Just a quick note. I've been told this is currently an alpha feature and will roll out further in the future. Although, apparently anyone can apply for alpha access after signing up at this link 👉mrc.fm/aivoice
The ideal would be if you could record a few paragraphs of a particular style and then use TTS to get that same natural, flexible style without having to record the new lines.
I’d say it’ll count the characters of the words you say (as it’s likely transcribing your speech). I’ll need to double check this. Let me know what you notice!
Does speech to speech also fix pronunciation? I'm not a native speaker and I'd like to use my own voice to narrate my books and then use speech to speech to fix pronunciation
Hi! I need to transform 20 minutes of voice audio, and always have it be the same voice, in a single transformation. How do you recommend doing it, taking into account that after 2 minutes the transformation can become unstable? I have a creator subscription right now. Thanks!!!!
Hi there! To achieve consistent voice transformation over a 20-minute audio track without instability, I recommend breaking the track into shorter segments and processing them individually before combining them back together for a seamless result using the Speech to Speech AI feature from ElevenLabs. Enjoy experimenting with this powerful tool!
Yes, I am using speech to speech. I understand it cannot be done in a single transformation. I'm giving a lot of freedom in the settings, that can make the voices in each new creation sound or behave differently, losing unity and coherence in the final result. Thank you very much for the response :) I thought one of the subscriptions would allow me to do this. greetings@@MikeRussell
What would be really helpful for authenticity is using voices that aren't that "good" and having maybe slightly bad audio. Such as a local dj that maybe smokes too much and there is a little bit of background noise.
I'm going to be starting a channel soon where I will need various voices for allot of different characters. Is this the eleven labs platform the one you would recommend?
Quick question: Does this AI technology have the capability to correct or neutralize accents? Specifically, can it transform a non-native English speaker's accent into a more standard English accent? Thanks for the insight!
No, this AI technology presently seems to mimic accents as shown in my examples. Hopefully in the future it’ll be possible to neutralise the accent. That’d be helpful.
Not just cool, profitable. I can see non native speakers like me, generating our own clone voice, and then getting the accent we want with the A.I For good or evil English is the main international language, and a combination of grammarly and this feature would be a revolutionary product most foreigners we would happily pay for. And now imagine this working in real time with call centers all over the globe…
@@enekotartun i have found altered studio is able to translate non native speech to an american or british speech while keeping the input speech's emotions in the output speech. they offer 3 minutes of free trial
Yes, the ability to easily generate different voices and accents is an incredibly powerful feature for content creators! Thanks for the kudos on my voices Dave. Ha ha! 😂
Yes, the quality of your microphone will impact the quality of a cloned voice as it serves as the input for capturing and reproducing your voice so there may be a slight quality change depending on which mic you use.
Can I upload a Japanese voice actress that I like to Eleven Labs and then do that feature if I speak in Japanese (I'm fluent), or does it not have the ability to do this in other languages yet? It seems to cost money to clone a voice or make one, hmm not even sure if it's capable of doing it in Japanese...*sighs*
You don't need any of those. Your voice is GLORIOUS. However, this tool is the game changer for people like me, who sound terrible, but can use the speech to speech stuff to run a faceless channel. CZcams would surely approve monetization to these voice recording because they sound so human.
Thank you so much for your kind words! I'm thrilled to hear that you're excited about the Speech to Speech feature and I completely agree, it's definitely a game changer for creators looking to run faceless channels and enhance their content with humanlike voices.
You're absolutely right, the 24 seconds will just be a barrier now. Still it’s enough to voice quick scripts. If it could be over 30 seconds we could voice an ad!
It's been incredible so far. Our wives are going to hate us. Can't believe how well it generates whispers. That blew my mind! _Try 'Trent Ashford' in the voices. There are two. You know them both, trust me!_
Your US accent was nearly as bad and offensive as my Australian accent. Let's just agree to leave each other's accents alone. We don't need an international incedent.
