Recording with the SR3D Binaural Pro
Vložit
- čas přidán 31. 07. 2023
- / michaelkrzyzaniak -- Please support me on Patreon.
/ @sr3d-microphones -- Check out SR3D's channel
In this video I go around and record a bunch of stuff with a binaural microphone.
Sorry for recording this voiceover in the middle of a river I thought it would sound cool but it turns out its just annoying to listen to. I'll record the next one indoors I promise 😇
The equivalent (Neumann KU100) is almost €8,000. Great job.😮
There are two modes of acquiring directional cues. One is quick, which is from frequency response and left-right balance. The second method is slow. Slow directional cues are not about phase or magnitude but about a difference between the magnitude captured in two positions of your head. There is a neutral position of your head and there is a special position only used for sound location. Neutral position is when you keep your head up and look straight ahead. The special position is when you tilt your head down and to the side. Every time you want to precisely position something by ear, you move your head to this position. The act of acquiring the directional accuracy is from switching between these two positions and comparing the tonal balance. That's how people tell if something is in front of them or behind them, above or below.
This is basically the argument for ambisonic recording, since you can in principle capture an entire spherical panorama of sound, and later you can tilt your head and listen in different directions. (It dosen't work very well.) I was trying to figure out a way to do this with the binarual mic: something like putting it on a tripod and spinning it around in different orientations, similar to capturing a panoramic photo with your phone. I have some ideas about how to do that, but of course it is too complicated and I am lazy.
Pinna, like eyes, can be "in focus" and "out of focus", the pinna shape are all slightly different, but some more so than others, and as such can "de-focus" the perspective of sound due to the deformation of the pinna (perhaps unique tones can be repeated causing confusion with a deformed pinna).
Unless you take a cast of your own ears, you can only hypothesise about the effects of the pinna, but cloning your own pinna, you can experiment with various angles, and when you do this you will realise that there is a fine tolerance of "focus" and "de-focus", however, using someone else's ear, or, an artist's impression of a pinna is a total waste of time as the perspectives would be confusion, unless of course you spent several weeks with the new pinna attached to your head/headphones, so that the new binaural infused information was learned (this can be learned just as wearing special glasses to make everything upside down, after a period of time, the brain will make the upside down image the right way, the downside is, when the glasses are removed, the learning process is needed again).
There is quite a bit of dogma around binaural in my opinion, and I believe it's simply due to standard papers and experimenters using the wrong pinna to come up with their results, it is IMPERATIVE that any tests done by, would be experimenters, to use their own pinna, or they will never comprehend the importance of the pinna, the head really has no importance, it's simply where we have our brains, eye's, mouth, and is a perfect location also for the pinna.
I have been working with binaural audio since 2013 when I realised how it works due to the pinna infusing the tonal differences before the sound enters the ear canal, if I had already been well-read on the subject of HRTF I would no doubt have been brainwashed into a way of thinking that I would never have realised the function of the pinna.
With a set of your own pinna, you can try extending the distance between the pinna, and other angles to your heart's desire, only to find that the distance between the ears isn't really that important (to a certain degree) as from the other angles that do have a huge impact of your perception of direction. Personally, my own ears, I can tell with excellent results the perception of direction, which make me presume that my pinna are not deformed and are quite a good example to be experimenting with, if on the other hand they were deformed, then no amount of experimentation would have me realising what the pinnae function was, and I would never have developed the SR3D binaural capturing device at all.
You seem to be regurgitating scientific papers from people who have studied the HRTF for many years in your comments, from no doubt scientific people that tested without their own pinna, so are actually hypothesised papers in my opinion.
There are papers out there with correct pinna related transfer function and very few videos describing the function correctly, as the majority of information out there is actually misinformation.
I find it an incredible subject, as when the function of the pinna is realised, you can then *see* how the pinna actually functions as a mechanical organ due to the folds and parabola creating unique tones - it sticks out like a sore ear! Without the pinna any perceived sounds would be heard inside the head, the pinna makes it so that we perceive the sound externally. Without the binaural information, the stereo image simply becomes stereo.
