But have you tried it?
The Nigerian gentleman is asking you to commit fraud, that is not nice.
Now what is it going to take for you to put two pictures in the freezer? The possible outcomes are none (EXTREMELY LIKELY) or an improvement in your lifestyle (VERY LOW PROBABILITY.) However the cost of doing so is trivial. So the risk reward ratio is very good. You cannot be positive about the outcome unless you try. Then you also can't prove a negative.
Now doing this seems completely ridiculous, but someone has actually put this out with their name and reputation behind it, so there is some real support there. Yes it seems really weird, but you really will never absolutely know until you try.
Now if you want to give me your bank routing information!
BTW yesterday went out to dinner with my brother and sister in law for their anniversary, I gave them a 12 troy oz. gold bar. Last time it was a bag full of a hundred dollar bills. So you can be comfortable giving me your bank routing number.
(For the non-English speakers a gold bar is not the same thing as a bar of gold!)
Yes of course, doing controlled listening tests is normally not that funny.
Minimum number of trials required at SL=.05 is 5 for onetailed tests (actual p=0.031) and 6 for twotailed tests (actual p=0.031 too).
Therefore we usually try to choose one favorite DUT in a sighted test before and run a controlled test later, a correct answer is given by picking the favorite DUT.
A few months ago you told me that tiny amounts of "PIM" are not masked by the several orders of magnitude more from the loudspeakers. How do we know?
No, have you? I have more important things to do, like trim my toenails.
And how do you know the Nigerian is asking me to commit fraud? Have you visited him and audited his records?
About covers a semester of tuition (another thing to surprise folks from Europe).
I use mono. Did you try to see what phase shifts you could hear? Have you ever tried limiting the bandwidth to see where you can tell the difference?
I don't do double blind tests, either I trust my instrumentation my ears or I get complaints when speech recognition gets too low.
The most interesting complaints I got were when I had replaced most of the low quality sound system in an arena. The staff still used their old source materials. The issue was why did some music sound good and some "tinny?" Taking away their cassette deck for "repair" solved that problem. (It is still waiting for parts!)
Do you have more data on the Alien's tests?
Actually as soon as I get a freezer I will try it! Can't hurt! You trim your toenails in the freezer? Weird! I hope you don't trim them blind, that could be painful.
Actually you are supposed to exorcise evil spirits while trimming your nails and toenails and then burn the clippings. That is because you have exposed area for them to enter! If I can find a cite, I'll post it.
I got the same Nigerian letter, he explains it is ill gotten money.
Now are you going to tell me phase shift and limited bandwidth are in the same category as frozen pictures, Nigerian windfalls or toenail clippings? Then Radio Shack has the stereo for you, or save more and go to Walmart!
Yes I did, I don't recall if it was 17.5 or 18kHz but I had some 24/96 files brickwalled digitally, I heard nothing so for ME this whole issue is a don't care. BTW on an FFT there was considerable energy above the noise floor being removed by the process.
Again making conclusions from artificially generated non-physical test stimuli is fraught with potential error.
Found it-
Kiryu and Ashihara “Detection of Threshold for tones above 22kHz.” – Convention paper 5401 presented at the 110th Convention, May 12-15 2001, Amsterdam.
Dave's .ppt is a hefty download but might amuse and certainly annoy.
www.davidgriesinger.com/intermod.ppt
EDIT jneutron, Dave's .ppt also has some slides on IAD.
Kiryu and Ashihara “Detection of Threshold for tones above 22kHz.” – Convention paper 5401 presented at the 110th Convention, May 12-15 2001, Amsterdam.
Dave's .ppt is a hefty download but might amuse and certainly annoy.
www.davidgriesinger.com/intermod.ppt
EDIT jneutron, Dave's .ppt also has some slides on IAD.
Last edited:
Actually sort of. If I am designing a no holds barred design I want to be sure I exceed any human capabilities. You certainly don't consider 9K too high?
If I am building for myself I want to go a tweak or two beyond what I think I hear just in case the limit is some other part of the chain.
The interesting case is in distributed ceiling loudspeaker systems. The coverage of each loudspeaker narrows with increasing frequency. If the high frequencies are rolled off in the EQ so that you can walk the area and not notice any changes in HF content that is generally more acceptable. However if you use pink noise for adjusting the HF cut it will be much greater than if you use music!
Depends on the conclusion. I just asked if people could hear anything interesting. If folks would have said yes the tone changes in xxx way, that would have been interesting. Since no one said that before looking backstage, no interesting data was generated.
I will post a few other files with the same question.
The phase shift and absolute phase arguments are ancient by audio standards. To meaningfully evaluate the sensitivity you need transducers that are both phase linear and free of polarity effects (2nd order harmonics?) such that a positive going pulse is symmetrical with a negative going pulse. This is much more difficult with typical magnetic transducers since the system is asymmetrical magnetically and mechanically.
I remember reading about a test done by some respected British reviewers or researchers in the early '80s (cannot remember the details) where they did this using a Quad esl and found no sensitivity to phase. They selected the Quad because it is a minimum phase and symmetrical system. There are legitimate issues around the non-linearity of air (you can compress it a lot more than you can expand it) but I don't know of anything suggesting the non-linearity is audible at SPL's we can tolerate.
In short, before audibility of phase shift or amplitude response can be determined the test system should be free of those errors.
A sudden phase shift (as in the test signal) will have sidebands all through the audio spectrum, (the click?). Both waveforms are pretty symmetrical around "0". A more interesting test would be several waveforms based on an assymetrical pulse, like John used for caps. As long as they all have the same spectrum but different phase relationships it would be interesting to see if anyone hears differences, and which transducers are used.
I remember reading about a test done by some respected British reviewers or researchers in the early '80s (cannot remember the details) where they did this using a Quad esl and found no sensitivity to phase. They selected the Quad because it is a minimum phase and symmetrical system. There are legitimate issues around the non-linearity of air (you can compress it a lot more than you can expand it) but I don't know of anything suggesting the non-linearity is audible at SPL's we can tolerate.
In short, before audibility of phase shift or amplitude response can be determined the test system should be free of those errors.
A sudden phase shift (as in the test signal) will have sidebands all through the audio spectrum, (the click?). Both waveforms are pretty symmetrical around "0". A more interesting test would be several waveforms based on an assymetrical pulse, like John used for caps. As long as they all have the same spectrum but different phase relationships it would be interesting to see if anyone hears differences, and which transducers are used.
Then, you initially asked in a wrong way
and me, jumped in after this, in a smart *** way. Sorry for that
Regards
George
I don't think his explanation for the difference tone (f2-f1) as generated by the auditory system cuts wood.
However, his general assertion that zero crossings make hair cells fire, and that it is the binaural time differences between that determine location is right on the mark. Not only that, it is highly pertinent to this discussion. Because it implies that it is possible for the brain to register time differences in low uS regions, without having the sort of corresponding ultrawide bandwith that is suggested by some.
One one count I think this ppt is too optimistic about the brain. Given that the fastest neurons are in the 10 mS region, it is more like a 100Hz computer than a 1Khz device.
The only thing really annoying I find in this ppt is that the test tones don't play.
vac
Then, you initially asked in a wrong way
and me, jumped in after this, in a smart *** way. Sorry for that
Regards
George
George,
No apology needed, a few folks downloaded it and none ventured a guess. You started a very interesting bit.
Now try your ear only on the next sample! You response is always welcome.
ES
- Status
- Not open for further replies.