knight rider online

Hello all,
I'm trying to gauge of there would be any interest in a custom text to speech (TTS) voice from William Daniels. The costs for building a voice are prohibitive for a single person, but in aggregate may be within reach. The custom KITT reproductions I've seen are almost perfect except for the voice coming out of the voicebox.

There are a few companies online that build such voices and I've reached out to some of them to get a ballpark figure. There would be fees associated with getting William Daniels to do a few hours of voice recording to form the foundation. I'm not sure if this could be done independently of any type of KR licensing, or that would be another set of expenses. Finally, the expenses for a shell company and the distribution and licensing would be minimal.

Based on the fact that William used his voice talents for the MIO GPS, this doesn't seem too far fetched. I've been attempting to to use various digital signal processing techniques for voice modification to replicate this for a long time without any success. With the advance of Kickstarter, the platform for executing something like this seems more within reach - even compared to a few years ago.

Ideally this could be extended to develop a turn key embedded device that could be "dropped" in the replicas, but one thing at a time.

Regards,
Jonathan

William's voice has changed since the 80's which is why I never thought bought a Mio. I have a few hundred recorded clips from the series. Is there any way you can create a computer program to analyze the voice pattern and create the perfect Kitt gps voice?

Creating a TTS voice is much more complex then it sounds - it's the subject of many PHD thesis papers. I know his voice has changed, but based on current voice creation technology, it's probably not going to make much difference. The voice will certainly be of a higher quality under a controlled studio recording with the sentences pre-selected for optimal training sets. Also, simple modifications could be made after the voice creation such as pitch to bring it back to 1980s level.

The voice could be used for GPS, but I would envision it could be applied more broadly. The underlying system would be similar to a Siri on iOS to generate an almost realistic experience. Of course, I'm interested just in the voice part first.

Sounds great. I'm currently using the British female voice for my gps guidance and android assistant. I just really hope you can get it as close to the 80's as possible.