GUMI VOCALOID vs GUMI SynthV (Comparison) 

I think Yamaha has Vocaloid 6 (as well as 5 when that was current) priced the way it is (and was) because it comes with multiple voicebanks included. Factoring those voicebanks in, Vocaloid 6 (as well as 5) actually costs LESS than Synthesizer V Studio Pro would once you pay for an equal number of voicebanks, which are priced roughly the same for each software when you buy them individually. (I'm ignoring the free LITE voicebanks for SynthV for now.) Yes, if someone brand new to singing synthesis ONLY cared about obtaining and using Gumi, then Synthesizer V Studio Pro and a paid AI voicebank is cheaper; but it's the bundle of voicebanks included with the editor that make Vocaloid a better deal. The very significant problem is that not everyone wants the voicebanks included with Vocaloid, and not just as in the previous example. As this video demonstrates, Vocaloid 6 leaves something to be desired in comparison to other current technology, and that alone is enough to put many people off of Vocaloid, even though you get more for your money. Is getting more voicebanks just because they're cheaper really worth it if they're not as good or not the voices you want? And THEN you factor in that Synthesizer V is getting far more frequent new voicebanks, more frequent updates which often add entire new features, AND the ability to get LITE voicebanks. Yes, I absolutely do agree that Yamaha needs to step up their game, but I don't think their price is the problem.

Yamaha has needed to step up their game since VOCALOID4, honestly. VOCALOID4 should've been a free update to VOCALOID3. VOCALOID5 was utter garbage and I wish I could get my money back. Dreamtonics keeps releasing free updates to Studio Pro that Yamaha would never implement, and if they ever did something even half as good they'd just call it a new version and have you buy it again. Honestly this comparison is a perfect example of how little effort Yamaha puts into VOCALOID. SynthV started off as a single uni student working on a thing for UTAU, and now it's a small company that vastly outperforms a multi-billion dollar corporation with decades of experience in the same market. The SynthV Gumi model is trained on the same samples that the V6 model is trained on. Yet V6 is just muffled engine noise.

@@Dojan5 I agree with your second paragraph completely. My gripe is that Yamaha will show glimpses of major potential for Vocaloid, and then either not deliver on them at all, or give out something half-assed, possibly charging people for it again. Dreamtonics on the other hand consistently comes out with new features for Synthesizer V Studio that don't cost anything more than what a person would've already paid, with possibly the only exception being when cross-language compatibility was added for paid AI voicebanks (since if a person didn't have Studio Pro they couldn't use that, if I understand correctly). Allow me to provide a quote from the Vocaloid Wiki as an example of my gripe: "On October 13, 2022, a DTM livestream was launched following the release of VOCALOID6. The livestream confirmed that a total of sixteen new vocals were upcoming at the time, all with their own color themes." It's been over a year since then. Aside from MAYBE Fuiro, WHERE ARE LITERALLY ANY OF THESE UPCOMING VOCALS???? Yamaha did experimentation on Vocaloid with AI back in the Vocaloid 3 or 4 era with Hibari Misora. There is a little bit of AI in Vocaloid 5 (as confirmed by Crypton on why they didn't make a Miku voicebank for Vocaloid 5), but it took until Vocaloid 6 for us to get true AI voices in Vocaloid. Oh, and may I remind you that one of the shining new features of Vocaloid 4, cross-synthesis, was immediately scrapped in Vocaloid 5. Job plugins too. Sure, Vocaloid 3 to Vocaloid 4 may not have been that big of a jump in technology aside from XSY and showing the pitch curve on screen (oh and triphthongs, which yeah, could've just been an update instead along with the pitch curve), but at least the Vocaloid 4 era was Yamaha's most successful. But to go from that to Vocaloid 5.... It's sad, really.

​@@BiohazardP the price is a problem because they don't let you buy the standalone editor. it's not very accessible

​@@BiohazardP vocaloid 6 is $225 without tax that's just not reasonable for the average person Edit: also the basic version of synth v is free as well as some lite versions of some voice banks while the pro version is basically $90 which is way more affordable than vocaloid...but when you look up synth v pro you see either variations of these two search suggestions "synth v pro crack" or "synth v pro free download" which I think is funny

😂😅😊 yes! Now she sounded like an actual human!

Fr! She’s my fave Vocaloid! I love Megumi Nakajima’s voice in rl too, so it’s nice to hear he lovely vocals shine through with SynthV! 🙏✨

The team behind Gumi said they would still work on updates for the vocaloid software, Synth V Gumi is just another option to get Gumi

@@theamazingkaleb3077 a lot of vocaloids can sound human you just have to tune them.

