How do Vocaloid5's native English vocals compare to Synthesizer V's Cross-lingual vocals?
Вставка
- Опубліковано 10 вер 2024
- things to note:
* vocals were tuned with synthv's auto pitch writer then ported into v5. it's the same vpr and tuning for both examples
* aside from a few spots I missed both lyrics in these sections have been adjusted to account for the way each synth splits the words (i.e. "baby +" in synthv and "ba- by" in v5)
* This was mostly a test to look at pronunciation, not voice quality or range, though obviously those elements are present
Synthesizer V has recently been given a feature where AI voicebanks are cross-lingual, meaning they can sing in other languages. From what I can tell it works by having the synth voice sing the phonemes of the other language. The results are amazing, which is why I wanted to compare them to the native Eng vocals Vocaloid5 has.
Comment what you think below, I'd be interested to see people's thoughts/reactions. My own opinion is below, but I'd ask that you watch the video first and form your own opinion before reading mine. (Hopefully you've already done that if you've read this far in the description, but you never know!)
MY OPINION BELOW, WATCH VIDEO BEFORE READING:
Obviously Saki's vocals are better in every way which. Frustrates me to say the least. Tone quality, clarity, enunciation, and just plain sounding better. I get that Chris is one of the default banks for V5 and so nobody's really focused on how good he sounds, but for the default banks of a program wouldn't you want them to sound better so people want to buy the program? It's not just Chris, either - Amy, Ken, and Kaori all are super hard to work with because they just sound bad. It sucks, quite frankly. I have to wonder how these voicebanks would have sounded if they'd been in Vocaloid4, or even in another vocal synth like UTAU or SynthV. (Yes, even UTAU would sound better with properly recorded vocals.)
CREDITS:
Song - Line Without A Hook by Ricky Montgomery
MIDI, VPR, Tuning, etc - Me
I WAS JUST THINKING ABOUT THIS like obviously we have native vocals like Eleanor and Solaria who are great but Dreamtonics managed to make a vocal synth that made Japanese/Chinese banks able to pronounce English more clearly than the highest quality English Vocaloids. The innovation and jump in quality of vocal synths we've had in the past 3 years because Synth V is INSANE compared to the last decade and a half of Vocaloid
SAKI BEING LEAGUES CLEARER AND BETTER SOUNDING THAT CHRIS,,,,,,, LOVE TO SEE IT
It really is amazing how much more care was put into synthv compared to vocaloid 5
Yeah, I definitely can't help but feel like the default V5 banks were intentionally nerfed to convince people to buy the premium version for the higher quality banks. Which is a shame, because the engine itself is actually good, from an audio quality standpoint.
What do you mean by the "premium version"? I just bought regular V5?
@@Sours56 There's a version they sell as the "Premium" version. It's the same as the basic, but they add VY1, VY2, Cyber Songman, and Cyber Diva.
@@AlterP Ohhhhh. Yeah, the quality difference between the other voicebanks I've used in v5 and the default ones is stupid. You'd think if they developed new voicebanks for a new engine they'd make them good enough to use regularly. I guess that's why all the demo videos I've seen use Cyber Diva or Songman instead of the defaults.
I'm waiting for this cover, I love this song totally 💕
Premiere going up Monday at noon!
@@Sours56 thanks. I'll wait for it
Update: I do wonder what it would sound like if you ported Vocaloid5's automatic tuning into SynthV. However, none of the VPRs I attempt to convert with UtaFormatix allow me to preserve the tuning for some reason. If I ever figure out how to get the V5 tuning into SynthV I'll make another video showcasing it!
How did you port the SynthV file into V5?
I used Utaformatix to convert it to a V5 file! Its a web-based program and it's pretty handy!
@@Sours56 does it also save the pitch or no?
For the most part yes, as long as you tell Utaformatix to Save Pitch Information. I've had trouble getting it to save tuning from Instant Mode, though
The only good sounding V5 standard voice is Ken imo
Ken and Kaori are definitely more tolerable, though I haven't really used them enough to appreciate them. If I had to rank them best to worst it'd go Ken Kaori Chris Amy
@@Sours56 Ken is unironically one of my favorite male vocals lol