I was both impressed and slightly disappointed. My own voice has a British accent and I struggle to do other accents... but my expectations of the AI was that it would use the target AI voice accent. Instead, the American females voices now sound like my British voice and a little raspy too. Watching your video, the tricks seems to be to mimmick the target voice... but not everybody has this talent. Trying different AI voices I realised the differences between them were now a lot less and it was more like the entire cast of Harry Potter. I wish there was a way to control how much it sounds like your own voice vs the target voice. I feel the AI can be improved here... identifying the stresses and intonations but preserving the target voice and accent. But for a first attempt this is impressive technology.
Thanks for sharing your feedback! It's great to hear your thoughts on the AI feature - hopefully future updates will address these nuances for a more tailored experience.
![I Replaced ALL my ADOBE APPS with these [free or cheaper] Alternatives!](/img/n.gif)
Just a quick note. I've been told this is currently an alpha feature and will roll out further in the future. Although, apparently anyone can apply for alpha access after signing up at this link 👉mrc.fm/aivoice
Well that answers my question brcause I'm a creator subs!
I love you AI content! What mic do you use? A Neumann U87? @MikeRussell
that just takes me to the website. No option to sign up for the alpha.
Here ya go! Direct link found elevenlabs.io/request-alpha-access@@WanderingMendicant-qd7mv
Yep U87!@@LumpiDeluxe
The ideal would be if you could record a few paragraphs of a particular style and then use TTS to get that same natural, flexible style without having to record the new lines.
Absolutely! It’s got to be something that will come in the future. Glad you’re enjoying speech to speech!
I agree. As it, not very useful for me. I like my voice.
Mike, I think your voice is better than any AI out there
Thank you, I really appreciate the kind words about my voice - I'm glad my human vocal qualities can still compete with AI systems!
@@MikeRussell Agreed! His voice is awesome.
Good to se Mike has such high self esteem. Its good to be a fan of yourself. 😊
Thanks for the kind words! Confidence is important 😊
How does this work with "character count" in the different price plans?
I’d say it’ll count the characters of the words you say (as it’s likely transcribing your speech). I’ll need to double check this. Let me know what you notice!
Do Not pay for this. there is free open source software that does the same task better for free on your own computer.
Does speech to speech also fix pronunciation? I'm not a native speaker and I'd like to use my own voice to narrate my books and then use speech to speech to fix pronunciation
No. I believe it does not.
Hi! I need to transform 20 minutes of voice audio, and always have it be the same voice, in a single transformation. How do you recommend doing it, taking into account that after 2 minutes the transformation can become unstable? I have a creator subscription right now. Thanks!!!!
Hi there! To achieve consistent voice transformation over a 20-minute audio track without instability, I recommend breaking the track into shorter segments and processing them individually before combining them back together for a seamless result using the Speech to Speech AI feature from ElevenLabs. Enjoy experimenting with this powerful tool!
Yes, I am using speech to speech. I understand it cannot be done in a single transformation. I'm giving a lot of freedom in the settings, that can make the voices in each new creation sound or behave differently, losing unity and coherence in the final result. Thank you very much for the response :) I thought one of the subscriptions would allow me to do this. greetings@@MikeRussell
What would be really helpful for authenticity is using voices that aren't that "good" and having maybe slightly bad audio. Such as a local dj that maybe smokes too much and there is a little bit of background noise.
You make a great point. Let’s see where this tech goes!
I'm going to be starting a channel soon where I will need various voices for allot of different characters. Is this the eleven labs platform the one you would recommend?
Yes, I would highly recommend ElevenLabs for automatically generating multiple character voices for your CZcams channel.
right on man! thanks for getting back. I like your takes@@MikeRussell
Quick question: Does this AI technology have the capability to correct or neutralize accents? Specifically, can it transform a non-native English speaker's accent into a more standard English accent?
Thanks for the insight!