@@theamazingkaleb3077 if you are listening to a synthetic vocal and wanting it to sound like a human..... you're embarking on a fool's errand. The appeal of vocal synths is that they _aren't_ human.

I'm not refuting the bullet points you got here, but "company X doesn't have as many technological features, therefore they lack passion" is a bollocks argument. Competence and passion are not coterminous. I'm not even defending Vocaloid either. As an outsider, I'd assume the management is riding off that name recognition. But I'd also assume there's some very passionate engineers at Vocaloid.

I'm not defending vocaloid either but Gumi V6 can sing in English too And SynthV allows all AI vocals sing in Japanese, Cantonese, Mandarin, English and now Spanish no matter their native language. Spanish is very new though, recently implemented with SAROS' release and mostly developed by Eclipsed Sounds

@LucyLerma Vocaloid6's crosslingual feature is more rudimentary than SynthV's. GUMI will be able to sing in Mandarin and Cantonese on SynthV, but the Chinese crosslanguage often leaves a lot to be desired due to how complex the sounds are. I only considered whether a voice can sing in a language fluently, so I did not mention them. Spanish is indeed in beta, and it's going to take time for it to be implemented onto other voicebanks, so I did not mention it either.

Just praying for Kiyoteru to be released in SV 🥺.

but pitch tuning freehand on synthv is way much easier.@@raumfahreturschutze

fr

I find it weird that people complain that a voice isn't as robotic in a demo that's supposed to show you the capabilities of the voice If you've EVER used a vocal synthesis software, making robotic outputs is VERY easy and not generally the only scope we'd want to be able to use the voices we paid for for

It's also really different when a voicebank is obviously *meant* to be realistic and still sounds robotic. It's too robotic for realism and too realistic for being robotic. They just ended up with a vb that doesn't appeal to any particular demographic

@@-Solidwater hell, even THAT'S not true technically, cos there are a lot of people who wanna see more synths either using a blend of standard and ai tech, or ones that have realism despite being more synthetic! (like cevio ai's 1.0 vocals) it's just that yamaha fumbled so badly that they couldn't even fill a damn NICHE. i think v6 sounds a bit better now after a few updates, it has some nice vocals... but it's still embarrassing tbh

The complains make even less sense once you realize you can literally just tune a voicebank to sound robotic on SynthV. Realism is an option of many, not the only one.

I love Gumi for engine noise honestly

Fr, I love Gumi BECAUSE of this unique sound. Not even Gumi, Vocaloids/similar synths voices in general. SynthV really steps up the game but what's the point if they sounds too real? I want the gritty and robotic voices since it has much more personality@@Yan0419

If you didn't hear Meguchi/Ranka in the OG GUMI all those years ago that's on you .-. I'm actually shocked it took you this long

They don't need to ruin gumi like they did teto

@@johnhead9267 Teto was not ruined, Gumi neither.

I was thinking the same :(

I have no idea what they did but with v5 that's when the change in VBs happened. They must have a new engine or ... something? Idk V2-V4 has a nice engine noise, the voices sound smooth. Even V1 has it's charm and sounds good. But v5/6 they changed something.

⁠@@htsunmikui don’t have the technical whereabouts, but it’s because Yamaha kinda decided to kill their own product fsr. They made such profoundly stupid decisions with not just the characters but the editor that it allowed for other projects to skyrocket in both quality and popularity. When V5 editor released, i was only a novice, and yet I immediately recognized that it would be atrocious to use.

I still don’t understand why they removed the XSY fuction😭 it’s amazing, the v4 engine is amazing for fine tuning vocals and now you can’t even manually draw pitch bends in v5 or v6

v6 ui with the v4 engine is my dream……. yamaha please

@@violetutau XSY got removed because people found ways to bypass the restrictions placed by companies and make fake voicebanks using XSY, creating whole new voices out of currently existing voicebanks :/

all of this ai crap needs to stop. having a fun feature allowing people to change the voice singing a song with little effort is fun and all, but forcing EVERYONE to use a crappy feature they don't want is dumb.

⁠@@mozzapple No one tells you to use it! And the companies behind Teto, Miku, Gumi etc. still have their Vocaloid and UTAU Voicebanks up for Sale/Download. And the AI Technology which SynthV has is actually pretty good, compared to the AI Technology in V6, where it just sounds like crap. And AI in general won't stop! We are in 2023, we have arrived in a new Era, where AI will be part in our lives and it's already too late to back off from it. Synth V was one of the first Synthesizing Programs which have AI Technology included to their programs and their Voicebanks. And it did so well, that Vocaloid had to counteract with that and included AI into V6, since we now live in a time, where now people want voices to be as realistic as possible. The companies know that and they also know, that they can't stay stagnant with robotic sounding voices like in V2 and in UTAU, because of how times have changed. As I said, you don't need to use SynthV or V6, neither does someone tell you that you HAVE to use SynthV and V6 in general. You still are able to buy or download their Vocaloid and UTAU Voicebanks, since most SV Voicebanks are only considered as an Append and are recommended to be used with the Vocaloid and UTAU Voicebanks.

