You have just purchased one of the latest Speech Dictation Programs. Your processor is "a bit" slow, and your computer is more than "a byte" short of memory. So, can you "run it" anyway? You try to "load it". Much to your surprise it loads. Much to your pleasure it runs. Congratulations, that is where your pleasure ends!
Bigger And Better Power Intensive Problems.
New "power intensive" applications such as Speech Dictation, and Optical Character Recognition [OCR] are now programmed to take "full" advantage of "all" the annually escalating processor -CPU- power. As well as the ever growing addition of minimum memory requirements.
Power intensive applications now draw on both CPU and memory. Increasing one over the other will not resolve their "power hunger"; as they need to feed on both.
In both Speech Dictation, and Optical Character Recognition a "starved" program is a "slow" program, and a slow program makes mistakes. Like oxygen deprivation in humans, "smart" applications that are "power starved" over time become "stupid" programs.
In Optical Character Recognition this can be seen in text quality, the ability to recognize different type styles, as well as obvious uncorrected spelling errors. In Speech Dictation it is like "Alzheimers", an inability to retain new words, gradual loss of voice recognition "training" and ability to recognize words, or respond to command phrases.
If you want your applications to work "right", look at the power output of your present computer. As well as the minimum power consumption requirements of all your programs.
Bigger And Better Noise Problems.
New Speech Dictation programs "hear" better. Noise suppression microphones are the last line of defense against external noise and are only part of the answer to noise problems.
The first line of defense against noise is your computer work space. [eg: fans, furnaces, air conditioning, radios, televisions, background music, wind chimes, open windows and doors, street noises, people noise, machine noise, light fixture noise, etc]. Take control of your environmental noise first.
Your Speech Dictation program hears "electronic noise" better than you will. All monitors generate electronic noise. Large screen monitors use more power and generate more electronic noise. Electronic noise can slow the ability of Speech Dictation programs to recognize and respond to voice input over background [electronic] noise. Consider moving your large screen monitor off -the- top of [and away from] your desktop computer body.
As Speech Dictation requires bigger processors, big processors require bigger power [transformer] sources. Powerful processor and power sources generate electronic noise.
The next line of defense is to place Sound Cards away from both the processor and power source. Sometimes, inside small computer cases, the Sound Card needs to be shielded. When space and shielding do not work with new 64 and 32 bit Sound Cards; go back to the older [less sensitive] 16 bit Sound Card.
The last line of defense is the microphone. Start "testing" with telephone company "quality" head sets. Then test hand and stand held microphones with [foam top] "wind guards". For those to Do -It- Yourself, cut the bottom out of a Styrofoam coffee cup and use it as a "directional - sound - cone" around the head of the microphone. Finally, test noise suppressing microphones. Be sure to "test" first and "try" before you "buy". At their best, noise suppressing microphones are very expensive. However just buying microphones to "try" can be an expensive collection.
Top | ACSP Home | SuperAdaptoid Column