I use SynthV and with their new engine update their voice Felicia can straight up do a whitney houston / bonny tyler power ballad and sound natural. I think it's pretty cool. I would love to have a million dollar studio - and a million dollar singer - but alas, sometimes this will have to do.
@@Arcessitor Most singers can not sing like Whitney or Bonnie!! Listen to the songs on the Billboard top 10 or on the radio or in any club - almost none of these singers can sing like they used to. Plus every single one is so autotuned they already sound like AI. LOL - the industry did it to themselves.
I do understand your ethical concern about putting vocalists out of work, and I appreciate that you're looking at it from an ethical angle, but how is this different from instrumental sample libraries? If I use a guitar VST to get a guitar track on my song, am I not putting a guitarist out of work in the same way?
The vocalists get paid for licensing their voices out for this, so they're actually making money WITHOUT DOING ANY WORK -- a win/win for everybody, if you ask me.
In Vocaloid every new note has perfectly flat vibrato and the timing is perfect. you have to manually add vibrato and tweak note properties (timing, breathiness etc.) to make it sound more realistic. it's the perfect vocal synth when you need an artificially tight vocal part. Modern vocal synthesizers like Synthesizer V and Ace Studio seem to take the opposite approach, where the application immediatly suggests expressions, especially in regard to vibrato. But you still have to tweak them a lot to get them to sound good, so you don't really get a workflow benefit from the supposed shortcut. This is why I like the simpler minded, and often non-AI-based, vocal synths more. But I recently got an NFR for vocoflex, just like I talked about in your chat when this video had its premiere and I noticed AI vocal resynthesis from existing vocal input makes so much more sense than vocal synthesis with AI. Cause when you only input text the AI has to guess what kinda expression fits to your lyrics, but without even knowing what language and our cultural background is. But when you perform a vocal take with your own mouth you automatically put an expression into it that you believe to be right for the text. When this expression flows through Vocoflex it is being converted into a comparable expression of the other voice and you immediatly get a convincing and fun sound. I wasn't much of a singer for years but now I record new vocal parts everyday and try to improve on my vocal skills to supercharge the potential of the plugin as good as possible. It's kinda challenging. That's what I like about it. Using AI for the sake of AI is stupid. And using AI to make a lot of cash is stupid, too. But using AI to push yourself to be productive in doing the things you like, that's nice.
The question of whether or not AI is destined to deprive live musicians from their jobs, goes way beyond singing alone. But as a product (compared to other products doing the same thing today) I think it is pretty impressive. I hope it remains a tool, and nothing more.
I don't sing but I can sing at least that well. Maybe it's just me but it sounds like somebody lying in bed trying to quietly sing along with the radio so he doesn't wake anybody up until he falls asleep. Admittedly, maybe I just don't get yet. Just sounded so dead to me.
I like both Synthesizer V and Ace Studio but as you say, it takes a lot of time to get it good. Even more so if adding VoiceAI by SoundID on top to change voice.
So I am or was very close to signing up to Ace Studio. I think you should have started by playing your track with the midi vocal melody you had already created. The comparison would have been helpful. I also think the vocal model you used was terrible and sounded flat and dull. I would like to see you send your files to Ace Studio and ti see how good they could make it. Ps wrote this at 28 mins.
Yeah, I agree. He just doesn't understand how to program it correctly (very flat results, and intentional, perhaps) ... I have heard/seen much better examples. ACE is not meant to replace real life singers, it is made for those with low resources & that have time on their hands and are willing to coach & put in the extra effort into programming it to sound like how they want it to... A review that is very biased & extremely resistant to the tools available to those that don't have the same resources & $$$ available to them. I don't know any singers and probably live too far away from anywhere, where it is feasible to get them here to record (paid or not), so this TOOL is a God sent for me and fits well into what I do in generally. Personally, I think ACE is a great tool for prototyping etc and if you're just doing EDM and the like, coming up with great hooks and ad libs is easy-peasy with ACE... Just saying 😁
Synthesizer V is amazing ... I use it to QUICKLY get lyrics "text to singing" and help with harmonies BEFORE I record REAL vocals. It is a game-changer for me to experiment with a variety of ideas, emotions, and styles. It can be EASILY tuned to create any type of vocal you meed. It's literally the most important plugin I own (besides my DAW) to get a song exactly the way I want it. Here are a couple of examples: Celine Dion - th-cam.com/video/T_VtD19t9ks/w-d-xo.htmlfeature=shared&t=200 and then this one too from Aladdin - th-cam.com/video/1zKrxuQ_4Bw/w-d-xo.htmlfeature=shared&t=108
Hi, this is a really good review. As for practical application, I also think it takes too much time to bring the material to an acceptable result. Special thanks for your civic position regarding AI, products like this have nothing to do with technical progress other than the goal of making money by depriving many people of their jobs.
