Class B EVPs suffer from an identity crisis, fraught with credibility issues and doubt.  As such, they're easy to misclassify.  In fact, the uncertainty about so many Class B EVPs has raised many a debate as to whether or not another system for classification should be used....

Class B EVPs, the "middle class" of Electronic Voice Phenomenon, often have the same identity crisis and credibility issues we associate with the social middle class:  some aspire to be A list-ers but are held back by their inherent native flaws; some are better than any other EVPs caught by the investigator, so they're assumed to be legends in the investigator's own mind.  Some are even victims of a profound lack of confidence and are cynically mislabeled "Class C."  And when I'm finished beating to death this extended metaphor, I'll get to my point:  Class B EVPs are slippery to identify.

The definition of Class B most often used by investigators is that they are audible but frequently difficult to hear without the use of headphones, and that fellow investigators will agree to disagree about them.  That doesn't inspire confidence, I know.  In fact, the uncertainly about so many Class B EVPs has raised many a debate as to whether or not another system for classification should be used.  For example, one proposed ratings system would use a scale of 1 to 5 to assess two main criteria separately, clarity and volume; these would combine into a decimal number, one of twenty-five possible EVP ratings, ranging from 1.1 (the very poorest) to 3.3 (average) to 5.5 (superlative).  While the ingenuity of the dedicated has inspired some creative solutions, the most accessible and common method for hobbyists remains the letter grade system, in all its primitive charm.  However, one very pragmatic method used to work around its ambiguity is to divide and categorize the middle ground even further, just as our third grade teachers did, and to add for a "plus" or "minus" when needed.  Some have criticized even this method as a failure of nerve to address the real faults that hold back an EVP from being Class A.  No doubt, the same psychological motivations that drive EVP analysts to over-indulge their Class C EVPs lead them to vacillate about Class B EVPs as well.  It's emotional for many of us, so I'm in no position to judge.  And it's humbling for all of us, for it reminds us just how imprecise our tools of assessment really are, despite how driven we feel to find precise and compelling evidence.

That being said, there's an awful lot of good to come out of the Class B EVP finding, because more often than not it is compelling enough to merit further attempts at categorization, and even explanation.  When a finding is less obscured by noise and less clouded by doubt, researchers are more willing to dissect it and pose theories without getting caught up in the metaphysical wonder of it (as they do with Class As) or the technical blunder of it (as they do with Class Cs).  Furthermore, with their advantages, Class B EVPs can be used more reliably to corroborate eyewitness accounts and video findings, which helps to construct a more compelling case for the paranormal activity.  Some of the factors that make for a Class B EVP are as follows:

Recognizable speech: When the gender, and even the general age, of the voice in question can be discerned, this works enormously in an EVP's favor because it can be connected to background information about the alleged haunt.  This doesn't automatically mean a positive identification of the voice is possible, but anything that brings the finding closer to the facts of the case's background increases the odds that the EVP is a reliable piece of datum.  Beware, however, of not stereotyping or profiling voices too readily.  There are many similarities in pitch and timbre between the voice of a mature woman and an adolescent male, or between a young woman and a boy; likewise, some male voices are higher pitched than others.  Leave some room for diversity in your expectations.  Furthermore, consider language and dialect.  A common (and, in my opinion, lazy) mistake made in EVP analysis is to relegate anything that sounds like gibberish to "Native American language."  Firstly, there is no "Native American language"; there are many nations of indigenous people in this country, most of which had their own language and/or dialects.  Before sitting down to analyze the audio, try to research which cultural attributes and eras might be relevant to a venue's background, and remain open to the expression of those attributes in the findings.

Details: Just as speech is more recognizable in Class B EVPs, so too is the content of that speech, which is all about the details.  Unlike Class C EVPs that often communicate one-word responses of no particular importance, Class B EVPs tend to be actual communication.   They are more likely to be contextual:  directly responsive to an investigator's questions or actions.  Even something as simple as "Get out!" is impressive, so long as a general consensus can be reached that this is what the EVP is saying.  The details of speech in a Class B EVP are more persuasively relevant, in other words.  In the lesson on Class C EVPs, for example, if the white noise example titled "Tumbleweed" had sounded less like a random utterance and more like a direct response to our opening of an attic door from which insulation was falling (e.g., "Blows like tumbleweed"), it might have been more relevant and contextually detailed enough to be labeled "Class B."  It would have also discounted the notion that the eleven-year-old girl was responsible for it, since she would need to have been physically nearby to have understood the context.

