LOU GENTILE, EVP Applicant

[CACTUSJACKmankin]

If the voices could be translated to the same words by several people without having a translation suggested to them, then that might rule out auditory pareidolia. The key is blind independant verification.

 

[DevilsAdvocate]

If the voices could be translated to the same words by several people without having a translation suggested to them, then that might rule out auditory pareidolia. The key is blind independant verification.

Not good enough. There has to be a control recording.

At some point in the recording there may be a "hiss" sound. So if 10 out of 10 people asked to listen to the recording and write down all English words that they hear all write down "six", would that proof that there was a voice?

No. It wouldn't mean anything. How do you control for that?

You have to have a control recording that could potentially have the same "hiss" sound and potentially have the same recordings of "six".

The test recording would have to score significantly better than the control recording.

 

[Tirdun]

Well here's my 2 cents worth for a format.

1- Setup requires two shielded sound rooms, one with a microphone, the other with a recording device.
2- The claimant sits in the first room with a judge, who will have control of the microphone, which will broadcast into the adjacent room when activated.
3- The judge will activate the microphone and signal the claimant, who will ask any question s/he wishes. This is played into the room with the recorder and noted for time.
4- The microphone is shut off and the recorder continues for a set period of time.
5- Repeat steps 3+4 a number of times.
6- Judging is done by changes in sound level on the recording that occur after the microphone has been shut off. Any increase of predermined decible level is considered a hit.

I think this would eliminate the option of cheating, decent sound equipment + shielding would remove random noise and the required change in decibles for a hit would eliminate any argument about what was said.

 

[Hellbound]

An interesting addition to that protocol, Tirdun, would be to use three rooms. THe third room would hold a recording device identical to the one in the second, but would not have any sound broadcast into it. Then, one could look for differences between the two recordings, or even have the applicant decide which recorder contains the answers to his questions. All properly blinded, of course.

 

[Hitch]

This is a tough one to test. I assume the "responses" will not be loud clear voice, but something faint. So the test of success will require some judgment. But with recordings, espescially bad ones, it is easy to hear voices and words when your task is to listen for them--I hear a hiss, "I think I heard "six"! Because the tet would require judges, there would need to be a control test to determine whether the test results are significant.

If you had a recording studio available, this wouldn't be much trouble. Otherwise, if the applicant accepts having the questions recorded and then played, then it would not be too difficult.

Whatever the protocol, the applicant cannot be in the recording room when the "responses" are recorded, and the judgment of the results will have to be compared to a control recording to determine statistical variance. This is going to be tricky unless you get a recording studio where you can control communication from one room to another, start and stop recording, and replicate the recording exactly for a control test, and get a number of "judges" to get statistically significant results.

On the page CACTUSJACKmankin linked to, near the bottom, there's a set of recordings:

Once we were all positive we heard the name Matthew (we didn't hear Jennifer until later playback), we all asked the same question again -- "Is Matthew your name?" -- We got back several responses of "No".

Fifth Contact

Fifth Contact: slowed down

Fifth Contact: "No" response

The static they claim is "no" sounds as much like "yes" as "no" to me. That's going to be an inherent problem with judging the results of any test of this.

That whole "judging" bit.

 

[Hellbound]

Exactly, Hitch, which is why I think we need multiple recordings, and the applicant must pick the correct one (i.e.-the one with his answers) from among a group of empty recorders.

 

[Orb]

IMO, the "voices" Lou is hearing is simply an artifact of how his recorder works. Notice that he is pretty specific about the model that he requires.

Apparently, the Panasonic RR-DR60 uses CELP (Code Excited Linear Prediction). CELP is a fairly common standard for speech coding, used by e.g. GSM (and probably other digital cell-phone standards), Speex (the speech codec), and other systems that need to store digital speech efficiently.

While I don't know all the details of how CELP works, it basically does something like this: When the audio is recorded, the input signal is divided into short segments (frames). For each frame, the encoder tries to recreate the sound wave of the frame using a synthesizer. The encoder outputs the set of synthesizer parameters that produce a synthesized sound that best matches the original sound. These synthesizer parameters are what is stored in the memory of the digital recorder. When the recorded speech is played back, the stored synthesizer parameters are used to synthesize sound, which hopefully will be resonably similar to the input.

The advantage of this method is that it makes it possible to reproduce intelligible speech at a very low bitrate. Also, since the coder usually is tuned to human speech, it can actually make the speech come out clearer and filter out background noise. That's why digital cell phones often can work resonably well in noisy environments.

One downside is that while the output might be intelligible, it can sound synthetic, and the voice might sound different from the original. Listen to the clips on the page CACTUSJACK linked. Lou's voice sounds almost like a synthesized voice (which it kind of is).

