Craziest Vocal A.I. Yet: Dreamtonics Vocoflex with Synthesizer V
HTML-код
- Опубликовано: 22 окт 2024
- Dreamtonics has been at the forefront of vocal synthesis with their Synthesizer V, but now, they've gone even further: The new Vocoflex offers a realtime A.I.-inside of a native plugin-that can radially resynthesize and remap real vocal performances with low latency and shockingly persuasive results.
Check it out here: dreamtonics.co...
►🎚Mixing Breakthroughs: mixingbreakthr...
►🎛Compression Breakthroughs: compressionbre...
►🎧 EQ Breakthroughs: EQBreakthrough...
►🔊Mastering Demystified: MasteringDemys...
►✅Become a channel member here: / @sonicscoop
Big thanks to Dreamtonics for supporting this channel. Get a special discounted intro price at dreamtonics.co...
The future is here. Dreamtonics has been at the forefront of vocal synthesis already with their Synthesizer V, which turns text and MIDI into convincing vocal performances. But now, they've gone even further: Their new Vocoflex offers a realtime A.I.-inside of a native plugin-that can radially resynthesize and remap real vocal performances with low latency and shockingly persuasive results.
Dreamtonics' Vocoflex can resynthesize vocals in real time, at low latency, inside of your DAW using native processing power. You can create your own vocal models quickly and easily, select or combine voices from a variety of presets, or create new vocal textures with hexidecimal code. It even works in a variety of languages.
Love or hate A.I. and vocal synthesis, you should know what's out there. This is what is happening NOW.
Check it out here: dreamtonics.co...
►See free audio tutorial videos with Justin here:
• Justin Colletti Videos...
►Get the free mastering workshop:
sonicscoop.com...
►Get the free mixing workshop:
sonicscoop.com...
►Win free stuff at
sonicscoop.com...
► ✝️ Subscribe to the podcast or leave a rating and review here:
Apple: podcasts.apple...
Spotify: open.spotify.c...
#musicproducer #dreamtonics #vocoflex #synthesizerv #aivocals #ai #autotune #mixingengineer #musicmixing #audiomixing #electronicproducer #audioengineer #justincolletti #sonicscoop #compression #eq #mixbus #mixingbreakthroughs #compressionbreakthroughs #masteringdemystified #mixingengineer #audioengineer #audioeducation
Have you tried Vocoflex or Synthesizer V? How do you use it?
Big thanks to Dreamtonics for supporting the channel! Get a special intro discount on Vocoflex at dreamtonics.com/vocoflex/?Blaunch
people might say it's wrong to do this things. but to the producers who doesn't have singing gift, it will be a life changer. I hope they will develope Korean soon.
Absolutly, i can't rely on my voice, so it's the only way to compose something solid because i'm hearing how the vocals harmonies sound with a 95% 'real' voice
exactly, vocaloid producers have been using similar tools as Synthsizer V for many years now and they created great songs with it. As a musician who cannot sing myself I am really happy this exists. I never saw people complaining about companies bringing out VST with full orchestras replacing cello players because everyone understood what an advantage that means for home or small studio musicians.
And also music made with vocaloid and synthV also have originated work for singers who like to sing vocaloid songs with their real voices (utaite) and I never saw an AI complaining about how humans would take their jobs away ;)
This stuff will make live music far more valuable. Authenticity is a big part of the connection we feel to music. If we are constantly skeptical of what we are hearing recorded then actually experiencing it live will be more desirable.
I 100% agree.
Live music will always be a niche thing.
Live music earns musicians more money today than recorded music does.
Of course, people will go to fewer concerts than they listen to songs at home. One is easier than the other :-)
But as far as the capacity to earn money, and to make lifelong experiences and connections, live music wins.
And this is coming from a guy who loves recorded music and recording music :-)
I think this is a use tool for vocal training and performance, thank you for showing the power of this VST
Absolutely. Happy to do it!
thanks for the Telastar story, I looked it up and you are correct, powerful perspective. Changed my life by giving more confidence with these tools
Awesome to hear!
for me this tech is what I have been dying for for a few years. I started making music with vst instruments, and wants to sing my own lyrics but can't because i live in flat with neighbours. and studio's are expensive. Now, I can go and record my own singing and train my voice into it, and then blend my voice with a touch of another singers harsh voice. and now just write the lyrics and have the software mimic my voice and lyrics without paying for studio time and without bothering anyone. finnally I can finish my tracks.
