My idea was to binaural record various speaker systems in the same room (my special designed listening room) so as to be able to make comparisons between them, especially blind. I tried the Head Acoustics system, which is highly regarded. The problem, to me, was that playback was simply not believable. I could compare the recordings to the actual right in the same room. Things just didn't gel. I tried different headphones and all of them were different but none of them convincing.
Ford had done this years ago in cars and got some decent results, but I was not involved in that project and so I don't know what it took for them to get it to work. They used Stax headphones on playback and the headphones were equalized per Head - that was not possible for me.
I think that binaural may work pretty well in a larger space recording original music, but recording a recording and getting it right is a lot harder.
Toole mentioned to me once that they (Harman) had got is to work but marketing would not let them use it outside of the company for fear that it might get out-of-hand. If this could ever be done reasonably and efficiently then people could evaluate speakers over the internet and not have to go out to hear them at some poor show that is likely to be a waste of time anyways. One could also do real blind tests of loudspeakers, something that only the big guys (like Harman) can do today. Great idea, I just couldn't get it to work.
Similar experience. I was involved in some work with both a Head Acoustics and the original Neumann dummy heads with Stax and Sennheiser HD600 headphones. Some people had HRTFs that were close enough and locked onto binaural sound but some did not. It was concluded that personal HRTFs were needed for any serious auralization work.
Ditto for me. I've tried very hard to record my own HRTF and use it with headphones. Very limited results.
The guys I know here on the forum who have tried the Smyth Realiser say it works amazingly well. But they also say that switching head profiles to someone else's sounded really weird.
The recordings on the site Mitch linked to were very nice, but didn't go much out of headphones for me. I would really love to hear that. The DTS demo I heard in Vegas was pretty good.
The guys I know here on the forum who have tried the Smyth Realiser say it works amazingly well. But they also say that switching head profiles to someone else's sounded really weird.
The recordings on the site Mitch linked to were very nice, but didn't go much out of headphones for me. I would really love to hear that. The DTS demo I heard in Vegas was pretty good.
I have done this too. I used the binaural mics linked to earlier and set up at the listening position and made a recording. Then I took the recording and the original and loaded them up into Audacity and lined up the start of the two tracks, down to the sample and then alternated between them every 10 seconds. This really reveals the tonal balance and the rooms contribution when alternating back and forth. If I still have it, I will see if I can upload it. However, to be fair, I really need to calibrate the microphones... In fact, I ended putting these binaural mics into Jeff's binaural ears and got better results.
Jeff, the maker of 3DIO claims that, "that the ear pinnae design and physical spacing provide 98% of the directional information your brain needs. The Free Space ears give you this information, and the discs give you the proper head-shadowing effect. Alternatively, a full dummy head with shoulders and/or torso will add very little, if any noticeable benefit to a binaural recording" And .." based on subjective listening tests, the Free Space has proven to provide a more open and "seamless" localization across the front, when compared to a full dummy head microphone."
From: FAQ : 3Dio, Free Space Binaural Microphone - Binaural mic for binaural recording
I have used the Neumann dummy head in the studio for a number of recordings and the recordings from Jeff's mics (his Free Space ears, but a different set of mics linked above) seems to translate the binaural effect much clearer for me, regardless of the headphones I tried them on. Comparing the "ears" between the two, there certainly noticeable differences in the shapes. Jeff's "mold" seems to be more detailed then the Neumann's. However, that's my subjective opinion, but I would rather have a repeatable and measurable test.
Yes, Ditto. tried HATS, Kemar and HEAD Acoustics, same conclusion. OTOH, heard a demo with a one of the torsos, played back over speakers with crosstalk cancellation, and while music tonal balance was poor, percussive sounds were incredibly realistic in their image placement. They could place a "kiss" behind my ear (coming from frontal speakers)
Last edited:
The technique can be conceptualized as applying weighting functions to the 'room curve". The test outcome will give a better idea of the sound presented to the ERP by the room, but will be so room dependent as to be nearly impossible to apply guidelines against. Interesting idea and I bet with time and $$$ it could be figured out, but it would be a huge challenge.
Interesting discussion. It is said that convincing externalization of sound sources requires the incorporation of head movements. Without it, you rely solely on imperfect HRTF's. Additionally, you use small head movements in order to separate directional sounds (direct sound and strong discrete reflections) and reverb. Not incorporating head movements into the playback system leads you to experience more reverb than you would in real life.
Just to be clear, I was commenting on recording a pair of loudspeakers in a room with playback over headphones that closely modeled the real situation. That was not happening. But binaural playback of a live performance can be quite good in general. I am not sure why the difference, but it was considerable.
I have always had a problem with this position. I mean lets look at the converse. If I hold my head perfectly still, then the sound should be different - it never is. How can that be if head movements are "required".
I hold my head perfectly still much of the time when listening, never is it any different than moving my head.
Agree. Head movement helps in localization, but for me never changes the overall tonal balance.
I've heard that for best results with binaural heads you should use molds of your own pinnea , and measure the freq response of your own ear canals and use this inverse eq on playback with headphones. The first bit makes sense but not so sure about the eq. This may be to adjust for the difference between regular distant sources and headphones.
Found it. Great info gathered from many sources.
http://www.linkwitzlab.com/Recording/Binaural_hearing_and_headphones (1).ppt
http://www.linkwitzlab.com/Recording/Binaural_hearing_and_headphones (1).ppt
The problem with this is that you actually do move your head. You move your head the entire time, until you sit down and try to keep your head as still as possible. But then of course you already know where the speakers are, you already have a mental map of your acoustic surroundings. When listening to a binaural recording, all is different. You not only cannot rely on head movements, you also have no visual cues as to where the sound sources are.
In the telecom world, the cocktail party effect was a big deal, as our devices often need to operate in zero SNR environments. Binaural hearing enables conversations in environments a phone would severely struggle in
The following study looked at the effect of head movement
The Cocktail Party Effect Explained ? Drink Factory
"In short - a quick turn of the head makes us reset our perception of what we're hearing"
That is, some head movement can allow our cognition to reset and reoptimize to a different perceptual solution.
The following study looked at the effect of head movement
The Cocktail Party Effect Explained ? Drink Factory
"In short - a quick turn of the head makes us reset our perception of what we're hearing"
That is, some head movement can allow our cognition to reset and reoptimize to a different perceptual solution.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.