Couldn’t you just take the AI track and import it into a track and manually manipulate the pitch with a third party plug in? The rendering was the red flag. That takes way too long
There is no reason to worry about this aspect of AI. The lion's share of money being made in music these days is live performance based. No one is paying even $50 to watch a computer making music and most people are not influenced to buy a product based on AI saying its good. Ace studio despite the hype of other TH-camr sounds like garbage . I kept trying to change singers to get something good and then realized they are all bad. A few give you hope when they are singing long notes without words but once there are words it sounds tinny and awkward.
AI vocals will never replace real singers but I do think there's a place for AI vocals in modern productions. However, if you're going to try one out then I'd suggest Synth V should be your choice rather than this one and try one of the better vocal libraries like Natalie, Solaria, Felicia (female). It's kind of incredible.
I tested ACEStudio and didn’t like its workflow (each change takes time because of cloud based processing) and price. I prefer to make AI voice models myself (RVC) and change my vocals offline. It’s very important to know that you need a good vocal material if you want good results. For example after recording my vocals I had to use melodyne in a quite excessive way in order to chage my sang vocals to a good sounding female voice.
A little too emotional for a technology review. This tool will only be able to replace real singers where none are needed, such as in advertising jingles etc.. Otherwise it is just a tool to experiment and try out new ideas. Except for the rendering times that kill the workflow, it's an interesting toy.
@@FoliaSoundI think we will accept AI, similar to how we accepted synthetic sounds or sampling. AI cannot replace the joy of making music together. As long as there are people who want to make music together, there will also be people who honor this. What is important is how we deal with it; creativity and a sense of community must continue to be encouraged. If we don't do that, I don't see a bright future with or without AI. Greetings, PS: nice channel
What a tedious way of working, good grief. If you want to sing , just learn how to sing folks.This is goig to cost you WAY too much time, which in turn seriously effects workflow. Respect for putting the vocals in the video man, I think it was perfect as an example. Too many companies are baiting young people into the illusion of being a producer, vocalist or artist. Just the mail of the company, explaining what the software is good for, is totally different then what they are selling online.... So, they are aware of fooling people. Just hopping on the AI-train. It also shows what people are willing to do for money, at the expense of your wasted energy.
I don't think you'd laugh at this track - th-cam.com/video/DlAIObTIAFA/w-d-xo.html - I understand people's reactions but I seriously think it can be compared to the age of steam where people thought all kinds of strange things and concepts about it, change will never stop, if there is one thing you can guarantee in this world and that is change.
Frankly? I would, just a bit. I know it's AI and it's meaningless in terms of human experience, human emotions or artistic intent behind it. Sorry to disappoint you :(
@@FoliaSound Why should I be disappointed? These are my own lyrics and video not AI so the artistic intent is my own, why did you not hand write your comment and send it to me by post? Oh! you used technology - th-cam.com/video/EZBRuzR9ETM/w-d-xo.html My video and lyrics.
Did you write this song? Did you sing it? Do you know how much human depth, most beautiful craft is there, when writing and performing music, do you know how profoundly satisfying and rewarding this is?
![Chris Langan - The Interview THEY Didn't Want You To See - CTMU [Full Version; Timestamps]](http://i.ytimg.com/vi/9miVG2xT5jY/mqdefault.jpg)
Heeey! What do you think about Ace Studio? Acheved same stellar results as me? :D Not using AI at all in your music making? Lemme know down below!
I use SynthV and with their new engine update their voice Felicia can straight up do a whitney houston / bonny tyler power ballad and sound natural. I think it's pretty cool. I would love to have a million dollar studio - and a million dollar singer - but alas, sometimes this will have to do.
@@Arcessitor Most singers can not sing like Whitney or Bonnie!! Listen to the songs on the Billboard top 10 or on the radio or in any club - almost none of these singers can sing like they used to. Plus every single one is so autotuned they already sound like AI. LOL - the industry did it to themselves.
May be it's just about knowledge
th-cam.com/video/76CTCMEJbtg/w-d-xo.html
I do understand your ethical concern about putting vocalists out of work, and I appreciate that you're looking at it from an ethical angle, but how is this different from instrumental sample libraries? If I use a guitar VST to get a guitar track on my song, am I not putting a guitarist out of work in the same way?
The vocalists get paid for licensing their voices out for this, so they're actually making money WITHOUT DOING ANY WORK -- a win/win for everybody, if you ask me.
