Ambisonics is bad technology

Ambisonics, a technology used to create spatial audio experiences, is bad technology for the home (and most public performances)Before arguing this point, let me define how I am judging technology:

A technology is only as good as the experience users have using it.

If someone were to argue that ambisonics is potentially good technology, then I would agree. But there are so many ways that ambisonic playback can be compromised that it is more than likely going to result in a poor spatial audio experience. It is the experience that the user actually gets that counts… not what the technology is theoretically capable of. And that is where ambisonics fails dismally. It is a fragile technology that insists on a slew of preconditions.

Here is a list of just some of the things that can go wrong:

  1. The user cant easily find/install a software decoder
  2. When the user does eventually install a decoder, there is no guarantee that it is a proper one (includes shelf filters etc.)
  3. Connecting one’s computer to a multichannel sound system can be problematic
  4. The user’s speakers may not be matched
  5. The user may not be sitting in the centre of the speaker array
  6. The acoustic of the playback room may interfere with the ambisonic image

Compare all that to the experience of a high def flat screen TV. Buy it, take it out of the box, turn it on. There it is. For ambisonics to become good technology, it must achieve a high level of consistency of quality experiences … not just have the mere capacity to deliver good spatial audio.

I’ve sat in concert halls that are particularly reverberant. Seeing an array of speakers my first thoughts often concern how the reverberation of a hall will interfere with the spatial image. Recently I was surprised when the spatial imaging was far better than I had anticipated … I later learned that spatial location had been implemented in VBAP (Vector Based Amplitude Panning), an alternative spatialisation technology. Here’s an example of how a technology with far less potential can actually be more robust and therefore ultimately deliver a better experience. I would say it is a better technology.

I think the future of ambisonics is in fixed installations in concert halls; where the installation can be tightly controlled, and doing head-tracked HRTF decodes on portable media devices. Both would allow ensuring a consistent quality experience. Making ambisonics become good-technology in the home is possible, but I think it would require all sorts of funky automated things such as speakers that know where they are relative to each other; built-in mics that can measure and compensate for room acoustics; and mechanical robot devices to push listeners into the sweet spot when they’ve inadvertently decided not to sit in the middle of the room!

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

20 Responses to Ambisonics is bad technology

  1. Bogus Name says:

    You discuss only the setting up of an Ambisonic system. The same argument can be made about *using* an Ambisonic system (or anything else). If a file is two-channel UHJ then you have to know to press the UHJ button and to select the appropriate level of Forward Preference. The alternative is for the file to press the buttons itself using metadata in the file.

    This is why I have been pushing so hard for file formats for UHJ, G-Format, etc. See:
    http://members.tripod.com/martin_leese/Ambisonic/UHJ_file_format.html
    http://members.tripod.com/martin_leese/Ambisonic/G-Format_chunk.html

    Unfortunately. metadata does not help with the setting up.

    Regards,
    Martin

  2. Bogus Name says:

    There are also gender differences in the use of technology. See this recent article in E&T Magazine:
    http://eandt.theiet.org/magazine/2011/12/what-women-dont-want.cfm

    In the past, manufacturers would try to sell technology to women by making it smaller and painting it pink. They have finally realised that this doesn’t work and, instead, have started asking what women actually want. Manufacturers are starting to understand that to appeal to women they have to concentrate on simply making the technology work.

    Men tinker, whereas women look at devices from a functional perspective (does it do what I want?).

    Regards,
    Martin

    • Hi Martin,

      Yes, my partner has very little patience for using things that don’t “just work”… I can see that point.

      The thing about ambisonics for me is that I used to tinker to get everything set up properly. But when I started creating compositions that I wanted to share with others I realised that I was actually asking people to tinker-to-buggery as well.

      It’s like inviting someone to play a game of tennis, then insisting that they make their own tennis racket. It will be badly made … and they may not enjoy the game because of the bad racket. Actually, most people wont even show up for the game because they never completed the racket, or thought that it didn’t work well enough to bother playing.

      That’s what it is like asking someone to listen to ambisonics. The reality, I suspect, is that the majority of people who have the capacity to listen to ambisonics properly are perhaps more interested in tinkering anyway. Ambisonics is a strange mix of elegant mathematics, geometry, audio engineering and magic… I think this is what pulls most people to ambisonics.

