Here’s a solution to music piracy…

Imagine if you downloaded an MP3 (or uncompressed WAV) from a site … which had no DRM (Digital Rights Management) in it whatsoever … yet when someone else listened to it (whether on your gear, or by copying the file) it just didn’t sound as good.

Not possible you say? … well it is! …. by using persomalised binaural renderings of surround tunes. That means that someone measures your ears, and uses those measurements to mixdown a surround tune to stereo. When you listen to that stereo mixdown, you hear what you would have heard if the tune was played back over a 5.1 system. When someone else listens to that tune, it doesn’t quite sound right … the ‘surround sound’-ness of the tune is somewhat lost … they cant tell whether a sound is coming from the front or behind.

Mark Waldrep of AIX records recently posted a report on the Sursound list about a demonstration of a very accurate (personalised) stereo mixdown of a 5.1 recording. Ofcourse, accessing such a mixdown would require that one has a set of personal ear measurements.

Clearly, the challenge here is providing a mechanism for consumers to obtain personal ear measurements. If that challenge can be met … then there’s a huge opportunity to generate audio files with no DRM, but whose value is lost when consumed by others.

Now, clearly, I’ve got my ‘business hat’ on when I say all this. I’ve spent the last month projecting figures on various business models for Ambisonia …. and this kind of scenario is a huge opportunity. Its an opportunity not because it has a kind of ‘natural’ DRM-effect built into it …but rather because it offers an iPod-compatible augmented personal listening experience… which is not experienced by others on the same physical digital file.

Such a model fits very well with Ambisonia. Imagine this … upload your personal ear measurements … pay $5/month and get all of Ambisonia’s content in personalised surround stereo mixdown! … the only challenge is … who’s got their own personal ear measurements?(!)

This entry was posted in Uncategorized. Bookmark the permalink.

7 Responses to Here’s a solution to music piracy…

  1. Nick says:

    It’s an interesting idea.
    There are many private and public (institutional) researchers presently working on ways to make personal HRTF generation easier, for example, by generating HRTFs from photographs of the user’s ears. In fact much improvement is possible (especially for horizontal localisation) only by customising for the user’s head dimensions, which affect ITD and IID.

    That said, there is also evidence that people can adapt to modified (and potentially other people’s) HRTFs and recover virtually the same perceptual acuity as for their own ears. – there’s a famous article in Nature Neuroscience journal about this:
    Relearning sound localization with new ears
    Paul M. Hofman, Jos G.A. Van Riswick, A. John Van Opstal.
    Nature Neuroscience 1, 417 – 421 (01 Sep 1998)

    with this introduction article:
    Of vulcan ears, human ears and ‘earprints’
    Fred Wightman, Doris Kistler
    Nature Neuroscience 1, 337 – 339 (01 Sep 1998)

    Also, many existing binaural products rely on people having quite a good experience with non-individual HRTFs – e.g. Dolby Headphone.

    It pretty much goes without saying the best binaural experience should be with your own HRTFs, but it’s possible to have a very good experience with non-individualised HRTFs as well.

  2. thanks for those references Nick.

    The whole binaural thing is very tricky, because it _is_ a degraded surround experience compared to speaker arrays … but at the same time it is an extremely accessible form of surround.

    Ideally, I’d like to be able to supply ‘best of breed’ binaural decoding on Ambisonia. Most of the binaural decodes I’ve heard have been disapointing. I think a proper binaural decode will require more than just the application of HRTFs.

  3. Henry says:

    I quite liked Derek Silvers take on music piracy:

    Which says,

    “Putting so much attention and energy into fighting piracy (as if, when solved, you’ll suddenly start selling 10 times more) – is misguided effort, distracting you from what you really need to be improving.

    That’s the real reason I often tell musicians not to worry about piracy. I’m not saying it doesn’t exist. But energy spent worrying about it is energy better spent working on what you know you really need to do. ”

    To this end I’m enthusiatic about seeing more UHJ or generic binaurals available, to address headphone listening. I imagine surfing to Ambisonia as an outsider and being disappointed at not being able to listen to any of the content immediately, almost regardless of the surround content.

  4. Nick says:

    Binaural decodes (at least mathematically) don’t have to be a degraded experience compared to speaker arrays.

    Imagine (or build) your best possible ambisonic listening room setup with your ideal speaker array…. then go and sit in the ideal listening spot, fix your head in place and run a process to capture the binaural impulse response from each speaker to each ear. Now you have a set of HRTFs that when convolved with an ambisonic decode intended for that speaker array, will give you the exact signals for headphones that you would hear if you sat in that array. This is an ideal binaural decode!

    But there the problems begin: first, the decode involves your personal HRTFs so it won’t necessarily work as well for other people. This you can only solve with some individualisation strategy. However, you could make a technique that comes very close to separating the effects of the listening room from the individual person, so you can keep the ideal room simulation and just change the listener (by using their anechoic, recorded or otherwise personalised HRTFs). This is still pretty impractical, so the best of breed still uses some form of generalised (non-individual) HRTF.

    The other big problem with listening to your own (mathematically perfect) binaural decode is the multimodal effects. Eg, if you listen to that decode outside of the represented listening room (particularly if you can see your surroundings), you’re likely to know you’re not in the same space, so you’ll perceive the simulation differently. In fact any actual difference might be perceived – eg the feeling of pressure of the headphones on your head can degrade the experience – or a bad headphone frequency response could massively damage it!

    In the end, the mathematically perfect solution is not perceptually perfect for many uncontrollable reasons.

    So binaural is really about making do despite practical limitations. Probably the most dramatic perceptual improvement would come from the inclusion of a good, diffuse but subtle binaural room presence, not personalised HRTFs. To decide what’s definitely best of various options is a tricky business in itself, requiring a decent perceptual experiment design.

    I guess whether or not an imperfect solution is worth implementing might come down to other business factors like a cost-benefit analysis…

    I dunno – if you’ve heard one less-disappointing binaural decode, why not implement that?

  5. Henry,

    You are absolutely right, that many (most, in fact) people would discover Ambisonia, think how cool it is, then feel disappointed with their inability to listen to the recordings.

    Binaural is one solution.

    Streaming AAC is an other. Streaming AAC is coming. See the tests here:

    Streaming AAC is not binaural … its a 4 speaker decode … Quicktime collapses the rear 2 speakers into the front 2 speakers, when it detects a stereo output. Not ideal, but at least you can listen. (still trying to find out if I can stop the rear speakers collapsing into the front).

    Binaural decodes _will_ come…. but not in the immediate future. I still havn’t found a binaural decode that doesn’t confuse front-back and has good externalisation (except for the Dolby headphone stuff, well … that’s the closest).

    BTW I’m not concerned about music piracy. I just found it very interesting that there is a ‘natural’ DRM possibility, by having personalised binaural decodes.

    I’m still looking for the right ‘generic’ non-personalised binaural decode.

  6. Nick,

    “run a process to capture the binaural impulse response from each speaker to each ear” … that’s recording personalised HRTFs isn’t? … and how would Joe Blow, average consumer with home theatre do that?

    I have read (including that recent post on Sursound by Mark Waldrep) that personalised HRTF’s (in combination with head tracking) can produce a ‘practically’ identical result to a speaker setup. It makes sense.

    But the challenge is how to deliver that. If it cant be delivered, then it is of little interest/use.

  7. nmariette says:

    Yes, well at the moment, it can’t be delivered practically. So the question remains – do you bother with what’s practical or forget about it altogether.

Leave a Reply

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

You are commenting using your 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