Quote:
Originally Posted by richard plumb can you mix a binaural recording from a master 5.1/7.1 movie soundtrack? |
I was going to just say "no" in response to that, but I guess it deserves a rather more detailed answer.
First, let me define what a true binaural recording is. A genuine binaural recording is made by constructing a model head, and then putting microphones in its ears. The head is not a simple construction, incidentally - the materials it is made from have to respond to sound in the same way as a real head would, so you have to choose them carefully to have the correct density and firmness, as well as shaping them like a real head.
So, what a binaural recording does is to record what is happening
actually inside the ears of a person who is listening to whatever is being recorded. You then use headphones to recreate that same sound inside the ears of a person listening to the recording. The two can match up very well.
The most common application would be to record something like a classical concert. But there are also a few recordings available of things like radio plays.
Binaural recordings aren't perfect, for a number of reasons, but they can achieve things that are impossible with speakers. Most notably, it is possible to make a sound appear to come from a position that is much closer to you than the speakers are. Speakers cannot reproduce the experience of having someone directly whisper into the listener's ear, but a binaural recording can.
If you want a quick sample, there's an "extra" on the DVD of
Monsters, Inc. which was recorded binaurally - Billy Crystal and John Goodman talking into the microphone/ears. The effect is really quite creepy!
Now, you ask can if we can create a binaural recording by processing a 5.1 or 7.1 soundtrack. The strict answer to that is no, we can't recreate a binaural recording as such. What we can do, however, is imagine playing back the 7.1 or 5.1 soundtrack over
speakers, and then calculate what the sound would be in the ears of a listener listening to the speaker-playback; then we can recreate
that as a stereo signal that can be fed to headphones.
This is
not a true binaural recording: it is using headphones to simulate speakers rather than producing a recording that is explicitly designed for headphones in the first place. It will therefore never sound quite as good as actually playing the original 5.1 or 7.1 recording over speakers would (where a true binaural recording could sound much better).
However, it can get pretty close to listening via speakers. And this is precisely what a system like Dolby Headphone does. In theory, I suppose, one could set up a real listening room, play back the movie soundtrack over speakers, and record it binaurally (which still wouldn't make it a real binaural recording) but a more practical way to do it is to reproduce that process mathematically: feed in the original digital audio values for a 5.1-channel recording at one end, and then mathematically calculate (as a new, stereo digital signal) what that 5.1 track sounds like inside a listener's ears.
(This is quite a complex calculation; it involves not only processing direct sound coming from the "speakers" but also reflections from the walls of the imaginary listening room, and all of those being modified by the sound-waves diffracting around a person's head and ears, and making the bones of his skull vibrate.)
Now, there are systems which can do Dolby Headphone processing in real time. In principle there is also no reason why you shouldn't pre-calculate a Dolby Headphone track and record it onto a DVD as a stereo track. There are, in fact, a few DVDs that actually do this - but, sadly only a very few.
What this thread was originally about is a system called "Smyth Virtual Surround". In a way it does the same job as Dolby Headphone, but it does it in a much more clever way.
Instead of using a standardised theoretical room combined with a standardised theoretical set of listener's head and ears (which are never going to be quite the same as the real listener's ears are) it directly
measures the effect of the listening room, and the speaker system, and of the person's head and ears, and then
customises the calculation specifically for that combination (even correcting for the frequency response of the headphones he is using).
SVS also has a head-tracking system, so that when you turn your head from side to side during playback the perceived position of a sound changes in exactly the way that it would if it were coming from your speakers.
The effect is apparently so convincing that you genuinely
can't tell whether you are listening to speakers or to headphones simulating speakers - there is no audible difference.
The downside of the system is that it cannot work without being calibrated to the
combination of an individual person's ears and a specific listening room and sound system. That means you have to have a surround-sound setup running to calibrate it, and the system will accurately reproduce every single acoustic flaw that the original sound system and listening room have. It's impossible to supply a "standardised" ideal room, because you can't separate the influence of the room and the influence of the person's ears. You could offer a standard room+system+ears profile, but that (head-tracking aside) couldn't sound any better than Dolby Headphone does, and would render the system pointless.