Also, if you try to record something other than human speech, it will likely be horribly mangled. The encoder will try to find the best match possible, and since it's only equiped to reliably reproduce human speech it's pretty likely that the output will sound a lot like human speech, no matter what the input sounds like.

Googling RR-DR60 and EVP suggests that that specific model is the tool of choice for EVPers. Probably because of it's ability to turn any noise into something vaguely human sounding.

Thanks Scoob for this very enlightening explanation from one who is hopelessly lost when it comes to electronics. I used to think EVP's were just people hearing patterns from random noise, like the way we can see pictures on a fuzzy carpet. Now I understand the device sort of does this by itself! Amazing! Thanks for the education!

 

[strathmeyer]

Well here's my 2 cents worth for a format.

1- Setup requires two shielded sound rooms, one with a microphone, the other with a recording device.
2- The claimant sits in the first room with a judge, who will have control of the microphone, which will broadcast into the adjacent room when activated.
3- The judge will activate the microphone and signal the claimant, who will ask any question s/he wishes. This is played into the room with the recorder and noted for time.
4- The microphone is shut off and the recorder continues for a set period of time.
5- Repeat steps 3+4 a number of times.
6- Judging is done by changes in sound level on the recording that occur after the microphone has been shut off. Any increase of predermined decible level is considered a hit.

I think this would eliminate the option of cheating, decent sound equipment + shielding would remove random noise and the required change in decibles for a hit would eliminate any argument about what was said.

Instead of trying to remov random noise, all you have to do is have a proper control.

 

[Hitch]

Exactly, Hitch, which is why I think we need multiple recordings, and the applicant must pick the correct one (i.e.-the one with his answers) from among a group of empty recorders.

But what if the "spirits" decide to talk through the "control" recorder rather than the "test" recorder? Spirits can be like that sometimes. They always seem to do the opposite of what a skeptic would expect.

 

[Hellbound]

True, but we can minimize the legitimacy of this complaint by, for example, making the control recordings before the actual recording, or in other areas, or both.

Woos will always have an "out", an excuse, no matter how well the setup. I'm not really looking for something that will prevent them from making excuses (I don't think such a thing exists), but simply a way to show the irrational and ad hoc nature of these beliefs to those not completely taken in by them.

 

[drfrank]

IMO, the "voices" Lou is hearing is simply an artifact of how his recorder works. Notice that he is pretty specific about the model that he requires.

Interesting stuff, skooob, I hadn't heard thought about it like that before.

I guess that makes distinguishing human voices in the output very similar to seeing `canals' on Mars - pareidolia operating on compression artifacts.

 

[Mendeli]

I don't suppose Kramer just completely forgot all about mr Gentile's application... but it's certainly been a while already after TAM4. Well, maybe with what all has happened since with Randi has had an effect on things... or maybe they've had some private negotiations. Anyways, it's curious...the silence.

 

[Hitch]

...
Anyways, it's curious...the silence.

Sorry. That last line made me laugh in this context.

 

[Tirdun]

I don't suppose Kramer just completely forgot all about mr Gentile's application... but it's certainly been a while already after TAM4. Well, maybe with what all has happened since with Randi has had an effect on things... or maybe they've had some private negotiations. Anyways, it's curious...the silence.

I dunno, with 2 million (Randi + other skeptics) I'd be making a riotous amount of noise to move things along.

Plus the book sales!!!:

How I Talk To The Dead
by Tirdun

[Cool swirly graphic here]

Only winner EVER of the 1 MILLION Dollar
Randi Paranormal Challenge!!

 

[KevinM]

The Reason

Ok first I'm a coleague of Mr Gentile and in fact was co author on both the original press release and on the application. As such I can probably clear a few things up.

1) The reason that we suggested the evps be analyzed independently is very simple. It gives the ability to compare voice prints against participitants and thus rule out some one whispering under there breath.

2) As to the suggestion of sealed envelopes or containers lets be clear about some thing. We stipulate the voices are of an unknown origin. No where in the application do we state or presume they are ghosts, angels, demons or any other spiritual being. Do we have an oppinion certainly but as I'm sure you'll agree establishing the phenomenon comes before proving the origins. We're currently looking into running this sort of test to see what sort of results we can obtain.

3) Kramer has contacted us and the delay is in relation to the bypass. Randi apparently is interested in being involved directly so as such we are happy to wait for his recovery and wish him well during this time of hardship. Mr Gentile and I both respect Mr Randi and were very sorry to hear about his operation.

4) As to the recorder unfortunately its been discontinued by the manufacturer. Getting another one of the same type may be very difficult which is why we suggested turning one of his recorders over to a qualified engineer to check for any trickery. If one can be found that is an original from the manufacturer then we would in fact prefer to use it.

5) As to the quality of voices given a large enough sample(and the sample laid down in the application WOULD qualify) Mr Gentile can get completely clear voices. The only manipulation that might be required is speeding up or slowing down the recording itself(a function available on the recorder) because some of the voices tend to come through faster or slower then normal.

