Investigating via android?
Posted 30 April 2012 - 01:24 PM
For instance digital audio processing. It is a common thing to run EVPs through some kind of software enhancement. While a limited amount of frequency filtering may not hurt (in some cases!) noise reduction is a killer. The various methods employed actually can alter the content of the audio. And when a recording is already questionable that may be enough to make it appear a voice is present when it is not. Especially bad if white noise is also present.
Some take the Ovilus seriously; it is not. Consider HOW the Ovilus works. One needs to know exactly what a spirit would need to do in order to create a message. The mechanics of it ARE important. When one takes that into consideration it is clear that the requirements are so improbable that to generate a meaningful message is simply not likely. But of course random words will once in a while come out and you may actually get something just by dumb luck. After all, even a stopped clock is right twice a day!
Posted 30 April 2012 - 03:05 PM
Can you recommend any resources to learn more about audio and how to analyze recordings in a way that reduces false positives?
I value your sage advice, so thank you very much!
Posted 30 April 2012 - 09:50 PM
The easiest way I found to analyze EVPs is to use the Rule of Three. Start with the original recording and play it for 3 people who have never heard it and have no idea what it may say. Each writes down his interpretation without consulting anyone else. You look at the results. If all three agree, you have something worth considering. If two of the three agree, and the third is reasonably close, also worth looking into If the third opinion is way off, the EVP is probably false or a result of paraedolia. And of course if none agree it is junk. Toss it out.
As for what you can do to process an EVP for analysis, there are only three acceptable practices. (1) You can amplify it to make it louder. That means you may also be increasing the background noise level. If by doing so you still can't hear the EVP it is not a reliable recording. But you may want to try number 2 before actually trashing it.
Number two is frequency isolation. You can reduce those frequencies outide the voice band IF they are creating some of the noise. That means cutting everything below about 15o Hz (300 Hz if the voice appears female) and cutting everything above 5 kHz. Listen to it again and if it is now intelligible then you have accomplished your task. If not, toss it out. BTW, this is bets done with an analog graphic equalizer not a computer. Analog equalizers can work with ALL frequencies simultaneously; computer equalizers must work each pass band individually using algorythms. These can destroy data by altering the signal.
The third practice involves ONLY those cases where someone has attempted communication. Often the investigator is clearly heard while the reply is much fainter. It is acceptable to set a limiter to reduce the loud audio to make the recording more uniform PROVIDED you set the limiter at least 10 db above the level of the EVP response. Thus only the investigator's comments are altered; the actual EVP evidence is preserved in its original form.
All of this is best done using analog hardware, not computer software programs. That will preserve the integrity of the original recording.I generally record on analog tape however if you use digital it is acceptable to play the recording back through analog equipment for analysis provided the minimum standards for sample rate and non-lossy files are maintained.
A final omment, any file you keep in digital format MUST be in a non-lossy format. That is WAV or PCM . MP3 is NOT ACCEPTABLE for analysis. You can use it for posting and sharing where only listening is intended, but if any serious study is anticipated, it must be non-lossy. That also means if you recorded the EVP in digital the recorder must also be non-lossy. After all, once data is destroyed it is gone forever.
Which brings us back to analog recording. By its nature, assuming good quality equipment is used, analog will pass the entire audio spectrum, at least between 20 Hz - 12 kHz. (Some will go higher). Thus by nature it is "non-lossy". Thus the lossy versus non-lossy is not a factor if analog is used. Simplifies things greatly!
Edited by CaveRat, 30 April 2012 - 09:53 PM.
Posted 01 May 2012 - 07:41 AM
You said: "A final omment, any file you keep in digital format MUST be in a non-lossy format. That is WAV or PCM . MP3 is NOT ACCEPTABLE for analysis."
So now I have to ask, what is the acceptable format, if WAV, PCM and MP3 is not acceptable?
Most of the EVP I've caught has been in response to investigator's questions, however, one was when my name was spoken; another was laughter in response to a joke I said - laughing right along with me.
I used a Sony Linear PCM ICD SX750 - is this okay and if not, what do you recommend I use?
My goal is to get real EVP - even if that means only picking up one once in a while - I'd rather find nothing than to make one up or think I hear one when I don't.
Another point I forgot is what about when you actually hear the whisper or answer aside from the recording? For example, on one investigation, I walked into the room and asked "Is anybody home?" and my fellow investigator heard a "Yes!" and marked it. Later on when listening, I heard it, too - at the time I wasn't willing to admit I heard it because it was so faint and I was afraid I was confusing the sound for something else. You can't argue with the human ear AND a recording, right?
What do you think of the Sound Forge program?
Sorry for all the questions, but I truly respect your knowledge and expertise. If I'm going to try to capture EVP, I want to make sure both myself and my crew are doing it right. We don't want to be one of those groups posting false evidence for the sake of having something to show. One of our locations is chock full of voices. The last time, something or someone was following me around, really getting a kick out of my jokes and what I was saying - this entity had a good time that evening!
Posted 01 May 2012 - 08:02 PM
Your recorder is OK if used in its PCM mode and highest quality setting with one drawback. According to the spec sheet the A to D converter is only a 16 bit process; I recommend a 24 bit process. The sample rate is fine as long as you use settings above 96 KBPS; the recorder has two options, 128 and 192 K. Either meets specs. I don't have the full operation manual so I can't say for sure what options you have when recording, but as long as you use the WAV / PCM setting and the file is saved as a WAV file you should be OK.
Can you use external mics? I prefer the mics for EVPs be placed about 12 - 20 inches apart for better separation to allow for spatial analysis. While placing the mics close and angled away from each other gives some separation the time lead / lag between them is not sufficient to do this process.. (It relies on the speed of sound the determine direction and sound arrives at both mics almost at the same time when they are close together.)
EVPs that are both heard and recorded do have more credibility but are still not irrefutable. It is possible something mundane may create a sound that is misidentified. Of course the more relevant or detailed the response the more valid it becomes. In other words if a question is asked, "What is your name?" and the reply comes back "Tom". it is not as valid as if the reply was, "My name is Tom". Reason is something could fall and make a muffled sound which you only think sounds like "Tom". But the likelyhood of something falling and creating a complete string of phonemes that you misidentify as "My name is Tom" is much less probable.
Regarding Sound Forge I have heard of it but never used it. Thus I can't give an opinion aside from the generalities that apply to all digital audio programs as mentioned in my previous post. I prefer analog for the simple reason that software only simulates the effects. Consider frequency filtering as an example. In analog I use a 5 pole Butterworth filter comprised of actual capacitors and resistors which generate the time base. Software simulates the same effects. But which would be the most accurate, the real thing or a simulation?
Kudos to you and your group for wanting to get serious and maintain quality over quantity. Now if only you could get some others to adopt THAT policy....
Posted 02 May 2012 - 08:26 AM
And I agree with you last statement - there are some groups out there that claim to have tons of evidence, but when you look at it, it's really questionable. In one photo, they claim there's a spirit on the glass window, but upon looking very closely, there is no doubt it's the reflection of the investigator holding a flashlight as they pass by the window.
I've been on many investigations and can honestly say that only 2 out of them all have resulted in interesting phenomena.
Fort Mifflin is one of them; the other was this old factory we return to again and again.
Posted 04 May 2012 - 04:57 PM
( And some claim to get EVPs every time they go out.....)
Posted 04 May 2012 - 05:17 PM
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users