Can you get your own voice into Synth V?
I have played up till 24 min in, and conclude without looking further that while its great new technology, it has a seriously long way to go before it sounds natural. This sounds very synthetic. The price is very high for such a experimental plugin even with the early adoption offer. Nice starting point, but has some way to go.
Around the 24 min mark you are hearing the lowest latency setting, which sounds a bit grainy. Further on I ask him to switch back to the higher latency modes which sound dramatically better.
@@SonicScoop Thanks for your reply. I had watched other videos on this plugin. If it can indeed produce realistic tonation changes then videos need to show it at its best in first few min. Dreamsonic own video does a inadequate job of using poor examples in their promo video. I own their Synth Pro, it's amazing.
Wow!
Agreed!
That’s my compa ❤
What's a "compa"?
@@SonicScoop Apparently Mexican slang for "buddy"
38:27: this is like a more-advanced Auto-Tune. Very interesting. Thank you for showcasing some possibilities.
And it seems most people define musicianship as musical skill you can do naturally, but will feel like it's "dishonest" for you to present musical ideas that are greater than your technical playing abilities; composers can't play the dozens of orchestral instruments it takes to perform the music they wrote, but I digress.
At least this tool will help songwriters to have their music actively listened to. People perfer their music with vocals, regardless of how it's made.
can we train the ai with our own voice possibly? or we can blend our own together?
Yes and yes.
Very interesting software. Some points though. From what I've read the output gets watermarked (you can't hear it), probably to prevent you from using voices you don't have rights to. Also I wonder how that would work with royalty free vocal samples. The specific sample is royalty free, but making a voice out of it might not be .... They have a registration process that involves uploading a copy of your passport or ID, which is a first for me. Also Synthesizer V comes with only one voice when shipped, which isn't much, and it's a voice oriented on the Asian market. Just some points to think about.
You can buy one of the bundles which comes with multiple male and female voices. Or you can buy additional voices as needed if you bought a copy with only one voices…
I'm aware of the Bundles but I think the program itself should offer more than just one voice. Services like Ace studio or Audime give you over 40 voices to choose from. I know they are subscription based but that gives you the advantage of being able to cancel at any time. The combination with Vocoflex is interesting though. All of this is thinking aloud. Trying to decide where I should spend my money to overcome the handicap of the voice I was born with... 😂@@BoAstrup
Hello, does it only work when uploading files? Can't you use it to sing live in real time? thank you so much
Yes, you can sing in real time with it :-)
If you import a real vocal and make changes, can you save the new voice that you have created with your own voice? I downloaded the Vocalflex manual but did not see anything here on the video or in the manual that gave instructions........ Also I read below that you have to submit a ID. I started the ID process today and will not purchase until I am approved and a code is sent...
Edit... I did find that all created voices can be saved in the PRESET section from understanding. I assume that if you started another project, you could load a Saved voice and use again? I just want to make sure this PLUGIN tool is going to be useful before purchase....I would assume if a Vocalist was recorded 20 years ago with a older MIC or not so great conditions, the old vocal could be inserted and a new modern day AI version of that vocal (voice) could be re created??
Wholly guacamole. 😱
INDEED.
And what song will it bring? 😮😮😮
Whatever song you write!
If you buy vocoflex why do you need synth v?
They are two totally different tools. Synth V creates vocals from scratch using MIDI and text. It’s a vocal synthesizer.
Vocoflex changes the timbre of an existing vocal performance (real or synthesized) into a different voice.
Hope that helps!
-Justin
funziona in italiano?
In order to purchase this they require passport scan/upload or drivers license . 😮 Is it just me, or is that crazy!!?
I’d imagine that has to do with government regulation around AI in certain countries. If they didn’t have to take that information, I’m sure they’d prefer not to have the barrier to entry. Why would they want it in place?
Will it work in Spanish?
YES.
@@SonicScoop I saw the entire session and saw it in action. I’m still waiting for approval to purchase the product. I never experienced anything like it with a purchase.
32:40 !!!!!!!!!!!!
hmmm interesting
Agreed!
@@SonicScoop another thing I've got to learn now... :(
What king of microphone he said is best? Dynamic or phase?
Dynamic.
(I think you meant condenser? I’ve never heard of a “phase” mic.)
no Russian language support? (or any other slavic)
I’m not sure that there is. Are you aware of any similar programs that have it?
Vocoflex should work OK, but not optimally across languages I believe.