No, this AI technology presently seems to mimic accents as shown in my examples. Hopefully in the future it’ll be possible to neutralise the accent. That’d be helpful.
Not just cool, profitable. I can see non native speakers like me, generating our own clone voice, and then getting the accent we want with the A.I
For good or evil English is the main international language, and a combination of grammarly and this feature would be a revolutionary product most foreigners we would happily pay for.
And now imagine this working in real time with call centers all over the globe…
@@enekotartun i have found altered studio is able to translate non native speech to an american or british speech while keeping the input speech's emotions in the output speech. they offer 3 minutes of free trial
@@Intellectual99 thanks a lot! I will try that one
@@enekotartun let me know how it goes for you.
bahahah i love the recordings you did. technology is amazing
Thank you so much, Lila! I'm glad you enjoyed the recordings, and yes, technology truly is incredible!
Can it turn a stuttering Speech to fluent Speech ( using person ‘ s own Voice ? )
No, the new Speech to Speech AI feature from ElevenLabs would mimic the stuttering.
Great! How many seconds can you record per day/month with the Starter subscription?
The Starter subscription allows 30,000 characters and free allows around 10,000 or equivalent.
@@MikeRussell but I mean with the speech to speech feature, how many seconds of recording per day/month?
Wonderful! Do you know if you can upload more than 50MB (speech to speech)?
I’m unsure of the upload limit for this feature.
Awesome Mike, I love this. I guess for now this only works with English voices or not. Looking forward to use it with Dutch voices as well.
You're right, currently this AI voice cloning feature only works with English voices well but I think more languages will come soon.
What is the Speech-to-speech word limits? can't find that online
The Speech-to-speech feature from ElevenLabs has about 5 min now I believe.
Mike, this is cool. I often just want a different voice an try to explain the inflection I want. P.S. loved the accents.
Yes, the ability to easily generate different voices and accents is an incredibly powerful feature for content creators! Thanks for the kudos on my voices Dave. Ha ha! 😂
Nice, but this feature is only available in English (unlike the text-to-speech feature).
You're right, while I tested other languages the results weren’t great. The best result right now is in English.
It speech to speech does not show when I sign up
You need to get the Alpha access. Link under my pinned comment 👍
Thanks bro. This helped me.
You're welcome! I'm glad I could help you out with your audio needs.
This is EPIC!!! Love it. This changes everything!
Yep - truly cool, right!?
How can I choose other languages than English for Speech to Speech? 🤔
You can use the Eleven Multilingual models however I don’t know how well it works in other languages.
When I got to my ElevenLabs/Speech Syntheses I’m not seeing that Speech to Speech option. I have the Creator Plan, do I need a higher one?
The Speech to Speech feature requires Alpha access. I posted a link under my pinned comment. Enjoy!
Thanks brother,I need also on nepali language
"Appreciate the support! I'll look into covering Nepali language options for future content, stay tuned!"
Well this will help since my voice sounds terrible in the microphone.
That's great to hear! Speech to Speech technology can definitely help improve the quality of your audio recordings.
I want to give a text and use that text with my own voice. How can I do this?
Yep, ElevenLabs allow you to clone your own voice and use text to speech too.
Subscribed!
Thank you, I really appreciate the support!
ElevenLabs, it'd be great to see some use cases that aren't so unbearably goofy.
Well, you know, my life is basically a collection of use cases that are unbearably goofy, so I'm just trying to stay on brand here! 😂
Man this is insane means one man can dub for so many characters in a movie even female and vice versa. Will this reduce the jobs of dubbing Artist?
It’s an exciting technolofy for sure but humans still have unique skills and creativity that are still highly valued in the industry.
Will your mic quality effect the quality of a cloned voice?
Yes, the quality of your microphone will impact the quality of a cloned voice as it serves as the input for capturing and reproducing your voice so there may be a slight quality change depending on which mic you use.
wow cool upload👍🏾
Thanks, I'm glad you enjoyed it!
