This started as an observation of my own, but over time other sound designers and re-recording mixers have also commented on this. This really started when Dolby Digital was introduced to theaters back in 1992. I know a lot of us back then were creating mixing that really tested Dolby Digital technical parimeters, but we are way past that now and should be creating mixes with a lot more subtlty.
One of the worst examples of sound design and mixing I have heard recently is Battle Los Angeles. It is unlistenable. Five minutes into the movie, and you are greeting with various loud gunfire and explosion that last the entire movie. I was battle fatigued after the first half hour.
I just wanted to say great thread! Barring all the emotional stuff, there is a lot of great audio talk to chew on. Soundtrackmxr seems to be the most in tune with my ideals in a sense. Trying to artificially create localization isn't easy. Seems like the most similar speaker configuration that would replicate what the sound designer is attempting to convey is just common sense.
That said, for music we are largely confined to a 2 channel world. Binaural and ambiophonics may be the best we can do. Utilizing ambiophonics on my home system does create some tonal issues. Running white noise through an impulse reponse of an ambiophonic system, then a matching EQ, and then reversing the curve by 100% might be a decent solution for those with a DAW and the ability to make their own impulses. Someone should try it I think.
Thanks for the audio education,
Ron
That said, for music we are largely confined to a 2 channel world. Binaural and ambiophonics may be the best we can do. Utilizing ambiophonics on my home system does create some tonal issues. Running white noise through an impulse reponse of an ambiophonic system, then a matching EQ, and then reversing the curve by 100% might be a decent solution for those with a DAW and the ability to make their own impulses. Someone should try it I think.
Thanks for the audio education,
Ron
Any equalization is probably best done 'to taste'. The overall frequency response is a mixture of direct and reflected sound, and they each have a different spectrum governed mainly by the choice of speaker geometry and the degree of cancellation. There is not much you can usefully measure without a dummy head, and even then there are choices to be made of what to measure.
Controlling the direct and reflected sound somewhat independently is probably the best way forward. This means additional speakers. Dipoles fed with the raw stereo, with the nulls aimed at the listener is one way to do this. Ralph goes a bit further...
Controlling the direct and reflected sound somewhat independently is probably the best way forward. This means additional speakers. Dipoles fed with the raw stereo, with the nulls aimed at the listener is one way to do this. Ralph goes a bit further...
Last edited:
Another way to play with the direct and reflected magnitude and spectra is to use distance and direction to the speakers. If the 'Ambiopole' is brought as close as practical, it dominates the direct sound. If the additional speakers are placed far away they have less affect on the direct sound and proportionally more on the reflected sound.
Although I am not allowed to use the word "diffuse field" on the grounds that there is no such thing as "a bit pregnant", the refections in the room redistribute both the recorded direct and reflected sound in a 'plurality' (patent speak!) of directions, altering, by means of listener's HRTFs, the interaural difference spectra. This can then be manipulated by altering the relative magnitudes and directions of the various components - direct-direct, direct-reflected, reflected-direct, reflected-reflected. Does it matter? The answer is to play and find out..
True, but interpretation of dummy head impulse responses are beyond the skills of many. Who even has a dummy head in his home ?
I do.. I have no clue of why I'd want to measure a room with one. All it would seem to do is make things more difficult vs. just using an omni mic.
Thanks David, I screwed up my thinking yesterday when I wrote that. I'm blaming it on Belgian beer. What I meant was 'make an impulse of your room in normal stereo, then one in ambio and make them match the stereo.'. It's still not perfect but it should fix the to al errors(I think).
. I have no clue of why I'd want to measure a room with one. All it would seem to do is make things more difficult vs. just using an omni mic.Thanks David, I screwed up my thinking yesterday when I wrote that. I'm blaming it on Belgian beer. What I meant was 'make an impulse of your room in normal stereo, then one in ambio and make them match the stereo.'. It's still not perfect but it should fix the to al errors(I think).
True, but interpretation of dummy head impulse responses are beyond the skills of many. Who even has a dummy head in his home ?