Synthesizer V could be harder. You could try typing phonetically in another language, but there would be a lot of trial and error and the results probably wouldn’t be ideal.
I could just use all my own sample singers in voco, skip Snyth v because of scrutiny and use audimee midi notes lyrics plug-in and use vocoflex for my voices live but with my owned voice samples for the model and voila! and avoid the scrutiny right?
What’s better about synth v vs audimee with its midi editor now and also vs audimee and melodyne or autotune? I understand vocoflex is the only one live I know of and the model.
Synth V is a totally different type of product. It is a vocal synthesizer that creates vocals from scratch using text and midi.
Vocoflex and Audimee are more alike. Vocoflex has the advantage of being native in your DAW and having a realtime mode.
Audimee has the tuner, pitch editor and automatic harmonies.
I would love to see the dreamtonics guy sing his best with best model for Spanish and than best model for English choosing cover songs for each that matches his voice 51% minimum his voice so we can see if it improves a soso singer? And by how much so let’s see end results than edit reverb compress and let’s bypass a b test? Two windows hear real and than hear vocoflex. Ha vocoflex should sell a course class how to train vocals phrases to match your timbre to ways of singing vs your 49% vocoflex mix.
Where can we hear the voices?
but they require you to give them your legal ID like a passport or driver's license!!
I didn’t know that. First I’ve heard of it, and I didn’t have to go through that process when I did it. Probably due to new international government regulations around AI?
I paid for a product, but I didn't provide my ID. As a result, they kept my money and didn't give me the option to download the product without risking identity theft. I really want the product, but I'm concerned about the possibility of their system getting hacked and my personal information being compromised.
You’ll probably have to take that up with whatever governments are requiring this information for those kind of AI technology.
If you’re not comfortable with it, then it may not be the product for you!
Such ID might only be required in certain territories as it’s the first I’ve heard of it.
What country are you in?
I posted comments about my negative experience because I am in the USA. Is iligal here I checked.Initially, they refused to return my money, and I declined to provide them with my passport. After several weeks of posting my complaints about them, they finally returned my money. It's disappointing because I liked the company's idea, but some risks are just not worth it.
They took the passport from.a.friend of mine and than did not send the links to download
Feels like identity theft to me!!
There may be ID requirements for AI tools on sone countries to my understanding.
If he can’t find his download link, have him reach out to support.
They are a legit company that has been around for years :-)
Vocoflex has same editable midi editor as synth v right?
Nope, they are two totally separate products. Vocoflex is just voice transformation.
Can I use Vocoflex to bait phony tech support scammers?
I should hope so!
There is no discount on that code it’s 199 just for vocoflex, I saw 199 for both already? Where code
There was an introductory discount when it first came out, but it may be over now. You can try the code SONICSCOOP at checkout, but I don’t know if it would still be functioning.
Hello Justin, to this year in mix con u could make the video of engineer Jaso with the NBA song Vette Motor?
He was considering doing that one, but I think he chose another. You can however, ask him any questions you want about that one in the live Q&A afterward!
If you sing off key can you map like span plug-in the wave oscillator to see and overlap to replace differences vs top singer mathematically ? Like an auto button or go to samples or splice or ai download 20 voices than vocoflex ai matches best voxengo span plugin suggested wave areas for improvement suggesting best voice for 49% for top melody harmony rhtymn nice vocals? 49% to them max so I can claim ownership in legal battle if a hit
Are vocoflex voices better than audimee?
man, its cool software, but the customer service is severely lacking. The forum link doesn't work, there doesn't seem to be a dedicated support email or anything like that (just a business inquiry email). Wish I hadn't jumped in so soon. Pretty weird. :(
Interesting! I haven't had any experience with their customer service. Maybe the usual customer service forum broke because they just released this plugin last week and weren't set up to handle the traffic? I'll ask.
@@SonicScoop there's a support email I finally found when going to one of their store pages. Just kind of embedded and hidden. heh. Thanks for all these videos you do, btw! Super cool to see how things work
Glad to be helpful! Yes, I’m told the normal support forum has crashed, probably due to high traffic from the release? I’m told you can email their support directly at support@dreamtonics.com and they’ll follow up immediately. Hope that’s helpful!
That AUTOTUNE is caused by vocoflex itself? sounds WIERD
No there in no tuning being in vocoflex. Miguel’s pitch fluctuates a bit as he’s singing I imagine it’s hard to have great pitch control when you are hearing your voice being resynthesized in real time with a little latency, and it is probably causing him to warble in pitch a bit.
