In this article we compare and contrast the different kinds of voiceover options that are available to us today. The article aims at making an informed choice.
There are two broad categories of voice options available today. Artificial or machine voice, and a natural human voice.
Artificial voices are of two kinds.
-
Synthetic Voices:
These are truly artificial sounds, developed using machines. The smalls small elements, called ‘phonemes’ are artificially generated and using software and algorithms and rules of orthography of the respective languages, are trained to pronounce words based on these phonemes.
These sounds lack the natural feel and are often described as robotic, artificial and lacking in emotiveness. The voices are monotonous, and completely devoid of modulation and emotion.
-
Artificial Intelligence based Voices:
These are based on deep learning. Human voices are recorded to cover a wide range of phonemic combinations and intonation, such that they can be reproduced to create new renditions for completely different texts. The quality of these voices is improving day by day as more and more work is going into artificial intelligence and deep learning and computational linguistics.
Human voices on the other hand, are recorded by a real human beings, while reading from a script. All the attributes of human understanding and rendition come into play and the listen is natural and emotive.
Human voice overs can be amateur or professional.
-
Amateur artistes
will usually record their voice overs using non-professional equipment such as phones and personal computers. Their voices are inexperienced and the recordings lack the professional touch. They will be marked by background noise, reverberation, breathing sounds, salivation and fricative noises.
-
Professional voice artistes
record in professional studios. The voiceovers reflect deep experience and keen knowledge of modulation and intonation. Depending on the level of experience, a professional voiceover will be clean of salivation, breathing sounds and reverberation. There is almost no background sound. All files are studio edited and cleaned for pleasurable listening.
Pros and Cons
The choice of voiceover takes into account the following factors:
- duration
- legality
- cost
- control
- production time
- purpose and
- output quality
Synthetic Voice
These are cheap and the creator has easy control. The creator can manipulate the voices merely by editing the text. However, the output is robotic and sounds unnatural. It lacks modulation and gets monotonous beyond a few minutes. These are best suited for simple navigation, interactive voice response systems, and narrations not extending a minute or two.
Artificial Intelligence Voices
These can be used for longer sessions such as e-learning modules and slightly longer narrations. These are cheaper and give greater control and allow for engineering and manipulation of intonation, pace and other prosodic features. AI voices are getting better and better and some voices are better than others. The advantage is that professional sounding voices are available where human voices are not.
The limitation is that it lacks flexibility especially when it comes to acronyms, foreign terms and switching between languages. This methodology is absolutely discouraged for entertainment purposes. Also, the amount of work required in coding the prosody is quite a lot – which makes it almost as expensive as human voiceover, without the flexibility that one has with a natural human voice.
Further, when we take a multilingual perspective, we will see that not all languages have evolved AI voice software. Whereas some countries have very advanced AI options such as English, French and German, many are yet to reach the desired level of ‘naturalness’ and ‘flexibility’ in the range of emoting. Some of the better evolved languages available besides English, French and German are Chinese and Arabic. Unfortunately, Indian languages are not very well developed in AI. Hence for a developer, it is not an easy choice and the same parameters cannot be applied across the board in multilingual projects.
There are several AI based text-to-speech software available today. Some of the more famous ones are :
Human vs. Artificial Voices
The dialectic boils down to cost versus quality ultimately. Turnaround time and convenience play an important role too. An artificial voice will give us the control and flexibility for a quick output. But, for professional presentations videos and films, a professional voiceover will provide clarity and emotiveness. The punch words are effectiveness and IMPACT.
Amateur vs. Professional Human Voice
While choosing among human voice options, we have other issues to consider. An amateur may take five to ten takes where a professional artiste will take only two or just one. An amateur may not provide a clean voice where a professional voice artiste will provide a hassle-free, studio-ready recording that requires no further editing.
WordPar International endorses the use of professional human voices as well as machine voices based on AI. We have a large network of professional voice actors as well as a group of linguists who will produce tailor-made voiceovers using AI software. Our team of artistes and linguists are reliable and produce high quality work.
WordPar discourages the use of synthetic machine voices and engaging amateur human voice actors.
What We Offer
WordPar International endorses the use of professional human voices as well as machine voices based on AI. We have a large network of professional voice actors as well as a group of linguists who will produce tailor-made voiceovers using AI software. Our team of artistes and linguists are reliable and produce high quality work.
WordPar discourages the use of synthetic machine voices and engaging amateur human voice actors.
1,104 views
The post Professional, Amateur and Machine Voice-over appeared first on Wordpar.