I think I've covered most of the major questions I will be checking back occasionally if any one has any additional questions.

Kevin Meares
Co Host "The Lou Gentile Show"

 

[petre]

Well resoned, a departure from the usual entry. It is no surprise Randi wants to get personally involved. Best of luck!

Since you'll be back for questions, I'm wondering if you've identified steps that may make testing easier that won't interfere with success? For example, will it work if the recorder is sealed in a box? In another room (maybe with a speaker to project the question, which is then turned off during the "answer" period)? Finding the limitations may ease negotiation later.

 

[Ducky]

Ok first I'm a coleague of Mr Gentile and in fact was co author on both the original press release and on the application. As such I can probably clear a few things up.

1) The reason that we suggested the evps be analyzed independently is very simple. It gives the ability to compare voice prints against participitants and thus rule out some one whispering under there breath.

2) As to the suggestion of sealed envelopes or containers lets be clear about some thing. We stipulate the voices are of an unknown origin. No where in the application do we state or presume they are ghosts, angels, demons or any other spiritual being. Do we have an oppinion certainly but as I'm sure you'll agree establishing the phenomenon comes before proving the origins. We're currently looking into running this sort of test to see what sort of results we can obtain.

3) Kramer has contacted us and the delay is in relation to the bypass. Randi apparently is interested in being involved directly so as such we are happy to wait for his recovery and wish him well during this time of hardship. Mr Gentile and I both respect Mr Randi and were very sorry to hear about his operation.

4) As to the recorder unfortunately its been discontinued by the manufacturer. Getting another one of the same type may be very difficult which is why we suggested turning one of his recorders over to a qualified engineer to check for any trickery. If one can be found that is an original from the manufacturer then we would in fact prefer to use it.

5) As to the quality of voices given a large enough sample(and the sample laid down in the application WOULD qualify) Mr Gentile can get completely clear voices. The only manipulation that might be required is speeding up or slowing down the recording itself(a function available on the recorder) because some of the voices tend to come through faster or slower then normal.

I think I've covered most of the major questions I will be checking back occasionally if any one has any additional questions.

Kevin Meares
Co Host "The Lou Gentile Show"

Thank you for your response. Just to be clear, I have nothing whatsoever to do with the JREF in any official fashion. I am curious on one point, though:

What protocol would be acceptable to ensure against cross modulation or radio interference?

 

[LordoftheLeftHand]

KevinM:

Previously I suggested a test protocol that would eliminate many of the potential problems with testing a claim of this sort. Would the following be an acceptable way to test this claim?

The applicant would be properly blindfolded next to a stack of cards. Each card would have a large number printed on one side. The recording would be started. The applicant would pick up the first card and show it to the room and ask the voices to speak the number that is on the card. After a set amount of time the recording would be stopped and the media would be ejected from the recorder and be labeled “1st Card”. New media would be inserted and the process repeated for the second card.

This would be done an agreed upon number of times. When finished the applicant would be given a sufficient amount of time to review the recordings. The applicant would write the number he thinks he hears on the media. When the applicant is finished, the number on the first card will be compared to the number written on the media. Repeat this for each card/media. If enough of the cards match the number written by the applicant on the media, the test would be considered a success!

This kind of test would likely require very little in the way of soundproofing or machine testing. Also there would be no reason to have some try to interpret what was said on the recording. The applicant can judge for himself what was recorded.

LLH
*note: I do not represent the JREF! While I like to think I have a good idea of what the JREF might require for a test, ultimately anything involving the Million Dollar Challenge must be done through the JREF.

 

[Flange Desire]

Ok first I'm a coleague of Mr Gentile and in fact was co author on both the original press release and on the application. As such I can probably clear a few things up. SNIP

Kevin - just a thank you.
Such clear communication is very much appreciated on this forum.

 

[Mendeli]

Sorry. That last line made me laugh in this context.

Ahh-a! You catched it! Normally nobody laughs at my gentle, yet intelligent hints, puns and stupid jokes...

Except this time even I didn't get it before you did, (insert curse word). It would have been so great to have been able to honestly claim it was intentional.

edit: Oh and about the recording device, it might be worthwhile to check ebay every once in a while. If lucky, such a device just might be found!

(and thanks KevinM for answering about the delay!)

One question though.

As to the quality of voices given a large enough sample(and the sample laid down in the application WOULD qualify) Mr Gentile can get completely clear voices.

As important or perhaps even more so as the fact of if the voices are clear or not, is the question of if the voices make any sense.

If you show a card with a number in it in a room while properly blindfolded and ask what the number is, (etc just like lordofthelefthand suggested) will it be more likely that the voice answers something like "five" instead of "green" or "rapsberry juice"? I know this might sound offending, but please bear with me, I'm honestly curious and fashinated about this!