  3. Elan Paul says:

    If somebody creates an integrated component that runs Ambdec and has a multichannel audio interface and provide information that includes links to affordable multichannel amps (e.g. Dayton Audio and HTD) and provides information on the correct speaker layout and links to loudspeakers with concentric drivers (eg. KEF, Tannoy) or other appropriate satellites then what is the big deal. Developing ambisonic systems may not be a no-brainer, but when people start to experience sound with height , they will realize that other systems don’t come close to reproducing sound the way we where naturally designed to hear, in 3 dimensions (and yes, there may be an argument for hybrid systems, but that doesn’t exclude what Ambisonics has to offer). The sonic “virtual reality” available with Ambisonics has the ability to transport people to other realms and take people out of there mind and body. Ambisonics isn’t for your living room, but what about other single user applications, such as game “pods” and relaxation and meditation parlors. What about as a tool for psycho-therapists? Etienne, if you have thrown your hands up, we are sorry to see you leave the playing field, but that game isn’t over! Your contributions to the field are greatly appreciated. It may be challenging now, but at some point, people will ‘get it’, and be able to get it. 🙂

    Elan

    • Hi Elan,

      ah … but I am not leaving the playing field! … on the contrary, my understanding is maturing. I have spent much time trying to help people get a good spatial audio experience with ambisonics, and the conclusion that I keep concluding, year-in year-out is that very few people are managing to arrive at the point where they can hear it.

      My point is that ambisonics has a long way to go — technologically — before people can and will ‘get it’ (at home). I don’t see that technological development happening right now. Asking people to install Linux to be able to access Ambdec is burying ambisonics, not exposing it.

      Please notice that the commercial entities having a go at pushing ‘spatial audio’ into the home are actually moving in the opposite direction to Ambisonics. These days, it is seen as a benefit that consumers could get a ‘spatial audio experience’ by using a single bar speaker. The benefit is they don’t have to have speaker cables and speakers all over the room. The commercial entities are recognising that even simplistic 5.1 is too hard for the consumer.

      Ambisonics is not about commerce … but business knows what people will and wont bother to do, and it is starting to recognise that even 5.1 is not consumer-friendly.

      • Elan Paul says:

        Etienne,
        Yes, you are likely correct when coming from that perspective, Ambisonics will never become consumer technology for the masses. And I agree that the capability of system is useless if the practicality is null. But I suspect, as experiential and interactive technology evolves, there may very well be a place for Ambisonics. Undoubtedly a very niche market at this point, understood. But as we know, sound has the ability to induce a shift in our mental and psychological (and sometimes physical) state of being. I believe that Ambisonics can be leveraged to enhance that affect and can have great value as such. Just not sure where yet! :). Glad to know you are still playing…

        Elan

  4. Thanks for writing this article, Etienne…..there isn’t enough critique of Ambisonics out there, currently. I’m not sure about your VBAP argument, however. Does that not suffer the same issues in that the speaker positions must be known (and how many etc.) for the system to work as the composer desired? How the different systems then break down when the system isn’t as desired depends on many factors, and is something that hasn’t been explored enough, as far as I’m aware.

    Of course, the Ambisonic decoder is a big barrier currently, which is why Ambisonic audio should be presented and distributed as speaker feeds…..the audience then need never know that it’s Ambisonics or VBAP or Pro-tools 5.1 panner, or Reaper’s or whatever. In they end they don’t, and shouldn’t care (generally). We’d never ask a comsumer to install pro-tools to listen to a mix made with their panner, we shouldn’t ask people to install anything to listen to Ambisonics either…..Ambisonics should be (and is in some circles!) a production format only at the moment.

    cheers

    Bruce

    • Hi Bruce,

      I think all spatialisation technologies will be knocked about a lot by the acoustic of the performance room … especially if it is reverberant. You probably know more about the subtleties of the various technologies than I do … but in this instance it turned out that only a semi-circle of speakers was used (limited to the stage). Ambisonics would not have fared well here … perhaps that is the principle advantage of VBAP … sounds can be localised within restricted layouts.

      The impression that the localisation was stronger than it would have been if done ambisonically was subjective… but I think I can explain it. Given the limited number of speakers (I think 4) I was surprised at how focused the sound source was. If 4 speakers had been used ambisonically (covering a 360 area … instead of a 180 deg area) the focus of the sound source would have been less. In this sense I guess you could say that I was projecting my expectation of the focus of sound location on what I knew was possible (ambisonically) with 4 speakers. To get that level of sound source focus on a 180 deg stage, with 4 speakers, in a reverberant environment, was impressive.

    • BTW … its interesting that one of Ambisonic’s projected strengths is its speaker agnosticism. Because this is also the feature that creates a huge barrier to entry in the form of software complexity / file format confusion etc.

