[quote title=Eric Bridenbaker wrote on Fri, 30 September 2005 02:00][quote title=Paul Frindle wrote on Thu, 29 September 2005 20:26]
Timeline wrote on Thu, 29 September 2005 19:40 |
A crash symbol can be measured to 27khz.
Hey Paul,
This is definitely something think about, it all makes sense! Hadn't considered any of this.
Question: If captured ultrasonics are subject to double intermodulation upon playback, would the same process also apply to captured frequencies under 20KHz?
In other words, are we hearing this double IM effect anyways, even in a recording that is bandlimited 20Hz-20KHz? Wouldn't for instance 10KHz also be subject to the same rules as 25Khz?
Does it make sense that intermodulation in a non-linear system will occur for all frequencies, whether in the audible range or ultrasonic?
Still trying to get a deeper understanding. Hopefully I'm wording this OK...
Best Regards, Eric
|
Yes very definitely. Intermod products will appear for signals within our hearing range if loud enough, both in the air and in the ear itself - although normally we tend to be unaware of the ear's problems IME, unless specific tests are done that make us concentrate on them.
The point would be that for the situation I described above the intermodulation caused by the air is really at the initial boundary between the instrument and the air because the sound pressure level drops dramatically with distance. So if you are listening from a meter or two away you are hearing mostly just one level of intermodulation. SO the subharmonic beating of supersonic HF you describe would occur mostly from this initial boundary.
But if you put a mic in front of the instrument and pass this at full supersonic bandwidth out of a speaker at similar levels you will have two such boundaries, the original instrument to air and the speaker to air as well. And this is assuming that both mic and speaker are hypothetically perfect.
BTW on another point - I can find no evidence that the ear itself is capable of creating subharmonics from supersonic HF. Once the signal gets beyond our hearing range it has truly gone as far as I can tell in tests I have done. This was something I didn't expect at all since beforehand I assumed that the mechanical bits of the ear would at least have some sensitivity to the HF and some non-linearity! But it seems not to be so for some reason.
The tests I did were of course inspired because our company was involved in designing high bandwidth reproduction systems - and there was a fair degree of pressure to find 'something' that improved with higher B/W. Sadly to date I have found none - if the most faithful reproduction of the original sound is what you are after