Save a ton o' time by batch processing your audio with my Audacity macros! www.RadioDJDude.com/shop Also, as with any chatbot-generated info, make sure to fact check what's important to you.
Some might see this as a new frosting on an old cake but I appreciate it. When I launch I want to choose six voices and use them in four hour blocks for short reads like these. For somewhat longer reads leading up to a track or covering opinion or a recent or historical event use a somewhat high energy instrumental bed at low volume and put strategic pauses in the read for impact and flavor but NO DUCKING in the music bed, just keep it low. These things allow you to create a running personality for each voice to the extent that you could even give each voice it's own on air name. At the 4 hour you could create a sort of dj handoff AI convo. Listeners will fall in love with these "people."
Sounds like a fun idea! Have you checked out Notebook LM from Google? They have a male and female voice that converses with each other. Sounds incredibly authentic. Perhaps they can be prompted to have a :30 sec chat about a particular subject. Could be interesting...
Could a twist on this be done with "What Unkown Holidays Are Today?" Y'know the ones like National Peanut Butter Day, National Paper Airplane Day, etc?
Question on this process. If you delete couple of these because they sounded bad, how do you go back and re-generate them? I had to switch voices in the middle of this because, while I could get the voice to sound good on ElevenLabs, but when I let the automation run, the voice was speeding through them, and stopping abruptly, like it was in the middle of a sentence, but it wasn't. So, after the first bunch, I deleted all of the bad ones and ran the rest. I have 22, so there's another 8 left for November that I'm not sure how to generate.
Yeah, unfortunately, the bots don't always nail the read. I end up having to re-do 5-7 out of 30. In make.com, there's a way to tell the spreadsheet module which line to start at. OR, if you only have a handful, do them one at a time in ElevenLabs, so you can tweak settings to get a good sound.
Never said I was creating something original. 😀 As mentioned, it's been around since the microphone was invented. BUT, my version is much snappier than most. No need to run a minute of rando history items. Boring!
![Twenty One Pilots - “The Line” (from Arcane Season 2) [Official Music Video]](http://i.ytimg.com/vi/E2Rj2gQAyPA/mqdefault.jpg)
Save a ton o' time by batch processing your audio with my Audacity macros! www.RadioDJDude.com/shop
Also, as with any chatbot-generated info, make sure to fact check what's important to you.
Some might see this as a new frosting on an old cake but I appreciate it. When I launch I want to choose six voices and use them in four hour blocks for short reads like these. For somewhat longer reads leading up to a track or covering opinion or a recent or historical event use a somewhat high energy instrumental bed at low volume and put strategic pauses in the read for impact and flavor but NO DUCKING in the music bed, just keep it low. These things allow you to create a running personality for each voice to the extent that you could even give each voice it's own on air name. At the 4 hour you could create a sort of dj handoff AI convo. Listeners will fall in love with these "people."
Sounds like a fun idea! Have you checked out Notebook LM from Google? They have a male and female voice that converses with each other. Sounds incredibly authentic. Perhaps they can be prompted to have a :30 sec chat about a particular subject. Could be interesting...
Could a twist on this be done with "What Unkown Holidays Are Today?" Y'know the ones like National Peanut Butter Day, National Paper Airplane Day, etc?
Great idea, Brian! I love that. "Rando Fact o' the Day!"
Question on this process. If you delete couple of these because they sounded bad, how do you go back and re-generate them? I had to switch voices in the middle of this because, while I could get the voice to sound good on ElevenLabs, but when I let the automation run, the voice was speeding through them, and stopping abruptly, like it was in the middle of a sentence, but it wasn't. So, after the first bunch, I deleted all of the bad ones and ran the rest. I have 22, so there's another 8 left for November that I'm not sure how to generate.
Yeah, unfortunately, the bots don't always nail the read. I end up having to re-do 5-7 out of 30. In make.com, there's a way to tell the spreadsheet module which line to start at. OR, if you only have a handful, do them one at a time in ElevenLabs, so you can tweak settings to get a good sound.
Some online stations already do this.
Never said I was creating something original. 😀 As mentioned, it's been around since the microphone was invented. BUT, my version is much snappier than most. No need to run a minute of rando history items. Boring!
@@RadioDJDude true, true.