Voice cloning.
Yes, that’s a thing now.
Voice cloning & synthetic audio
We’ve been doing some creative 3d model work for the Coincade team and wanted to create a storyline for the main characters. It led us to play.ht and a few other AI tools that took the band-aid of our ignorance off.
AI Voice is a thing now. Marques, or MKBHD on YouTube did a short about this phenomenon with the JayZ track a couple months ago:
There has been a meteoric rise in digital content. And where this is quantity, there is automation. Based on my limited understanding, voice changers, models and AI tools were created for scalable and segment-able content.
When Cadbury launched their Celebrations campaign with SRK, the ad caused a shockwave in the digital and marketing industry. Although video and Audio components are both used, the AI element of audio is particularly impressive. Keen eyed audiences can find some frame drops in the video but the audio is impeccable. For India, this could be a milestone in custom ad campaigns and the use of AI. Next step is probably to scan SRK for a couple minutes with mo-cap and make the entire ad on a (fairly expensive) computer.
There are 3 specific trends that are guiding this tech
- The Rise of Vocal Synthesis: The advent of vocal synthesis has revolutionised the way we perceive and interact with human voices. Utilising cutting-edge algorithms and software, this technology enables the replication and modification of human vocal characteristics. From aiding individuals with speech disabilities to empowering creative endeavors, vocal synthesis offers a pathway for expressing oneself through a myriad of unique voices.
- Vocal Manipulation: Voice Morphing 101: The art of voice manipulation presents a captivating realm of possibilities, allowing individuals to modify pitch, tone, and gender, transforming their voices in unimaginable ways. Through the use of specialised software tools, one can seamlessly transition from a booming bass to a melodic soprano or even experiment with fantastical vocal effects. This opens doors for artistic expression, theatrical performances, and a plethora of entertaining endeavors.
- Vocal Deepfakes: Impersonation on Steroids: The emergence of vocal deepfakes introduces a new level of impersonation that can be both awe-inspiring and cause for concern. By employing vast datasets of an individual’s voice recordings, these algorithms recreate their voice patterns with astonishing precision. While this technology offers exciting creative potential, it also raises ethical questions surrounding the misuse of voice impersonation, highlighting the need for responsible application and discernment.
Besides the trends, there are also ethical considerations. While the prospect of assuming different voices for amusement or artistic endeavors holds inherent appeal, it is crucial to recognise the potential for misuse. Unauthorised impersonation, voice-based scams, and deception raise concerns regarding privacy, trust, and personal security. Adherence to ethical standards is essential to prevent harm and preserve the integrity of communication channels.
The Future of Voice Modulation
As technology continues its rapid evolution, the future of voice modulation appears promising and enigmatic. Speculations abound regarding a world where voice customisation becomes a norm, allowing individuals to dynamically adapt their vocal characteristics to suit different contexts or even assume virtual personas.
As we venture forth with this technology, there will be a rise of Audio verification systems, digital identity hacks and content fakery. The scope of benefit and failure is equitable.
This largely depends on the part of the user.
Which side are you on?
Stay curious. Stay Secure.
See you next week.
The views and opinions published here belong to the author and do not necessarily reflect the views and opinions of the publisher.
Be the first to comment on "Hacking the human voice"