How to fake appends?

[gumilytime]

The vast majority of the banks I like don't have appends. Do you know how to make them sound like one? The ones I use most don't have them.
- Power
- Sweet (understand sweet like GUMI's)
- Vivid
- Deep/Dark
- Mellow Adult

 

[Alphonse]

Other than Sweet and Adult (you can lower/raise GEN), you can't. The parameters in Vocaloid are not natural to the point where they sound like alternate recordings.

 

[mobius017]

As far as I know, what @Alphonse said is correct.

I will throw in a suggestion to look at plugins, particularly equalizers (EQ). EQs let you raise/lower the volume of ranges of frequencies in your sound. You can use it to make the voice's higher pitches quieter and its lower pitches louder, for example, making it seem deeper overall. It's not going to be the same as having an append--it will sound the same, just pitched differently, and appends are more than just pitch differences--but EQ can change the sound a lot.

Compression can also change the sound a good bit, making it sound more powerful and adding a kind of crispness to it.

I'm not an expert with these (someone who knows more might be able to do magic that I can't), but my experience, again, is that these won't magically create an entirely different voice, they'll just make the same voice sound (noticeably) different. But without appends, those and the Vocaloid parameters are basically what you have to work with.

Among the parameters, I'd recommend looking at BRI, GEN, CLE, DYN, and even Volume, as those can affect how cute/smooth/crisp the vocal sounds, too. We have a Tuning guide in the Resources section that has good information about parameter usage--check out some of the books.

 

[IO+]

You can manipulate a voice to some extent by stacking different vocal styles and blending them using gain control. But first, you need to create a base singing style as foundation first. Similar to how RGB colors combine turn it into Cyan, Magenta and so on, you can mix vocal styles with gain control automation, as shown in the picture below.



The beauty of vocal synths is that they never go out of sync (although some voice banks have different attack timings). Because of this, you can stack them like pancakes indefinitely. However, it's important to note that most voice banks have different formant structures, so proper EQ control is essential to avoid phase overlap issues that can hollow out the singer's voice. I should also point out that harmonically stacking vocal layers with different gender characteristics is a good starting point.

I find this method very similar to XSY, but with more control. It can be quite effective in certain situations—for example, when trying to de-Mikuified a Luka SP

 

[mobius017]

@IO+: For this technique, am I correct in thinking that you would basically be:

1. Finishing the song's lyrics/tuning completely
2. Duplicating that finished track several times
3. In each duplicated track, changing the base Vocaloid parameters that apply to the singer as a whole (i.e., in Piapro, making singer "presets," rather than adjustments in the real-time tuning area)
4. Using gain automation to make those different tracks quieter/louder during the course of the song

 

[IO+]

@IO+: For this technique, am I correct in thinking that you would basically be:

1. Finishing the song's lyrics/tuning completely
2. Duplicating that finished track several times
3. In each duplicated track, changing the base Vocaloid parameters that apply to the singer as a whole (i.e., in Piapro, making singer "presets," rather than adjustments in the real-time tuning area)
4. Using gain automation to make those different tracks quieter/louder during the course of the song

Correct, I recommend to finish the tuning first because it's much easier to see a whole picture of the song that way. It save time and sanity.

The hardest part of this method is that you can't rely on what you hear in the editor, that means you don’t always know when to stop tuning or keep going because stacking them like a pancake it's always sound bad and as i mentioned, EQ and phase control has to be done carefully.

A good place to start is by bringing the volume of all the vocal tracks down. Then, choose your best track and use it as the foundation. After that, begin bringing up the other tracks and EQ them one by one.

Phase overlap is unavoidable, but it can be managed by using phase rotation or by shifting the phase 180 degrees, though it will always leave some artifacts.