First I need to emphasize that this isn't intended to be an exhaustive treatise on the subject of voice talent marketing... it's just few ideas for some "Do's and Don'ts" that you can definitely benefit from. Someone once said: "Intelligent or not, we all make mistakes and perhaps the smart mistakes are the worst, because so much careful thought has gone into them." I love that! The info here might help you avoid a few smart mistakes in marketing!
Most text to speech voices are created by synthesizing complete phrases from separate words; each of these words is actually a record of a human spoken word. In order to create a well synthesized text, each recorded word must have features associated with rules that the voice talent should follow.
To prepare text to speech voices for universal use, speak with an arbitrary manner where a word doesn’t have emotions like anger, laziness, happiness, sadness etc... Arbitrarily spoken words give an artificial sense to the text and are considered a real advantage.
To prepare text to speech voices of a good quality, speak loudly and clearly, sound confident, and do not mumble. If you made an error, you have to record the word back and continue.
Remember that every word will be a part of a whole phrase which in turn will be a part of a whole text or a presentation, remember that communication is the key to a successful presentation. If you are short of time, always start with the most common and most repetitive words in the daily life, and chose the least repetitive words in daily life last. If you have extra time, aim at perfection and add as much vocabulary as possible.
Pause between words. Allow a little time between words so that the synthesis software doesn’t mix successive words together. Don't race through your dictation and always leave time to take your breath.
When using audio-visual aids to enhance your recording, be sure all necessary equipment is set up and in good working order prior to the recording. If possible, have an emergency backup system readily available. Check out the location ahead of time to ensure arrangements are suitable for your recording.
To prepare text to speech voices of a good quality, beware of ambient noise, there are special recording rooms that isolate those noises, try to rent one, but if you can’t, try to choose the quietest place as possible, also choose late hours of less activity around you to guarantee less ambient noise.
To prepare text to speech voices for universal use, always put in mind that your digital record shouldn’t need farther processing after you are done with it, which means you need to choose a universal recording format, and choose audio settings that are able to work for many applications, for example telephone lines don’t work with stereo recordings, nor do the programs that deal with PC and VOIP telephony, so mono recording is highly recommended to have a wider market. The same thing applies for the sample rate of your sound; choose an 8K sampling rate because analog telephone lines can’t carry the 16 K recordings!
Finally, when you start marketing your recordings, make sure your contact information is easy to locate and up-to-date. Do not try to dazzle them with your packaging. Concentrate instead on the content of your demo, providing them with examples of your best work.
More Technical Articles
Text-to-Speech and Voice Recognition Videos
Text-to-Speech Homepage