Can I upload a Japanese voice actress that I like to Eleven Labs and then do that feature if I speak in Japanese (I'm fluent), or does it not have the ability to do this in other languages yet? It seems to cost money to clone a voice or make one, hmm not even sure if it's capable of doing it in Japanese...*sighs*
Yes, you can upload a voice actress of your choice to Eleven Labs. I think speech to speech only works in English right now however.
Amazing!
Thank you! I'm glad you found this new AI technology as mind blowing as I did.
Very nice .very good.
Thank you so much for your kind words!
You don't need any of those. Your voice is GLORIOUS. However, this tool is the game changer for people like me, who sound terrible, but can use the speech to speech stuff to run a faceless channel. CZcams would surely approve monetization to these voice recording because they sound so human.
Thank you so much for your kind words! I'm thrilled to hear that you're excited about the Speech to Speech feature and I completely agree, it's definitely a game changer for creators looking to run faceless channels and enhance their content with humanlike voices.
Speech to speech does show when I sign up
You'll need to sign up for the alpha to access the speech to speech feature.
Thank you, I'm able to do
Wow!! Can you do Speech to Speech into another language?!
No, I tried it and it’s a little but wonky although I’m sure it will improve.
It's great BUT it's limited to 24 seconds for now. Look forward to the day when that limit is lifted.
You're absolutely right, the 24 seconds will just be a barrier now. Still it’s enough to voice quick scripts. If it could be over 30 seconds we could voice an ad!
Muito bom 😂❤
Thank you for your positive feedback!
Vielen danke
Merci for your positive feedback about the ElevenLabs Speech to Speech feature!
@Mike Russell : AI is getting way out of hand. I will only use if I have time restraints on the project. Music Radio Creative is all I ever need.
Understood, traditional audio skills and tools still have an important place!
It's been incredible so far. Our wives are going to hate us. Can't believe how well it generates whispers. That blew my mind!
_Try 'Trent Ashford' in the voices. There are two. You know them both, trust me!_
I'm glad you're enjoying the latest AI feature from ElevenLabs! Must try a whisper soon 😉
Круто!
Thank you! Glad you liked it!
how to add accent in recorded voice--> English American accent to English German accent
Not available right now as far as I know.
As a Texan, that accent was... definitely not right. But I can't pull off a British accent for the life of me, so...
No worries, accents can be tough to nail - practice makes perfect!
We can sing?
Give it a try. Would love to try it out!
Yikes. The voice used for "I'm from San Francisco where everything is magical," seemed a bit too stereotyped.. Be careful with that
I appreciate your feedback and will definitely keep that in mind for future videos. Thank you for watching and sharing your thoughts!
Great videos! But, my oh my, your American accent is super insulting lol.
I'm sorry if my American accent is not to your liking. Won’t give up the day job!
Your US accent was nearly as bad and offensive as my Australian accent. Let's just agree to leave each other's accents alone. We don't need an international incedent.
Thank you for your feedback and apologies for my awful US accent!
That San Francisco accent was awful!
Thanks for the feedback! I'll keep that in mind for future videos.
Lol that American accent was terrible!
You know it. If only AI could help me with that!
I was both impressed and slightly disappointed. My own voice has a British accent and I struggle to do other accents... but my expectations of the AI was that it would use the target AI voice accent. Instead, the American females voices now sound like my British voice and a little raspy too. Watching your video, the tricks seems to be to mimmick the target voice... but not everybody has this talent. Trying different AI voices I realised the differences between them were now a lot less and it was more like the entire cast of Harry Potter.
I wish there was a way to control how much it sounds like your own voice vs the target voice. I feel the AI can be improved here... identifying the stresses and intonations but preserving the target voice and accent. But for a first attempt this is impressive technology.
Thanks for sharing your feedback! It's great to hear your thoughts on the AI feature - hopefully future updates will address these nuances for a more tailored experience.