​@mozzapple you really are stupid aren't you? Look up the groundbreaking research of Dr. Dennis klatt. These are all derivatives of that design.

@@mozzapple at least the "AI" of SynthV isn't AI in the way that Vocaloid does it. You're able to turn it off in SynthV, and "AI" voicebanks just render differently, it's different from AI that steals. But I agree that AI is super annoying and I don't want it around lol and it'd be nice if Vocaloid didn't force it on people

​@@mozzapplejust tell it to the company who made that

Thanks for that info! I actually love how she sounds here in her new voicebank. The one I can't get used to though, however, is teto :(

I kinda disagree. I think Vocaloid has a different vocal processing than SynthV what allows this not realistic tone. Obviously that engine noise also helps, but when you take that videos of Vocaloids being turned into SynthV AIs you see that even the old ones sounds different in SynthV. Tbh I love both sounds! The realisticaly and non realistic ones!

​@@pedrovitor5324you know those videos are just Synth V voicebank with a voice changer on them, right? They're not actual Vocaloids in Synth V

@@corasundae Not all of them

Synth v exist since 2017,although they only start using AI voicebank around 2020

Many other Synths have been ported to Synth V recently as well. One of the most notable examples of this was Kasane Teto. A lot of the banks have free versions too, I'd recommend looking into it! (It's also less expensive.)

how did you not know about kasane teto

you have alternatives like CeVIO, good ol UTAU, SynthesizerV, NEUTRINO(the pioneer of AI bank),A.I. Voice, VoiSona, VOICEPEAK and more. By far SynthV gaining popularity pretty fast, a lot of Chinese bank moved to it and Teto resurged as one.

​@@karayura10 NNSVS and Diffsinger can get honorable mentions as being the AI counterpart of UTAU. I wish Diffsinger can gain more popularity. It has the potential to rival as the open source version of SynthV like how UTAU is to Vocaloid.

Sheesh, strong take. I agree with the last line. Another thing is that you can easily tune realistic voices to sound robotic if you prefer that, while it's much more challenging to make robotic voices sound realistic.

What I get from this comment is that you're as in love with your own opinion as those you criticize.

I agree. What do you think of the new Teto voice? For whatever reason I can't get used to the new Teto voice :( I loved the roboticness in her utau bank. Also, since Gumi is now on SynthV, I hope Luka comes next!

exactly

I'm of the camp that says a more realistic base is always better as going backwards to something chipspeech/cv utau/po-uta is actually insanely easy both in engine and post-processing compared to basically an impossible task going from something like CV utau to something realistic

​@@RogueError617 Every time I listen to Teto SV, as impressive a product as she is, I just end up crawling back to A_II's songs. (He's my favorite utauP.) What UTAU is and how its voicebanks complement some people's music perfectly is just something to behold. Irreplaceable.

@@raccoonchild I couldn't have said this better! You've articulated my thoughts about this so well. I'm so glad somebody else gets it.

That would be epic

Wait until you discover consonant tuning and pitchsnap, sometimes I wonder if people who make these comments even understand enough about using these programs

@@Hezkun not sure if this is directed at me but i've used these programs for years. i'm talking about the sound of the voice and not the pitch

​@@Hezkunwait until you discover v6 is ass and nothing you say will save it from being ass

well one reason she sounds different is because synthV doesnt have the vocaloid engine noise. so its only natural. no vb will sound alike when switching engines. they all have differences some more subtle then others

I fully agree. I've been listening to vocaloid songs for more than 13 years now, GUMI is my favorite, I probably didn't skip Mozaik Role even once when it came up in shuffle in my music throughout these years. And I don't feel the "GUMI" vibe from SynthV version at all. I never had any issues with sounding robotic, maybe it's just nostalgia but early vocaloid songs are still the most emotional to me despite voices not being finely tuned. But even if we put the whole robotic/realistic issue aside, if I listened to the audio only I would never tell SynthV version is supposed to be GUMI.

@@JustYumeChan Right? It's like the Kasane Teto SynthV bank. It sounds fine (I even really like some uses of her), but it doesn't have that edge that UTAU Teto has. That little mischievous edge in her voice that makes her sound playful and more like the chimera and living joke she's meant to be. SynthV Teto just sounds like her more grown-up and reliable sister.