I have asked ChatGPT to use my written notes that I have used in previous replies on the pinna, I asked it to condense all of my notes into one "paper", it wrote the first one as this (I wonder if you can follow it?):
The human pinna, or outer ear, plays a crucial role in binaural audio perception. Unlike the common misconception that it serves as a collector of sounds, the pinna is primarily a binaural infusion device. Its intricate structure, distinct ridges, and parabolic shape work in harmony to induce unique tones based on the direction of sound. This binaural information allows us to perceive the spatial orientation of sounds - whether above, below, in front, or behind us.
Contrary to widespread belief, the pinna's purpose is not to collect sounds but to modify them before they enter the ear canal. If the pinna were designed for sound collection, it might resemble the ears of animals like cats or horses. However, its unique shape enables the generation of differential tones based on the direction of sound, facilitating the brain's ability to determine the source of the sound.
The pinna is distinctively visible, providing a tangible way to understand its function. It stands as the only organ in the human body where the mechanism of sound modification is observable. Without the pinna, binaural information would be absent, resulting in an internalized perception of sounds within the head.
In the context of binaural audio, replicating human hearing involves creating a microphone that mimics the individual's pinna. By recording and playing back through headphones, the left and right channels remain separate, enabling the perception of sounds above, below, in front, and behind. Casts of one's own ears are recommended for accurate recordings, as the unique shape of each individual's pinna affects the tonal qualities of the induced binaural information.
While some literature incorrectly describes the pinna as a sound-collecting device, understanding its true function enhances the comprehension of binaural audio. Experimentation with one's own pinna through accurate recordings can provide a first hand insight into the intricacies of binaural perception. It is emphasized that using someone else's ears or artistic models for experiments may yield inaccurate results.
In summary, the pinna serves as a vital component in binaural audio, infusing directional information into the ear canal and allowing us to perceive sounds externally with spatial orientation. It is a unique organ with observable mechanics, often misunderstood in conventional descriptions that attribute sound collection as its primary function.
I have one question. I haven't done much binaural myself, but I've studied headphones quite a lot. It seems like none of the guys capturing binaural ever use compensation EQ in their files. They all sound more nasal than stereo recording and I'm not sure why this isn't considered coloration by binaural people. When headphones reproduce these recordings, they play tham into your own pinnae and ear canal, so you get your own directional finger print from that. Because the recording was already done with some ears, you get this coloration twice. This is 16db of unnatural ear gain! A bit too much.
I think the assumption is that the headphones play sound directly into your basilar membrane and your pinnae and ear canal don't transform the sound very much. I have no idea how good this assumption is. I hope some of the binaural purists can come and weigh in on this.
A good set of headphones would be required for the correct tonal colouration, as you put it.
I was quite lucky using the Sennheiser HD280 Pro's, however, I recently tried a popular Sony brand that's been out for decades (Sony MDR-7506) and they were very "toppy" and the colouration was completely wrong for binaural, they may very well be perfect for editing as they are used in the trade, but, for a natural sounding "binaural tone" they are completely useless in my opinion.
Assuming and presuming will not give you a guess that is accurate, you have to *know* and realise what it is you are dealing with to offer a good opinion that stands it's ground.
If you use speakers rather than headphones, you can use your own pinna to advantage, as it will infuse somewhat other binaural information on top of the existing binaural information so that instead of a sound moving *with* your head, it can (to a certain degree) stay in position as you move your head, giving a more realistic perception of being there, though the positions of the speakers do have to be within a certain position, there's a certain tolerance that gives a good response, and then it falls apart, neck speakers are quite good too, but not so much for turning your head that much, though I need to experiment more with them.