  5. Too bad you have not tried Ambiophonics. I would love to have your comments on any of the Ambiophonic systems and gizmos now available. See http://www.ambiophonics.org for links to free apps, free plugins and the latest hardware components. There are some not there including a new Finnish loudspeaker that includes Recursive Ambiophonic Crosstalk Elimination-The Tabulum Qlara. I don’t sell anything or try to make money at this.
    Ralph Glasgal

  6. David WAreing says:

    It is quite difficult to get people to try anything new, particularly if it involves moving speakers..

    It took me most of a day searching the mass of Ambiophonic ‘stuff’, and trying various audio editors, to discover in the small print the simple fact that renaming a file from .amb to ,wav enables it to be read into AudioMulch and that c:/wvunpack.exe y:/jhroy_rain-over-the-tent.ua will make the four channel wav file y:/jhroy_rain-over-the-tent.wav

    It does not seem that anyone has tried, or is likely to try, an alternative way of decoding Ambiophonic files

    http://electro-music.com/forum/topic-49067.html
    “I have recently made an Ambisonic to Pan-Ambiophonic-thingy decoder. I hope you can try this Pre-processed files as you are probably one of few people set up to do so.
    Stereo-dipole – front channel 1.2 , rear channel.5,6, Channel 3.4 empty.
    Source file from here:
    http://soundofspace.com/ambisonic_files/58
    My dual Ambiophonic file here:
    http://www.megaupload.com/?d=7MX2W3FR
    AmbiFusion-DW-jhroy_hot-air-balloons.aac
    You should hear 3D ie. including height. ”

    Ambisonic recordings are just fine, but the idea that the maths supports the idea that one can recreate the soundfield, at even for a small blob of space, by making standing waves and interference patterns , is either wishful thinking, or somewhat dishonest. The process is irreversible, IMO.

    David.

    • Hi David,

      Do you mean ‘ambisonic’ instead of ‘ambiophonic’ in the second paragraph above … I’m confused (I’m not familiar with ambiophonics at all).

      I can take the blame for both soundofspace.com and the UA format … both are my work. I know it is all under-documented. I’m a one-man-band and open-source support keeps failing me.

      I think you are right about the difference between ambisonically recorded vs encoded sounds. Whilst the maths for encoding point sources of sound ambisonically appears quite simple, actually encoding the behaviour of sound in space is very complex (and requires pretty good knowledge of psychoacoustics).

      • David Wareing says:

        Hi Etienne,
        My previous attempt to reply does not seem to have made it to here, so I will try again.

        Yes it seems I had a ‘brain fart’. The source file is Ambisonic and the pre-processed file is for an Ambiophonic ‘PanAmbio’ arrangement of speakers.

        The speaker geometry I am using for this is a narrow rectangle, 30cm. across (between speaker centres) and about 2.5 metres between the front and rear pair.

        Additional information to Raph’s http://www.ambiophonics.org/ can be found here:
        http://wwwnew.isvr.soton.ac.uk/FDAG/vap/html/sd.html
        http://ambisonic.info/info/ricardo/ambiophonics.html
        http://www.filmaker.com/papers.htm
        http://www.princeton.edu/3D3A/

        My present filters are more optimised for intensity stereo than the pure X-talk cancelling filters described, which are aimed more towards recordings using some sort of microphone which includes a real, or simulated head shadow effect, and interaural delay, This should make them well suited to Ambisonic recordings. How it it is that it seems to reproduce height is a bit of a mystery, but I have some crackpot theories to work on..

        David.

  7. Pingback: Ambisonics is a bad technology? | A J Horsburgh

  8. Pingback: Audio Central Magazine » Audio digitale: l’ascolto (è) consapevole?

  9. Jean says:

    Hi Etienne,

    What do you think about this announcement? Could this be a game changer? What is the difference with existing ambisonics developments?

    http://community.calrec.com/technicolor-pushes-for-universal-3d-audio/

    Thanks

    • Ambisonics is now about 40 years old and is in many ways an analog conception. It is difficult to configure for home and is incompatible with the existing library of LPs, CDs, DVDs, MP3s, etc. Now it is much easier to get a full circle of direct sound localization in the horizontal plane using just four speakers and any 4.0 media and just 4 microphones or electronic music sound channels. (An Ambiophone if absolute perfection for acoustic recording is desired). Also the many apps and components are compatible with the existing library of DVD, BD, and even CDs and LPs where 2.0 files can produce stages of 180 degrees and 4.0 files (5.1 DVDs etc.) a full 360 degrees if the recording is like Avatar or similar. But this type of new technology is for home use only, not theaters or mass listening.

      Ralph Glasgal
      glasgal@ambiophonics.org

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s