Volume: We want nothing more than our presumed spirits to speak clearly into the mic.  If it were easy, however, Class A and B EVPs would not be so rare.  What causes Class C EVPs to lack the gravitas they might deserve is their inability to surpass other competing sounds in the environment, leaving too much chance for mistaken interpretation.  The volume of Class B EVPs takes out some of that risk because the investigator will have taken enough care to inventory those competing sounds while she's investigating, so that they can be ruled out later.  It doesn't always work out, especially since the environment inside the recording device or, more commonly, the microphone cannot be known and therefore cannot be catalogued during the investigation.  A single hair, brushing against the mic, will probably leave an acoustic signature; so too might a gust of wind or a fly alighting.  Inside the equipment, noises are being generated also; anything with internal moving parts can be a culprit:  springs can react to changes in air density and temperature; motors and fans can leave audio artifacts; equipment housing contracts and expands.  Better pieces of sound equipment can lower the risk of this, but microphones are still apt to be affected by electromagnetic sources, radio waves, and so on.  As I've stated elsewhere, our sound equipment is designed to hear better than we can.  Don't be surprised by its refined abilities.  Celebrate them, but be wary of any potentially invisible acoustic sources that are not paranormal.

Layered voices and superimposed sounds: EVP specialists Debbie and Mark Constantino have devoted study to an entire category of EVP characterized by "layered" voices.  In their theory, a cross-chattering EVP is harnessing the vocal patterns, voice modulation, and intonation of one or more living people in the room so as to generate a spirit communique -- kind of like cutting letters out of a newspaper to piece together a ransom note.  [PPI researchers are currently examining data that may discount this phenomenon as nothing more than an artifact of a digital mic.] The Constantinos have also identified a subgroup of superimposed EVP wherein spirit voices mimic the words of the investigator immediately after they are spoken (and sometimes even before they are spoken).  By nature, such a finding can never be better than a Class B EVP no matter how impressive or unusual it may sound.  And filtering does little to unknot the two or more voices.  Spectral analysis can probably verify that separate voices are present, but beyond that, it is clear that the paranormal voices are inextricably bound to the acoustic sources in the room, and that will always make for shifty and unreliable findings.  One would think that capturing a clearly spoken EVP would be enough to evoke widespread support, but bad luck scuttles a good deal of these findings of interest that might have been proof positive.  A more cynical way of thinking about EVP voices coinciding with other voices and sounds is that it's just a matter of bad timing.  Because we don't know if we're receiving answers to our questions in these situations, neither can we know that we are talking over those answers.  We know nothing conclusive about the so-called physics of EVP communication, so even my cautions should be taken with some skepticism.  However, there seems to be some lag time between questions and contextual replies, as though the sound were traveling over great distances or more slowly.  At times, it's reminiscent of a speakerphone:  you have to be careful not to shut down the other side of the conversation because you can't always hear the off-handed responses in time.  When both parties are talking at once, communication is lost to both.  In EVP recording, when "spirit" voices coincide with those of the living, it damages both the details and the recognizable features of the voice, thus adding more risk to the findings.  Adhering to interrogation protocols that 1) allow for ample time between questions, and 2) dictate how other investigators should interrupt the interrogation when it is necessary, may help to reduce the risk of superimposed EVP . . . unless, of course, you want to capture this kind of EVP believing it is a phenomenon unto itself.

Once again, these common Class B features are not all inclusive, nor are the ones provided as black-and-white as I have described them.  In fact, the one universally true thing we can say about Class B EVPs is that there are a lot of grey areas in their identification.  However, the Class B EVP brings you closer to real paranormal sleuthing, which makes time codes and an organized plan for synchronicity indispensable.  Whether you plan to investigate in a team or go solo, recording the start time of your audio in adjusted GMT will later help you to cross-reference your audio with other recorded media and efficiently locate the time codes of your selected findings.

Examples

The following annotated waveform videos will illustrate why Class B EVP are sometimes more intriguing than they are reliable. You can probably hear these events of interest without the aid of headphones, but they're still recommended to enjoy the full effect.

As you can see, the common theme of these EVP samples is not mistaken identity, as it was with Class C, but with doubtful circumstances that, despite the quality of the recordings, hold back the an audio clip from assurances of it being paranormal.  Try to remain clear-sighted and honest about these factors:  get a second and third opinion other than your own.  Although it may seem frustrating at first that others don't agree completely with your interpretation of the audio anomaly, it's important to know which aspects of the EVP fail to be generally convincing or fail to yield a general consensus.  When these two ingredients come together, you'll know it, and you'll have found your Class A EVP.