• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

Computer (voice) as Host/"Talent"

kfbkfb

kfbkfb

For several years, the NOAA radio stations have used a computer voice to "read" the weather info (although the word "cloudy" is slightly mispronounced).

I listen the 102.5 Jack FM here (unfortunately moved to LPFM several years ago) quite often, it occurred to me that the voice could easily be a computer voice reading prepared text rather than a human voice.

With AI technology and existing computer voice technology, what do you think is in the near future for - Computer (voice) as Host/"Talent"?


Kirk Bayne
 
Given modern tech, I think it would easier to make realistic-looking, but graphicly produced TV news anchors than radio or VO talent. The problem with synthesized voice, is it's hard to get the tone, sibilance, and inflection points to sound natural.
 
Kelly A said:
Given modern tech, I think it would easier to make realistic-looking, but graphicly produced TV news anchors than radio or VO talent. The problem with synthesized voice, is it's hard to get the tone, sibilance, and inflection points to sound natural.
Click to expand...

If you watch the Weather Channel, they're able to do fairly credible local weather forecasts by editing words and phrases into full sentences. This same thing was done on automated radio stations in the 70s and 80s. There's no need to do synthesized voices unless they really want to.
 
The station I work for has a service where a 'voice' does the weather forecast, time and temperature. It's straight out of a computer and sensitive to local names/places. While I joke she sounds much like Alexa, it's a bunch of words spoken by a real person assembled by the computer as needed. I appreciate how the station has it worked up. Nothing is worse than a Tornado Watch where every word is read (including the 48 counties in the watch area). Ours simply says "for our listening area" or only for the counties affected (we only reach 5 counties). The interesting thing is one program I looked at for another station a couple of years ago had the ability to rephrase the forecast or severe weather warning each time it aired. I don't recall the number of variations before repeating.
 
TheBigA said:
If you watch the Weather Channel, they're able to do fairly credible local weather forecasts by editing words and phrases into full sentences. This same thing was done on automated radio stations in the 70s and 80s. There's no need to do synthesized voices unless they really want to.
Click to expand...
in the case of The Weather Channel, they usually have one of their meteorologist do the reading. in more recent times, it's Jim "if you see me in your town, that means bad things are about to happen weather wise in your area" Cantore that does the voiceover work for their LOT8s.
 
Will_H_69_9 said:
in the case of The Weather Channel, they usually have one of their meteorologist do the reading. in more recent times, it's Jim "if you see me in your town, that means bad things are about to happen weather wise in your area" Cantore that does the voiceover work for their LOT8s.
Click to expand...
It's still automated to an extent. He is not reading every forecast. So there is some AI involved.
 
There is a pretty limited vocabulary of weather conditions for Jim Cantore to pre-record. So rather than building up words from phonemes, they are building sentences out of phrases. Because English speakers tend to pause after every few words, it sounds reasonable, given good mic technique and rest during the recording sessions. It wouldn't be good if the phrase "one hundred fourteen degrees" sounded different than the other phrases because it was recorded on a different day or at a different facility.

The first few seconds of a Local on the 8s is made up of several phrases: "Currently in our area", "heavy snowfall and", "twenty-four degrees".

As far as a completely synthesized voice being a DJ, I'm not sure what the point would be. If you have someone writing a script for the computer to announce, there isn't any labor being saved -- just set up voice tracks.
 
PTBoardOp94 said:
It wouldn't be good if the phrase "one hundred fourteen degrees" sounded different than the other phrases because it was recorded on a different day or at a different facility.
Click to expand...
You mean like the infamous TV infomercials from Ron Popeil back in the day when he'd decide to change a phrase or the price of an item he was hawking, and they'd drop in an update voiced by him, but recorded on a different day/location and it was incredibly obvious that it was an overdub?
 
Last edited:
I remember that KCXL 1140AM, in Liberty, MO, used to have an overnight music show called "Transmitter Memories", that was hosted by an AI.
 
I would say give it time. Even ten years ago, Alexa was what Max Headroom would have been in the mid to late-70's.... "maybe, someday computers will be that smart and effective..." It's already here. They just need to iron out the glitches.
 
boombox4 said:
I would say give it time. Even ten years ago, Alexa was what Max Headroom would have been in the mid to late-70's.... "maybe, someday computers will be that smart and effective..." It's already here. They just need to iron out the glitches.
Click to expand...
Except that the character Max Headroom was a real person, edited and effects added to make him sound and appear as a computer generated character.
 
Some You Tube channels use computerized voices. It's on several channels that are doing "The Life and Sad Ending of..." videos. A lot of these are terrible about mispronounced names and words. Some are so bad they're doing videos like this on people who are still alive! (n)
 
I have heard a definite improvement in delivery of AI voice on YT videos (generally some historical ones where the voice is probably AI doing the equivalent of audiobook style delivery). There are still some glitches, but I can't see how this technology won't eventually be used for liners and possibly other announcements on radio, especially once they can be edited via the right software (spacing between words, inflection, and the like).
 
www.cbsnews.com

Synthetic Media: How deepfakes could soon change our world

Synthetic media, better known as deepfakes, could be a goldmine for filmmakers. But the technology has already terrorized women who have had their faces inserted into pornography. And it could potentially disrupt society.
www.cbsnews.com www.cbsnews.com
^^^
It is without a doubt one of the most important revolutions in the future of human communication and perception.


In the near future, radio "celebrities" (licensed by the real celebrities) could talk live about different things on different radio stations at the same time, thanks to deepfake voice tech.


Kirk Bayne
 
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.


Back
Top Bottom