Embrace the change and stay ahead of the curb. Those digging their heels will be left behind.
In Vocaloid every new note has perfectly flat vibrato and the timing is perfect. you have to manually add vibrato and tweak note properties (timing, breathiness etc.) to make it sound more realistic. it's the perfect vocal synth when you need an artificially tight vocal part. Modern vocal synthesizers like Synthesizer V and Ace Studio seem to take the opposite approach, where the application immediatly suggests expressions, especially in regard to vibrato. But you still have to tweak them a lot to get them to sound good, so you don't really get a workflow benefit from the supposed shortcut. This is why I like the simpler minded, and often non-AI-based, vocal synths more.
But I recently got an NFR for vocoflex, just like I talked about in your chat when this video had its premiere and I noticed AI vocal resynthesis from existing vocal input makes so much more sense than vocal synthesis with AI. Cause when you only input text the AI has to guess what kinda expression fits to your lyrics, but without even knowing what language and our cultural background is. But when you perform a vocal take with your own mouth you automatically put an expression into it that you believe to be right for the text. When this expression flows through Vocoflex it is being converted into a comparable expression of the other voice and you immediatly get a convincing and fun sound. I wasn't much of a singer for years but now I record new vocal parts everyday and try to improve on my vocal skills to supercharge the potential of the plugin as good as possible. It's kinda challenging. That's what I like about it. Using AI for the sake of AI is stupid. And using AI to make a lot of cash is stupid, too. But using AI to push yourself to be productive in doing the things you like, that's nice.
This is a nice summary of our chat conversation, thanks :)
The question of whether or not AI is destined to deprive live musicians from their jobs, goes way beyond singing alone. But as a product (compared to other products doing the same thing today) I think it is pretty impressive. I hope it remains a tool, and nothing more.
I don't sing but I can sing at least that well. Maybe it's just me but it sounds like somebody lying in bed trying to quietly sing along with the radio so he doesn't wake anybody up until he falls asleep. Admittedly, maybe I just don't get yet. Just sounded so dead to me.
Regarding voice overs - do you think this situation is temporary or it will stay like this and voice-over artists won't be needed?
It will get worse and worse, this is what I believe. Some of VO people will quit their jobs foor good.
Yeah, they need to train it to compose. The thing with vocals is it's an instrument. Generating any instrument performance is a giant task
This is also supposed to be used for people who dont have access to a vocalist lol
I like both Synthesizer V and Ace Studio but as you say, it takes a lot of time to get it good. Even more so if adding VoiceAI by SoundID on top to change voice.
So I am or was very close to signing up to Ace Studio. I think you should have started by playing your track with the midi vocal melody you had already created. The comparison would have been helpful. I also think the vocal model you used was terrible and sounded flat and dull. I would like to see you send your files to Ace Studio and ti see how good they could make it. Ps wrote this at 28 mins.
Yeah, I agree. He just doesn't understand how to program it correctly (very flat results, and intentional, perhaps) ... I have heard/seen much better examples. ACE is not meant to replace real life singers, it is made for those with low resources & that have time on their hands and are willing to coach & put in the extra effort into programming it to sound like how they want it to... A review that is very biased & extremely resistant to the tools available to those that don't have the same resources & $$$ available to them.
I don't know any singers and probably live too far away from anywhere, where it is feasible to get them here to record (paid or not), so this TOOL is a God sent for me and fits well into what I do in generally. Personally, I think ACE is a great tool for prototyping etc and if you're just doing EDM and the like, coming up with great hooks and ad libs is easy-peasy with ACE... Just saying 😁
Synthesizer V is amazing ... I use it to QUICKLY get lyrics "text to singing" and help with harmonies BEFORE I record REAL vocals. It is a game-changer for me to experiment with a variety of ideas, emotions, and styles. It can be EASILY tuned to create any type of vocal you meed. It's literally the most important plugin I own (besides my DAW) to get a song exactly the way I want it. Here are a couple of examples: Celine Dion - th-cam.com/video/T_VtD19t9ks/w-d-xo.htmlfeature=shared&t=200 and then this one too from Aladdin - th-cam.com/video/1zKrxuQ_4Bw/w-d-xo.htmlfeature=shared&t=108
Just listening to it. Pretty impressive, but I'm still against AI in general...
For the celiene dion one i thought it was her for like 20 seconds before I realized lol, so good
Have you tryed Mvoice One?
doesent really matter, take the generated vocals and then replace them with Audimee to actually get some emotions from much better singers :-)
Hi, this is a really good review. As for practical application, I also think it takes too much time to bring the material to an acceptable result. Special thanks for your civic position regarding AI, products like this have nothing to do with technical progress other than the goal of making money by depriving many people of their jobs.