At 3017 he is singing , he sounds much better? , or we hear a Synth v voice?
nope. no demo-no interrest.
There is the free Basic version with minimal features and max 3 tracks. Still great though.
sounds...
Go on...
@@SonicScoop terrible. Don’t you think?)
In the super low latency real time modes, it sounds kind of grainy. I asked him to turn that off around 20 something minute mark, and then it sounded much better again.
Also, how do you feel about the presenter singing voice will weigh into how good you think it sounds :-)
wondering why they don´t give a Demo or trial... suspicious.
I would love to see the dreamtonics guy sing his best with best model for Spanish and than best model for English choosing cover songs for each that matches his voice 51% minimum his voice so we can see if it improves a soso singer? And by how much so let’s see end results than edit reverb compress and let’s bypass a b test? Two windows hear real and than hear vocoflex. Ha
Humans are still exists!
True story! An AI likely wouldn't make that grammatical error X-D
Kidding, kidding, but yes. Human music is the best kind of music. Which is why I recommend very human ways of using these things that most may not consider.
-Justin
@@SonicScoop 😂😂😂😂😂😂😂😂😂😂😂😂😂😂😂
Hvv
What does that mean?
WTF so what we don't want to pay singers now. Sure you could do that with A.I. but why? Humans make music. It's a human thing. Tec should make life better for everyone, not make people redundant. Maybe I'm old.
There are a lot of ways this could be used BY singers. For example:
A singer can use Synthesizer V to extract MIDI data from their vocal performance to aid in composition, or transcription, or to use their voice to trigger a different instrument, or even use it to write multiple harmonies, or even an entire orchestral score, just by using their voice
A singer can use Vocoflex to remap their vocal performances to different timbres to create persuasive harmonies for their own voice using multiple voice textures instead of just one.
That's just scratching the surface. We talk about a lot of use cases like this in the stream.
-Justin
I'm interested in AI to do guide vocals for my songwriting. Just a better way for terrible singers like me to convey what I want from the actual human singer.
You think most people that are using this are going to be hiring singers? Most of the top singers now s days can’t really sing that great anyway. Some of the best singers I have seen usually are singing better than label paid people here on RUclips.
I don’t care how someone makes music as long as it’s good. I do get tired of hearing these ai moral policing comments tho.
@@ladyville3 Put it this way if anybody is using this in place of singers they won't make any money on sales or streams.
Ok voice disguise is fine but not serious. 😅what about changing bad singer to good? What’s the ten bullet point workflow all plugins needed to buy is ok. Make free alternative list too please. I have melodyne. And many more. Workflow?
I am worried about getting rights for non-pc hip hop or Edm songs. Do the woke manifesto commanders tell me yes or no or how to avoid the pc music police ? If I merge my voice 51% and synth 5 voice 49% me and use for singing live how does your politically correct approval board get involved? The woke music policy what are their rules? For the standard $199 I can use all voices or woke team checks those too before I put on Spotify distribution? So if I don’t buy the extra voices than no royalty woke police promblems? All royalty free on standard and extra voices? How long does it take for approvals and what are chances? Than it’s royalty free forever? What if I use paid voice synth v and run it through voco and doesn’t sound like them anymore? This is good for bad singer djs like me to sing their own songs live and skip singers and Twitch and kick live stream or Tik toks live rocks!?
It’s too expensive.
Nothing crazy about this. Very fed up with all this AI nonsense. Luckily the hype is about to blow over, in my opinion. It still sounds absolutely bland and boring and lifeless man, and it always will, be honest. I know you hear the same I hear, you're a very smart and musically trained man.
I just don't see the point, it would never ever fool me. I cannot see any beneficial aspect or advantage to this. Why on earth would anyone use this seriously, besides as a one time gimmick?
Auto-tune and talk boxes and vocoders at least have their own character. They sound synthetic, but in and interesting way. They don't essentially try to sound real, per se (auto tune and melodyne stuff tries to sound real sometimes, yes ok). But in general they are designed to open up new ways and new colors and lead to funky stuff. In the same way I play piano, but I love playing synthesizers as well. Synthesizers aren't generally made to sound like a piano, they're something else (obviously exceptions exist).
