this link to band-limited pink noise. Burn this file, the Left-Right Width Test, and any desired demonstration files above to an audio CD. With no crosstalk cancellation processing, the player set for loop repeat of the pink noise, and SPL meter fixed at the listening position, adjust each channel individually using level controls at the power amplifier inputs so that the signal (-18dB below digital Full Scale) produces precisely 85dB SPL from each speaker individually.
    Still with no crosstalk cancellation processing, play pre-cancelled music excerpts above to verify speaker and listener positioning. Now with crosstalk cancellation processing, play pink noise through only the left input of the system and, if necessary, adjust speaker separation slightly relative to a fixed listening position so that at the listening position the sound is perceived at the extreme left only. Repeat iteratively using the right input only (for extreme right perception) until the widest stage width is achieved, up to 120 - double stereo's 60 - but with none of stereo's "hole in the middle" or distortion of tone color, especially for important central sounds.
    I suggest you also listen to the other music files also on that page after you get good results with the width test.

    Now I will list what you need to try it with all sorts of music.
    System always under construction

  • #2
    Part II - Process stereo files into ambiophonic sound in real-time

    Part II - How to do this in real-time to any music you already listen to.(setup without pictures)


    removed because they don't work right. Stick with the audiomulch instructions on their page for now.

    EDIT: Update 3/5/09 - How to use it with winamp. Loading VST plugins into Foobar is similar.

    Things You Need
    -Computer with soundcard
    -Pair of speakers, and amp
    -VST Plugin Bridge
    -MOSC MAP ambiophonics VST plugin
    -Left/Right width track
    Setup your speakers so they are approximately 15-30cm apart. As shown here:

    1) Install Winamp
    2) Install VST Bridge
    3) Extract Mosc's VST plugin into any folder-or just put it into winamp's folder for easy retrival
    4) Open Winamp, goto options, preferences and select DSP plugin from plugins section.
    5) Select VST bridge plugin
    6) Click Configure Active plugin
    7) Click in VST plugin box and "load dll"
    8) Next browse to wherever you extracted the Mosc MAP VST.dll and select it
    9) Make sure the Ambiophonic button is green and enjoy any music now processed ambiophonically.
    10) Sit back about 1-2m away. You can use a Left/Right width test track also available in the electro-music thread to find optimal listening position. Notice how you can move your head around without a severe loss of the sweetspot.
    System always under construction


    • #3
      Nice to see some possible Ambiophiles on the forum. Such a setup is perfect for my show car and is the main reason for my carpc (only way to get a realtime ambiophonic processor without spending big dollars).


      • #4
        I've been researching it for little while but really dug into this past weekend. So you are familiar with this methodology. Care to add your thoughts? I'm a complete n00b but I'm soaking this up like a dry sponge.

        Do you plan on using Foobar or are you going to go the Console/VST plugin route?

        How do you plan on implementing the speaker location part of it? I haven't found a suitable location in my vehicle for speaker placement that would allow this. I could use my current setup, but the spacing is far greater than what they suggest of ~50cm.

        Is your car free of a center console?
        System always under construction


        • #5
          I was planning on using Audiomulch with VST plugin for the realtime dsp combined with a compatible player (something that runs under media portal)

          Speaker location is easy for me - Show car is getting a complete dash rebuild so speakers in the center of the dash will be an easy fit - my issue is they have to be easily removeable as it doubles as my SPL car. I also plan to have a 3 speaer rear stage for the ambient/room effects, as my cabin area is very small (walled 4wd).


          • #6
            ahha. So you are pulling out all the stops. Show car as in IASCA, MECA, USACI, or EMMA?

            Have you actually used Ambiophonics before then? If so, do you prefer method A or B? Have you considered VBAP since you are going all out?
            System always under construction


            • #7
              Show car as in IASCA, MECA, USACI, or EMMA?
              We don't have either of those formats down here, just local shows/sq comps and dBdrag.

              But yes - going all out is prety accurate.

              Have you actually used Ambiophonics before then?
              Not in real time processing, done a little playing around with two speakers and the pre-processed tracks is all. Was suitably impressed that i started planning a proper setup.


              • #8
                Nice, I'm not going to ask where you got that, guessing it was probably custom made.

                Down here as in Australia/New zealand? or just southern US? Just wondering.
                System always under construction


                • #9
                  Sorry - New Zealand (thought I'd set my location on my profile - obviously I had not).


                  • #10
                    Ah I see. You new zealanders are on top of your game. Abmolech is another New Zealander I believe and he got me pointed to this stuff.
                    System always under construction


                    • #11
                      Seems to be an interesting concept...though I'm not sure I comprehend 100%.

                      At first I thought it was the concept of getting more lifelike sound out of a system. Then I thought it was aimed at getting that sound out of different speaker placement. Then I thought it was aimed at getting that sound out of a pair of speakers placed within relative close proximity.

                      Now I'm just not sure.

                      I don't see anywhere that describes what it is and HOW they propose to go about achieving it. In other words, the actual method. From what I've read, it's nothing more than a theory at this point that's not even made it to testing...

                      Reading this link:
                      vaguely discusses that it's a method of treating the listening space to achieve more realistic sound but it doesn't go into hardly any detail about what makes it so different from currently sought after methods of room treatment. Perhaps I'm just dense?
                      Jan Bennett
                      FS: VW MKIV Bezel for 8" Lilliput - 95% Finished

                      Please post on the forums! Chances are, someone else has or will have the same questions as you!


                      • #12
                        The main principal is a crosstalk eliminator combined with close proximity pair of speakers for the front stage to give greater width and better sound localisation than stero.

                        All surround speakers are simply there to simulate room acoustics.,


                        • #13
                          The close proximity part is what I thought.

                          Just seems odd that they would claim all of this with a 24 speaker set-up...a room should sound life like
                          Jan Bennett
                          FS: VW MKIV Bezel for 8" Lilliput - 95% Finished

                          Please post on the forums! Chances are, someone else has or will have the same questions as you!


                          • #14
                            Actually a major improvement can be seen with just 2 speakers. The extra speakers recreate more ambiance/reflections but is not necessary however he states if you want to extend the listening width from 110-120deg to 180 or even 360, you can do so by adding more and talks about how to do it.

                            RED-You should try it with the samples they have and then use foobar to play any of oyur normal listening music through the convolver with the impulse files.

                            Basically all this theory, is pretty much the EXACT opposite of how "stereo" (equilateral triangle) sound works but works with normal everyday recordings, unless it was all synthesized.

                            It has been done, but plenty of research continues to improve it. There are even consumer grade processors out there that have used an early form of this. Tact being one of them. I think Method B is based off the newest research.

                            Look at the figures on that page, a picture is worth a thousand words sometimes.

                            It basically takes binaural (normal human hearing)/(headphone listening) and externalizes to loudspeakers.
                            System always under construction


                            • #15

                              There is something fishy about the way the foobar convolver plugin works and it doesn't do justice to this method. I experimented with the audiomulch method and it's definitely doing something different and it's much much much better. I read up on the convolver plugin and it appears to not work properly with those impulse files and it needs cross channel convolution capabilities.


                              If I can find another way to use another convolver plugin inside winamp or foobar I'll revise the experiment instructions. There appears to be way too much reverb and too much phase issues and I think it's because the foobar convolver is not setup for multichannel impulses.

                              Stay tuned...

                              P.S. The audiomulch method works great though Bye bye stereophonic sound, hello ambiophonics.
                              System always under construction