castleofargh
Sound Science Forum Moderator
- Joined
- Jul 2, 2011
- Posts
- 10,443
- Likes
- 6,063
Sure, listening situations vary a lot, but if it ever looks like 2 out of phase side by side frontal sound sources outputting the same square wave signal, call me.A 'normal listening situation' varies a rather lot. For me it's a CIEM that sits a few mm away from my ear drum. As the paper points out, broader firing speakers smear the timing far more than the audible limit... that's not a limit of the human ear, it's a limit of your rig!
I included that link as it shows some pretty crazy time audibility and speculated that might come into play in explaining the EEG results so consistently... Far more substance than the rebuttals so far! If you take the time to read the post this is pretty clear. Now, what exactly have I misunderstood?
And no, I don't really care about allowable phase variance on a 20k sinusoid, it doesn't relate to max audibility in humans, nor does it somehow circumvent Shanon's law. If you want to talk about 'misuse' that one is WAY further off topic lol.
There is no point in introducing yet another situation that isn't remotely realistic, to use data from that as an argument for everything else. It's just not good practice and is probably not at all conclusive in the way you implied it to be.
The more important point or as important IMO is that you shouldn't just casually associate numbers cherry-picked wherever from whatever experiment, on the basis that they show the same unit. Otherwise, I could show that when light particles hit me at nearly 300000 km/s, I remain mostly fine with that, therefore I have no reason to worry about a truck hitting me at a mere 50km/h.
What you did with that paper is not as dramatic, but it's similarly grounded in bad logic. We happen to have another very similar false case for this, which is MQA using a paper about interaural just noticeable delays(a test made to get the very best possible result out of it, instead of realistic music listening, but still factual for that particular testing). They took that time value and, like you, misused it to argue a need for more than CD's sample rate. Which was wrong because outside interaural discrimination (which is used to determine the horizontal provenance of a sound), humans aren't nearly as discerning and the JND phase then goes way up. But it also was wrong because it was used like you with a formula that does not in fact define the time resolution of a 16/44.1kHz file or any PCM resolutions.
Again, just picking numbers and giving them a new life because they happen to share a common unit is not serious work. Can I notice 7Hz on my headphone? Not at my typical listening level, at least. Can I notice the ground shaking at 7Hz and moving everything by half a meter or more? Yeah, I sure can. Let's try to do better.
Your presentation of the human ear is IMO a case for demonstrating that the ultrasonic content, if of influence, is unlikely to be of hearing influence. The audible high frequency range is already quite packed at the entrance, having to share a more limited number of hair cells(maybe why our sensitivity at high freqs is lower from the get go? I'm not entirely sure). They're also the hair cells that get damaged the most as we age or experience damaging levels of sounds (being at the entrance, and being for similar amplitude, resonating with higher energy, bad engineering if you ask me^_^, for what some think to be quite important).
I couldn't find pure tone audiograms by age going above 8kHz. Standard medical tests are logically more concerned about speech than anything else, and it doesn't require an expensive rig. But nowadays, I know it exists and is being used in some research. I think we nonetheless can confidently infer that it's not going well for high and ultrasonic frequencies as we age. Which again is completely in line with the physical model of an ear and hearing sensitivity in general.
There are other papers testing ultrasounds for skin sensation and one for eye wobbling and probably more. One of those might be a better cause for brain trigger due to sensory inputs. Because the ear has to start with some serious handicaps in that area, then age comes to close the coffin on that hypothesis.
The paper, about unconscious blablah, might be correct about the Alpha wave readings and it being caused by the difference in ultrasonic frequencies. But pretty much anything after that is opportunistic guessing turned into a narrative in favor of ultrasounds. I don't remember the details, but when this came out and was discussed at length here, I did go and read the reference papers and then some more on brain waves and interpretations(some stuff is near incredible BTW). About linking some emotions or focus levels with alpha waves in the brain, my takeaway was that it's way more complicated than the ultrasound paper even vaguely and cautiously made it to be. Caution, you didn't exactly keep when saying something like:
We enjoy it so much more that no conscious testing manages to validate that we do in the very same research. So much joy and relaxation that it does not manifest in any testable way... It's the NFT of relaxation, worthless. Tentative correlation is the most the paper does with that.There exists objective data that not only humans not only hear the difference in music sampled above 44 kHz, and that we enjoy it more too!