This tries to sound real. But I have trained ears and good speakers man. Frank Sinatra and Ella Fitzgerald turn around in their graves. Any good dedicated vocalist sounds so much better and more captivating. Any vocalist that is having a good time and is enjoying the singing sounds better. Heck I'd rather listen to a 10 year old kid from the music school at the music school concert that is talented and enthusiastic than to this.
I really don't understand why you're promoting all this AI tools buzz stuff lately. I respect you immensely for all the mixing and mastering video's, but I really don't understand why you do this. They must pay a pretty penny
It's interesting, at almost the same moment you wrote this to me, I got an email from another regular viewer who wrote:
"Justin! Bro.. that synthesizer V system is wild!!!! I downloaded after watching your podcast yesterday and instantly made a banger track out if it... What? Keep up the good work man."
So there's definitely a lot of different perspective on this.
I do talk a lot about use cases that could be interesting to a truly musical person. They may not be obvious at first.
For instance, someone who sings well could use Synthesizer V in reverse to instantly transcribe their vocal parts to music notation, or have those vocal parts be re-played by a totally different instrument.
A vocalist could potentially create a whole symphony work by mouth, even if they're not great with music notation!
Alternately, a good vocalist could use something like Vocoflex to create stacked harmonies with different vocal timbres, instead of a just a single timbre.
In this case, using the "artifice" of vocal re-synthesis would make the artifice of stacking a single human voice less noticeable, and it would sound more like an actual background section.
As for how lifelike and compelling the performances are, Vocoflex only changes the timbre of the performance, so how lifelike and compelling the performance is still comes down to the vocalist.
Synthesizer V is where you could make the argument of the voice not being as expressive and emotionally resonant as a good singer... but duh! I agree.
That said, Synthesizer V could DEFINITELY be more compelling than a bad signer, and there are a LOT of those X-D
That said, my preferred applications for it would be as a writing and demoing tool, or using it in reverse to instantly transcribe or transform sung vocal parts.
Whenever I approach this topic, I try to bring a LOT of nuance to it and look for use cases the average person might not think of, but I guess that's easy to miss if you don't watch the whole thing... which you probably wouldn't want to if you dislike the topic out of the gate! So I get it :-)
That said, I don't think Vocoflex sounds that great in its lowest latency realtime mode settings. It's too grainy that way. I had to ask him to turn that mode off.
But, in its normal higher latency settings, I think the timbre change works well enough to use for background harmonies from a single singer, in the context of a full mix, for SURE.
Hope that helps!
@@SonicScoop Thank you for your response, I appreciate you replying and things you say a lot
How about for me as a vocal arranger who can no longer rely on my voice dude to answer autoimmune disease who still wants to sell his a cappella arrangements and vocal part splits? This is a huge tool for that. Synth V and this and I have all vocal part ranges covered which even when I could sing, I couldn't sing low low bass up to high soprano 1. This is a huge blessing for me!
Unfortunately, the support is also a disaster. No one explains how to make this v studio sing. But now I have beautiful sine tones for 200 euros. Really, really crap
I’m told the normal support forum has crashed, probably due to high traffic from the release? I’m told you can email their support directly at support@dreamtonics.com and they’ll follow up immediately.
I just bought it. I'm very disappointed, there are no voices. Just a horrible sine tone. By the way, the installation wasn't exactly user-friendly.I really hope that they help with support. Otherwise I've just wasted 200 euros. >Apparently there should be at least one voice included. The voices in your video apparently have to be bought separately at great expense. You should know that before you buy.
I'm looking at the pricing now for you. I believe the main bundles for the Synthesizer V are $279 currently on sale for the synth with two synthesizer voices.
Vocoflex expands that to basically an infinite number of voices through resynthesis for an intro sale price of $159. (Or it can be used on its own.)
It looks like you could instead get Synthesizer V AND Vocoflex together for an intro sale price of just $199... But with that super inexpensive version, you'd have to buy Synthesizer V voices separately.
However, you could spend $60-$70 for a single voice for Synthesizer V, so for the same $260-$270 total, and you'd have a single Synthesizer V voice, which you could then morph into a potentially infinite number of voices for the same price as the basic $279 two-voice bundle.
(I suppose some people want to use Synthesizer V mostly to turn vocal performances into text and MIDI rather than the other way around, so I guess that's why they have a less expensive option?)
For anyone reading this later on, these sale prices may not be relevant. The $279 bundle for instance seems to list for $349 normally.
Maybe they could be even more clear about the way they package the bundles, but I seemed to figure it out after a few moments.
I hope that helps!
-Justin