Ahem... blush... I thought everybody did..
I do.
Thanks David, I screwed up my thinking yesterday when I wrote that. I'm blaming it on Belgian beer. What I meant was 'make an impulse of your room in normal stereo, then one in ambio and make them match the stereo.'. It's still not perfect but it should fix the to al errors(I think).
The problem is that you have no idea what the spectrum of the direct sound at the ears is, unless the epsilateral (main) and contralateral (canceling) waveforms are superimposed _after_ shadowing by a head. A pinnaless head or sphere would give a simpler picture than an actual head. With just two speakers, correcting the reflected sound will just bugger-up the direct sound. I aim for a flat mono, direct signal at the ears, which leaves a muddy sound if you hear too much room sound. The mono speaker signals can be seen just by adding the two (near and far ears ) IRs together. The peak I get in the speaker drive signals corrects for the notch you would otherwise get due to the interference of the two arrivals at each ear.
My intial impressions are the hd219 are great for binaural, budget or not.
It seems the DTS guys knew what they were doing there, as I suspected.
They have foam between the driver and ears (to cut down relfections?) which is one thing I want, and don't have to add.
I have a bust of Beethoven, can I put mics in his ears? He can't hear anyway.
Or would this give everything a Teutonic sound?
I understand what differences you would see David, I just don't see how it benefits an enthusiast vs. just a regular omni mic measurement. You can know what's going on just as well with 1 mic and no head. There's not much you can do about head shadow issues if you'd even think of them as issues. You were born with those issues. Yes I know you weren't born listening to two sources making the same noise as in stereo. Precisely why I like the idea of SS.
What is "SS"?
I may misunderstand what you are trying to measure. It might well be a useful measurement. If the sounds you are interested in have random phase, then an omni mic anywhere is probably useful. The direct, signals doing the cancelling need to be coherent, and with precise amplitude and phase relationship, otherwise both the cancellation and frequency response ' goes to pot'..
For a mono, central source the cancellation drops out during the processing so that both the speaker feeds and signals at the ears should be (more or less ) flat and 'in' phase with little problem for listening or measurement. The problem is worse for signals panned hard to the side, or stereo reverb, where none of the signals you can easily get your hands on are are anything like flat. The speaker signals are then 'out' of phase, larger and with small amplitude and timing differences. It is not easy to undestand until you try making your own filter. You then get faced by the problem that as the frequency goes lower, the head shadow effect goes to zero and the required speaker drive levels become infinite. It is then time to do some head scratching...
Surround sound.
It's too easy to gate a measurement to see what any one speaker is doing and we also know how that correlates with tone.
Measuring the quality of a sound image from 2 speakers isn't anything I've seen solid research on. I mean I know of no way to measure it in a quantified fashion. I'd love to read it if it's out there. We know a lot of contralateral reflections make for a sense of spaciousness and that early reflections can effect tone, envelopement, and source broadening. I've just never seen anything that says exactly what creates an optimal set of compromises for all.
It's also not hard to find the acoustic center using white noise and watching the old one toothed comb disappear.
It's too easy to gate a measurement to see what any one speaker is doing and we also know how that correlates with tone.
Measuring the quality of a sound image from 2 speakers isn't anything I've seen solid research on. I mean I know of no way to measure it in a quantified fashion. I'd love to read it if it's out there. We know a lot of contralateral reflections make for a sense of spaciousness and that early reflections can effect tone, envelopement, and source broadening. I've just never seen anything that says exactly what creates an optimal set of compromises for all.
It's also not hard to find the acoustic center using white noise and watching the old one toothed comb disappear.
True, but interpretation of dummy head impulse responses are beyond the skills of many. Who even has a dummy head in his home ?
Not only do I have a dummy head....I also possess castings of the finest ears in England!
Not only do I have a dummy head....I also possess castings of the finest ears in England!
England?
England, Wales......somewhere around there.
You expect a redneck from Florida to know geography?!
Lucky you have a sense of humour.. Enjoy!
Australian Newscaster asks Americans Which Country to Invade Next (language NSFW)