UTAU Tuning and Use: Realism vs Instrumental Quality?

[WendytheCreeper]

So I want to gather some opinions on this. When it comes to how an UTAU (or Vocaloid or whatever) is tuned and used in a song, what is your preference: to have it sound as human as possible, or to have it sound more like an instrument of focus?

For example:

In one song, you could be tuning your singer to sound human, with human-like pitchbends, breaths, a "natural" quality to the voice where you aren't putting too many effects or weird things to the voice.

In another song, though, you could be putting a whole lot more focus on how the vocal sounds as an instrument of focus in the mix of everything else. The voice might sound cartoony or more like a synth. There might be a whole lot of effects on it, or it might sound robotic.

A good example of the concepts I'm trying to illustrate could be, say, the use of Vocaloid in World is Mine vs how it is used in DYE. please don't kill me over your opinions of the tuning of these songs, I know they're old but they get the point along for me

What are your preferences?

 

[chunter]

I have no preference, sometimes the voice is not the lead instrument at all. That versatility is part of the advantage of using it in the first place.

 

[HulderBulder]

I don't enjoy it when some uses breaths on every note with a rest after it in a ust. It feels too breathy and excessive imo. Like the singer is badly trying to act like they're struggeling to breathe.
Same goes for growls. When the growling sampler came alot of people apllied W to the majorety or the whole ust, making it sound like there was something wrong with the voice.

I don't think one should avoid to use them at all, but instead use them sparringly. Think where you would use it if you sang. You might breathe after long or high notes, perhaps after many small notes with tiny rests? Also, you might only growl a little on a few notes, maybe when you want to give it some extra umph?

I think one should go for what idealy sound best, and not nessesarly the most human. Mostly because it's not an actual singer, so doesn't suffer from the limits of a singer.
Hopefully my rambeling thoughts makes sense.

 

[Kiyoteru]

I tend to not use a lot of tuning effects anyway (like breaths, growls, etc.) because I feel like they're contrary to the treatment of human vocals in the mixing process. Then again, I mostly do electronic music, or louder music in general, so I guess that means I often treat vocals as an instrument. The last time I did vocalsynth work featuring subdued instruments with a focus on the vocal parts.... I can't remember exactly, but maybe it was my christmas cover (which was entirely a capella). In that case, I used a vocalsynth that was distinctly non-human, Chipspeech, so there wasn't any point to making it sound more "organic" to fit the genre. I have an instrument perspective in general, then.

 

[Chianachini]

To me, it depends on the message of the song. World is Mine is conveying a very different message compared to DYE and uses different techniques to say it. It's up to the artist to decide how to do this--either using the voicebank as either a traditional singer or more like a VST instrument, just like how they would decide it's BPM or what scale to follow.

That also gives cover artists the freedom to convey the message in a whole other way, sometimes changing the meaning behind the song altogether. Art is about freedom--there are no barriers to self expression.

 

[WendytheCreeper]

forgetting about own threads like

For me, I think there's a time and place for using something realistically and using it not realistically. If you are writing the song to be very lyrics-focused, and the instrumentals are more acoustic in nature(rock, classical, etc), then it's safe to say that it would be weird to tune your UTAU as if you were using them in something very electronic. It would clash and be awkward.

The nice thing about realistic vocals is that they can work in a both "realistic" and "instrumental" quality. You could stick realistic vocals into a more electronic or instrumental-focused song and theoretically it'll still sound good.

The same cannot be said for, say, using the vocal more like a synth. At least in more acceptable and non-experimental music. (Though damn I should make it a goal to write a live orchestra piece featuring live played Alter/Ego because that's my jam) (Also we need more theremin in classical music)

 

[Ame。]

I always go for instrumental quality if I can.