@@sr3d-microphones This is even more interesting because the Sonys that you mentioned are basically binaural headphones with bumped bass. Apart from the bass, their response was shaped after a so called diffuse field equalisation target, which is a non-directional tuning that is supposed to localise a recording in the middle of your head. It's stripped of it's own directional cues the same way that a transparent amplifier is stripped of it's distortions. What you were hearing is exactly the problem that I'm pointing out. HD280pro lessen that effect but still it's there. You'd need something that deducts the coloration in the recording from the diffused field target to have headphones that perform well with this kind of recording.
@@rhalfik
Your term "binaural headphones" is a little confusing, as binaural, from what I understand from much research, are the implicit tones that the pinna infuse to the sound before it goes through the ear canal, so "binaural headphones" are not something that I understand, unless they actually infuse "binaural" information onto the sound that is being played back, and I can't really see that happening, I think perhaps the use of the term "binaural headphones" is actually a misplaced understanding, either from the manufacturers, or their end users. Many manufacturers use the incorrect term "binaural" on many microphones, which are simply stereo microphones, for example.
You state, "Apart from the bass, their response was shaped after a so called diffuse field equalisation target, which is a non-directional tuning that is supposed to localise a recording in the middle of your head."
If any sounds are perceived "inside" the head, the sound that is being listened to is not binaural, but either mono, or stereo.
Can you please explain, "It's stripped of it's own directional cues the same way that a transparent amplifier is stripped of it's distortions." As I can't seem to get my head round that.
My understanding is that the binaural infused tones that give us the external perceptions are indeed played back in recordings from binaural capturing devices, and the higher the bitrate, the better this seems to be, though the Sony's I mentioned were simply to toppy, whereas the HD280's seem to have a flatter response, and am sure there are many other headphones that could be better or even worse for true binaural capture playback, but the tern binaural is most definitely misused as a common misunderstood term for things which are used but not binaural at all.
@@sr3d-microphones You're saying a lot of interesting stuff that I haven't processed yet, so I'll just explain what I meant by "binaural headphones". Headphone tuning is complex and it depends on their purpose. A popular tuning for example is Harman OE2018, which is based on a stereo set of speakers and it's meant to give you the most similar listening experience to a pair of speakers in a living room. It's the perfect tuning for listening to music and other stereo recordings because stereo is primarily a speaker tech. Headphones were adopted for stereo recordings later and were unpopular so the industry forced this tech to adapt to whatever was already made for speakers. It's the right way to tune headphones if your recordings are meant to work well on both headphones and speakers.
However this tuning contains certain features that can give you directional cues other than what's recorded. Usually people hear the sound being more forward, at their forhead for example. By tuning I mean measured frequency response. For binaural listening, it's a good idea to use a tuning that is neutral from HRTF perspective, meaning that it positions the sound in the middle of your head. A theoretical version of this tuning is called a diffused field target. It's a directionless, blank canvas, on which you can lay the binaural recording. It's a bad tuning for leisure because it has a lot of highs and not much lows. Sony basically took this target and added bass, which your body transmits through bones anyway, so they simply adjusted it to taste. The problem with it is that it's very bright, especially around clarity region, which gives a lot of power to S sounds and other details. I think the first pair of diffused field tuned headphones were AKG K240 DF, which were used in broadcast in the 70s and 80s. They didn't have this bas bump, so they sounded super thin and were pretty much unlistenable for stereo recordings. I'd call them "binaural" though. The most renowned DF headphones were Sennheiser HD800, which also were quite unpopular among audiophiles so they got replaced with HD800S, which has tamed highs, so it's less DF and more HOE2018. These are all the headphones that are meant to reproduce a neutral, directionless sound in such a way that it has no direction. However if the recording contains some directional cues, they'll be reproduced accurately. It's like having flat frequency response, but since in headphone world there is no flat, its the closest equivalent with this purpose. At least that's the theory, but I don't see it working out in practice. As you noticed, Sonys tend to be to bright on these recordings. So in the end I don't understand who made a mistake and where. Who should adapt, headphone people or microphone people? :)