Good point - now can we move on?
it has something to do with x5, as it also can play hi res files. if you are un-interested it is simpler not to read about it.
Here we go again, X5 thread taken over by Sound Science lurkers
sorry for this, but it is a topic that concerns music, the reason fiio x5 exists, and file formats used with it. why not take part? it is interesting.
Well, I may be able to produce ultrasonic tones – by drawing them by hand. So I may do the experiment myself, but not today. If it only were possible to downsample to a sampling rate of say 16 kHz and do the experiment three octaves lower. But the X5 wouldn't play the file anyway.
Since there are microphones specially developed for hi-res recordings*, I would be perplexed if one of these recordings would use an ultrasonic filter. What would it serve for?
Here's an
article which confirms that (at least some) hi-res recordings contain a considerable amount of ultrasonics.
* http://en-de.sennheiser.com/studio-condenser-microphone-digital-recording-systems-mkh-800-p48
http://en-de.sennheiser.com/condenser-microphone-cardioid-guitar-acoustic-bass-brass-mkh-8040
http://de-de.sennheiser.com/kondensatormikrofon-cardioid-studio-recordings-orchester-mkh-8090
http://www.sanken-mic.com/en/product/freqpola.cfm/3.1000800
http://www.sanken-mic.com/en/product/freqpola.cfm/3.1000400
[just for measuring purposes:] http://www.bksv.com/Products/transducers/acoustic/microphones/microphone-cartridges/4138
okay, the examples you have given are totally correct.
Ultrasonics can be recorded, as this is also done with dolphins.
But no ultrasonics leave laboratories, as they cannot be anyways interpolated by DACs which already have a low filter implemented.
You can actually test this, create a 40Khz or 30 Khz tone, and try to record straight from x5's line out. Or any other DAC for that matter.
You will never be able to get ultrasonics, there are 3 steps that i know of in a recording process that erase ultrasonics.
Also, the article you provided has mostly noise, continous noise in that area, not notes or variable anything. like recording the air, but a very steady air.
Even i need to implement a low pass filter into my project to be able to interpolate samples correctly, it is almost impossible to interpolate samples if they are over 22.5Khz as most interpolation algorithms use processing power, from there on, there is much more power needed for calculations.
Sub harmonics created by them, as i said earlier, if they exist, they are already recorded. this is why interpolation and digital reproduction must eliminate subharmonics like it does with noise, because already recorded harmonics would overlap with new ones, if created, and this would raise THD noise, by quite a considerable margin.
Again, sorry for taking this into x5 thread, but it is my DAC right now, data is about it, and i guess that discussion is not entirely pointless. there are more people interested, and it is usefull to have.
Also, i do not start this on sound science, because over there i would be told that resampling does not make a difference, because it does not exist, where i demonstrated in two occasions, that the differences exist, and are audible.
EDIT::: a final word for this would be that no headphone nor speaker produces ultrasonics, and harmonics tend to appear after the sound is already in air, as i understand it. so unless using headphones or speakers producing the ultrasonics, the ultrasonic harmonics, even of high order should not appear.