Try synthesizer V
I will, many people mentioned it to me!
Yes, it is much better, especially the interpretations
User error? 😁🧡
Dude use it with edm music then take the sample throw it back into a DAW and mix it down cause i just think its the track you are using lol
Couldn’t you just take the AI track and import it into a track and manually manipulate the pitch with a third party plug in? The rendering was the red flag. That takes way too long
There is no reason to worry about this aspect of AI. The lion's share of money being made in music these days is live performance based. No one is paying even $50 to watch a computer making music and most people are not influenced to buy a product based on AI saying its good. Ace studio despite the hype of other TH-camr sounds like garbage . I kept trying to change singers to get something good and then realized they are all bad. A few give you hope when they are singing long notes without words but once there are words it sounds tinny and awkward.
AI vocals will never replace real singers but I do think there's a place for AI vocals in modern productions. However, if you're going to try one out then I'd suggest Synth V should be your choice rather than this one and try one of the better vocal libraries like Natalie, Solaria, Felicia (female). It's kind of incredible.
I tested ACEStudio and didn’t like its workflow (each change takes time because of cloud based processing) and price. I prefer to make AI voice models myself (RVC) and change my vocals offline. It’s very important to know that you need a good vocal material if you want good results. For example after recording my vocals I had to use melodyne in a quite excessive way in order to chage my sang vocals to a good sounding female voice.
Hahahaha.... Dear/poor Ace studio...... You've made a video against them. They will curse the day when they wrote their first email to you : )
They're ready for Eurovision lol.
Absolutely! :D
A little too emotional for a technology review. This tool will only be able to replace real singers where none are needed, such as in advertising jingles etc.. Otherwise it is just a tool to experiment and try out new ideas. Except for the rendering times that kill the workflow, it's an interesting toy.
I'm fine with going emotional, because all this AI thing is not just technological - it's social, economical, existential in every sense. Greetings!
@@FoliaSoundI think we will accept AI, similar to how we accepted synthetic sounds or sampling. AI cannot replace the joy of making music together. As long as there are people who want to make music together, there will also be people who honor this. What is important is how we deal with it; creativity and a sense of community must continue to be encouraged. If we don't do that, I don't see a bright future with or without AI. Greetings, PS: nice channel
Ace studio AI is very limited. It's okay for soft rock and pop, but the available voices can't replicate a powerful hard rock or blues singer.
chipmonks on crack..........Great Show
Thanks, James! Yep, pretty much :)
Wybitne raczydło xD już wolę kupić jakieś wokale :D
To samo u mnie!
What a tedious way of working, good grief. If you want to sing , just learn how to sing folks.This is goig to cost you WAY too much time, which in turn seriously effects workflow.
Respect for putting the vocals in the video man, I think it was perfect as an example.
Too many companies are baiting young people into the illusion of being a producer, vocalist or artist.
Just the mail of the company, explaining what the software is good for, is totally different then what they are selling online.... So, they are aware of fooling people. Just hopping on the AI-train. It also shows what people are willing to do for money, at the expense of your wasted energy.
Very well said, thanks!
@@FoliaSound ☺
Sorry to say, but its the worst review video I have ever seen.
Ok.
calm down with your 13 thousand subscribers bro
Helllooo@ What do you mean exactly? :) I get emotional sometimes, I'm fine with that.
I don't think you'd laugh at this track - th-cam.com/video/DlAIObTIAFA/w-d-xo.html - I understand people's reactions but I seriously think it can be compared to the age of steam where people thought all kinds of strange things and concepts about it, change will never stop, if there is one thing you can guarantee in this world and that is change.
Frankly? I would, just a bit. I know it's AI and it's meaningless in terms of human experience, human emotions or artistic intent behind it. Sorry to disappoint you :(
@@FoliaSound Why should I be disappointed? These are my own lyrics and video not AI so the artistic intent is my own, why did you not hand write your comment and send it to me by post? Oh! you used technology - th-cam.com/video/EZBRuzR9ETM/w-d-xo.html My video and lyrics.
Did you write this song? Did you sing it? Do you know how much human depth, most beautiful craft is there, when writing and performing music, do you know how profoundly satisfying and rewarding this is?
If not, ditch this AI and start learning to create your own music for real, see what kind of a proces this is, thank me later. Cheers!
@@FoliaSoundYou know what, I've been thinking about this, and when all is done, you just gotta have fun in your life.