Reasons for recording a talking bird
& suggestions for how to do it
Reasons to record a bird's speech

  • To permit friends to hear one's bird
    speak, as many birds do not speak
    when strangers are present.

  • To prove the identity of your bird
    through its speech.

  • As a history for study and
    familiarization by a new caretaker.

  • To listen to a bird's speech in order
    to gain proficiency at understanding
    the bird's dialect.

  • To permit others to study and
    analyze a bird's utterances,
    especially in investigations of
    meaning.

  • By listening to a bird's words,
    phrases, and sentences, one learns
    whether his parrot is a participant in
    interspecies communication by
    conveying  messages through speech.
You are here:
Recording
Birds
Arielle says what most Americans say,
click the play button to hear her words.

No transcription provided, as everyone
will recognize what Arielle said.
Recording speech sounds ultimately leads to comparing the words
using a computer as a tool for analyzing the sounds.

To see the sound spectrum contained in my speech and Arielle's
speech click on the word to the right.
Spectrogram
Arielle understands speech and speaks thoughtfully using English
words, phrases, and sentences.
Arielle perches at the right side of her
gym close to the microphone. The
scheme permits recording Arielle's
talking sessions from inside the house.
I improvised the microphone support  
from a tripod salvaged from a junked
projection screen.
Photo by the author
The basics of recording bird speech

How to make your own bird-speech recordings

Recording bird speech can be done with comparatively primitive machinery, and that is precisely how I
began my adventures with recorded parrot speech. A simple and inexpensive method is to use an audio
cassette recorder; the most basic and easiest to use is a self-contained portable recorders that many people
have stored away in a closet somewhere. Such a machine has to be placed close to the bird's cage or the
perch from which the bird speaks. The disadvantage of this approach is that the internal machine noise can
mask the desired speech, especially if the bird speaks softly. An excited parrot can overload the recorder
unless care is taken to position the machine a distance from a loud bird.

To obtain improved results, a separate directional microphone will yield better results. A general-purpose
cardioid microphone (one with a heart shaped sensing pattern) will reduce undesired sounds from positions
behind or to the side of the microphone. The user should experiment with the microphone position to obtain
optimum results. If possible, use a home component tape recorder, since it is likely to offer better fidelity
than many portable machines.

Avoid a common error: recording the tape at too high a level. Recording at excessive levels will cause the
sound to be distorted. Even when correctly recorded, an effect called "print-through" causes loud passages
to be recorded onto passages that are at a low level. After a few weeks of storage, this effect causes
pre-echos and post-echos upon playback; the intermingling of the magnetic fields on the recording distorts
the stored information. Despite the limitations, audio cassette recorders are available widely and represent
one of the easiest methods to capture you bird's speech sessions.

Due to the problems associated with print-through on audio cassette tapes, I switched to recording
Arielle's speech onto compact discs. The CD format does not suffer from the effects of print-through and
potentially offer a more permanent, archival, storage for data. People wishing to make recordings on other
digital recording devices are encouraged to use whatever devices they find convenient.

Recordists using digital devices should be aware that sound files, depending upon the resolution selected,
can consume a great amount of memory, and, as a result, are likely to require better management than I
find time to do. Digital systems often distort severely, so attention to the recording level is exceptionally
important. When first starting, I advise hobbyists to steer clear of expensive devices which might prove
extravagant if you lose interest in the project.

Whatever equipment you choose, the important thing is to make some recordings!


My purpose for recording

My bird-speech recording started to show people that Arielle could speak a variety of expressions. After a
while I came to recognize, not only was I hooked, but also that I had a full-time research project on my
hands.

Arielle generally speaks softly, so hearing her voice can be difficult. Because I had been listening to her
through glass doors (visible in the picture above) for years, in the past many words were either inaudible or
not understandable due to the intervening glass.

Confusion can result when listening to a series of extemporaneous phrases. If the listener does not initially
perceive the words, there is a chance that he will completely fail to comprehend the message. When one is
writing the words on paper, it is difficult to listen to rapid speech and write a transcription simultaneously.
Often by the time I figured out and noted a specific syllable, I had forgotten what she had said previously.
It is safe to say that I might have understood many unrecognized phrases if I had been able to refer to a
recording. Complicating listening is the fact that Arielle talks, sometimes continuously, for 12-20 minutes.
She does not repeat unique messages for a covert observer who might be listening to her words inside.

I began tape recording Arielle's vocalizing sessions several years ago. Now, I record her daily as she
speaks while perched upon her play gym. The amplified microphone placed outside close to her (shown in
the picture above) much improved my ability to hear her vocalizations. Using headphones for monitoring
makes it easier to concentrate upon her words, as the earphones minimize household distractions. The
sound is considerably clearer with the attenuating effect of the glass barrier removed. This system is so
simple that I should have made the connection years ago.

While Arielle's speech is sometimes difficult to understand, the availability of a program to review her
pronouncements permits repeated listening to statements. If her speech is slurred, likely target terms can
be determined from the context of her speech.

During the first two months of recording, she generated about two-and-a-half hours of articulations to
analyze. Arielle spoke 426 permutations of phrases and words as established by counting the occurrences
on the transcription for the recordings. There were numerous repetitions of common parlance. Her current
rate of production, which is variable, has increased to 80 minutes of speech recorded over a period as short
as two days.

Before recording her speech, there was no second opportunity to comprehend Arielle's expressions.
Compared with simply listening through the door, I now hear much more of Arielle's expression. It is
difficult to write as fast as Arielle can speak; if she starts at a high rate, it is certain that one will miss parts
of her monologue. By playing back the program, it is possible to fill in the words missed during simultaneous
transcription and real-time listening.

Over the last years, the macaw's speech production has increased markedly; either that, or, perhaps, I am
simply more aware of her vocal sessions. I suspect that Arielle's vocabulary will continue to grow, because
she seems to be warming up to perform something more spectacular and more astounding than what she
has already done. Arielle is a bashful individual, she generally speaks covertly when she is outside on her
perch with no people present. I am in an awkward position, since I have been unable to determine what her
clandestine plan is, but, in time, she will communicate it to me.

E-mail address: Mike@ParrotSpeech.com