I think it's probably because of the engine noise maybe? To be honest she sounds perfectly capable of sounding like a clearer version of her vocaloid VB based on how her timbre sounds. Sometimes engine noise can add stuff to a singing voice, giving it more character that's not actually there, but can help the voice out. Flower, Rin and Len, and most Vocaloids to an extent for example, really use that extra space the engine noise adds to have "sub-text" (I don't have a good word for this) I think a big example is Teto, because she DOES sound like Teto, but due to the lack of UTAU resampler processing, some people swear she sounds different (TBF, Teto is an extreme example, she didn't have multi-pitch during her UTAU days which really amplified the UTAU noise)

This is the one thing with the AI voices that kill me a bit. I don't know if it's because of engine noise or processing or what, but it's just a bit too smoothed over that it looses the original charm entirely.

@@NeoNovastar I get what you mean.

Hope they help Saki reconnect to miku's old v2 voice acting to re enact the beautifulness of her v2

Synthv gumi is now my other favorite!

If you're talking about how realistically each one sounds, I think it's because Vocaloid's initial goal was to make "Vocal Androids" and I hear Synth V's is more aiming for realistic sounding voices. But ye I agree ... Vocaloid may have started this vocal synthesizer stuff but they're being left out now... I think it all started with V5. Hopefully Cryptonloids migrate to Synth V coz I heard NT's not that good either

come on V6 still have its charm, btw did you see vocaloid β studio?

Nah. Gumi SV sounds like her voice provider Nakajima Megumi.

I actually think the opposite case, I think her synthV bank is closer to her pre-V6 sound The vocaloid engine noise seems to get worse and worse every gen, maybe you associate the vocaloid noise to gumi?

If you have heard Megumi More specifically her ranka lee era(which was the base for gumi) you will notice it sounds a lot like her

She sounds more like Gumi than her voicer to me.

@@huhuhuhurricanestreamwow7937 if i wanted to listen to Megumi or Ranka lee or any other real singer, i'd listen to them. but i want to hear classic GUMI.

@@kiramekikuroi4408 no one got rid of "Classic Gumi". She is still being sold. People are still using her.

Is because has not the robotic effect of vocaloid, is the same if you heard with attention.

Alrite 1. judge it by standards of realism better sound, better UI, better customization, better Tuning, Ease of Access, Cross Language, Cheaper 2. takes away a lot of creative control What Do you Mean by this exactly ? If you gonna Write Djent or Metal Song doesn't Matter What the Vocal synthesizer are you Still in charge of "writting" the song So i dont understand what do you mean by this sentence ? How different Vocal synth gonna kill your writting process? 3. the reason vocaPs never make the switch Because they already have well estabilished Fanbase and People doesn't like change mostly Remember when twitter made everyone mad and people Switching to competitor ? Doesnt last long Remember When Mixer launched ? people still stay on Twitch Apple literally Fuck everyone on each new release being anti-consumer and People still buy it. "Name" is more important than anything, especially since people really attached and made their identity to it 4.the reason both v5 and v6 failed Nah it just Yamaha being got out-league by SynthV for AI stuff, how come Corpo 50x larger than Dreamtonics couldnt catch up with the tech ? 5. sheer multitude of options you have to tinker with the program and create your own sound You could still do that with any other Synth Vocal you know, it just matter of taste. You don't Need to make it V4 are more superior than SynthV, You just like it Better. Some people Love their 60s les paul But in lot of way Strandberg is better Guitar 6.but that's never really been vocaloid's criteria for ""good"" Again Preference, Nostalgia, Bias etc 7.complete creative control Every Synth Does that, Coming From the guy Who use V3 to SynthV. my Creativity Doesnt Change from me Switching Engine but from trial,error and experience. 8.tuned by ryo or kz or giga They could still Do the same on SynthV tho, what you talking about ? you pretend as if its imposibble to do spesific tuning on V6 or SynthV? 9.blank slate SynthV also this, you realize they "do" tune on SynthV right ? and there's huge different between untuned and tuned work right ? right? 10.no automatic tuning, no automatic realism Its an Ease of access, you could disable it if you dont want it. u using Smartphone right now, right ? why not old nokia ?oh right, because its better and ease of access for literaly everything. But you could go to Emergency mode and make it work like old phone for communication only. It's there if you needed, its not there if you dont need it. 11. i've worked with daws for decades Coincidentaly i worked on DAW too, But i dont have brand loyalty and i would switch to anything if its better.

think about this one, when you make them sing together they are gonna still sound great doe

@@MrBlind-kr4vq well of course, her voice was beautiful from the beginning c:

Bet most people wouldn't be able to tell the difference when it's not explicitly stated

@@Hezkun Idk I listen to her for a long time so I think the difference is clear as sky is blue, I personally love the robotic touch in v6

But she doesn't.

Well... V2 was like a vocaloid ranka lee.