I have to be fair and admit that I'm picking on some of the things that trouble me most, but in your long post, I agree with and find reasonable a lot of what you wrote. I don't try to mean that you're only talking total nonsense, but you do take liberties with the data that you probably shouldn't. As mentioned, there are a bunch of other papers that aren't nearly as optimistic or opportunistic(gotta please the funding overlords and get results no matter what). I could take similar liberties as you do, go pick up that old research between SACD, CD and MP3 where MP3 was slightly preferred in blind testing (not statistically important of course), and make up a story to support the benefits of MP3. The moment we let our own desire influence our interpretation of data, we're no longer doing science. Seeking confirmation and finding it in the vast population of research isn't that hard. But are we ever correct with such an approach?
TBH, I never got convinced by any of the rationals around higher frequency content for better timing. The minimum delays we pick up are, in accepted hearing models, not based on actual discrimination at those speeds, but on analysis of events that allow inference of such timings by the brain.
For example, our hair cells aren't able to trigger as fast as the highest frequencies we can hear. For lower frequencies there is a direct relation between triggering and frequency, on top of tonotopic activation of cells validating a specific area of resonance that itself meant a certain frequency for the signal. But for higher freqs, it's only the location that resonates most with a signal that ends up interpreted as a higher tone the actual signal isn't sent to neurons in any way other than magnitude of stimulation in a given area.
Another example would be the interaural delay discrimination, we again rapidly run out of recycling speed for the cells and neurons and the tiny delays we notice are more likely about getting a certain signal pattern that the brain can somehow analyze to figure out from peak amplitude, etc, that each ear is getting a somewhat similar signal with a delay. It has no relation AFAIK, with higher frequency content allowing a smaller timing.
For ultrasounds and ideas of transient stuff, at best I can agree that at a given time it would provide higher total energy to also have ultrasounds. I agree because that's physics, not opinions or hearing. But is that energy detected? Apparently not, given how blind tests don't validate such an idea. Could it still be enough of a difference for some more neurons to get triggered? IDK. There is a lot going on that ends up nowhere. The brain, again as far as I know based on modern models, receives way more data than it is capable of handling, and the accepted idea is that the brain effectively uses a bunch of tricks to discard most data and focus on dealing with what it deems relevant. There is a thread out there in the section about some possible local cleverness for neurons where some signal supposed to make it trigger its own impulse, doesn't, based on some rules theorized to be linked to number of occurrences and what not(super deep stuff). I don't know how much of such activity reaches the brain to still end up dead in the water on its way to some final area or conscience or automated reactions, or whatever subconscious concept we use (not very popular in the last decade apparently). But I wonder if maybe that kind of ultimately rejected data might also create visible readings on brain wave patterns? If that's possible, then it's even more work to try and make actual sense of what's measured. I've seen 2 rather long videos on youtube(World Science Festival probably, at least for one of them), on various people involved with decrypting the brain(one has a fairly solid system to help paralyzed people talk, the vocabulary isn't amazeballz, but it's something). Anyway, that's a domain of study where a year does make a difference, so while I seem to spit on the paper you linked, I'm actually very curious and optimistic about similar lines of work(hopefully for something a little more important than "why should we care about ultrasounds in music?").
I think there is enough research to agree that ultrasounds can be detected in some ways, but arguing that it has a positive impact on music is more heavily reliant on wishful thinking than it is on actual research, and we kind of all know why if we're honest about it. Marketing+audiophile philosophy+personal desire to be special and notice stuff that most don't+non-test that audiophile call critical listening tests. Those I think are more than enough to explain why people grasp at any paper straw if it means hires is bettererer in any, and I mean any, way. The monetary incentive on research is about 100% skewed toward trying to find a difference, and a beneficial one at that, because a giant industry is behind it, while 3 guys with a beer belly are behind saying that as far as we know, 44 or 48kHz handles everything it needs to handle.
And yet, we have so little to show in favor of hires. That alone is almost enough to convince me that it's all BS if I'm being honest.
It's a weird thing really where being able to do something becomes a need to do it for reasons TBD. We don't have giant hobbyist groups pushing for TVs with the spectrum for birds or mantis shrimps, we'd like even better dynamic range and some color accuracy because we basically all clearly and demonstrably notice when it's improved, but there is no push for UV wide TVs. That would be stupid and dangerous, but you get the idea, better timing of the picture, better whatever phase discrimination for important stuff we'll invent as the marketing guys get creative with it.