Tsurumaki Maki has a release date!

Hi all,

Just wanted to make you aware that AHS has opened preorders for Tsurumaki Maki, who releases on June 18 with four voice banks (AI and Standard banks for both Japanese and English). This means our first full English voicebank is coming very soon!

And yes, there will be lite versions made available some time after her release!

You can read more here: 弦巻マキ|製品情報|AHS(AH-Software)

Maki’s demo songs also dropped yesterday (technically this morning JP time). This is a playlist with all her demos: Synthesizer V 弦巻マキ 公式デモ曲リスト - YouTube

「いいね!」 4

YES! I believe in Tsurumaki Maki supremacy and finally Eleanor will have an English friend.

「いいね!」 1

I had a look at it yesterday. There doesn’t seem to be any obvious way for non-Japanese users to navigate the web pages, and it’s not clear how to purchase the voicebank from outside Japan. Google Translate via the Chrome browser stops when it gets to the final payment page.

So even if I wanted to purchase it, I can’t. :no_mouth:

But the lack of any apparent English support on the website makes me leery of purchasing the voicebank.

The CeVIO AI version (yes, I know - not the same as SynthV) has two English sentences in the video:

Hi, I’m Makii Tsurumaki. CeVIO AI Tsurumaki Maki Talk Voice English is a text-to-speech software based on the voice of Manami Tanaka, a voice artist. Just type in your favorite sentences and it will read them out in English right away.

That third sentence is the sentence that’s used to show off the “Fine”, “Angry”, “Calm”, “Sad” and “Mild” voices. And it pronounces “read” wrong… five times in a row.

If they can’t bother to check the primary sentence in the demo - or even notice it that it’s wrong- I can only imagine that the English language support is next to nothing.

Given that Japan is likely the primary market and English support is doubtless not a cost-effective proposition for them, that seems like a good decision on their part.

But it’s likely going to keep Eleanor Forte the primary voice for English speaking SynthV users - at least, until a distributor that supports English speaking clients starts selling it.

Most of the demo songs on the “English” page are marginally in English - which is apparently typical for English-infused Japanese songs.

“Dancing in Binary” sounds OK, but honestly I have trouble making out the lyrics, and I’ve listened hard to the song a far too many times:

Show kind of feel I do inside
Moonlight falling down the snow
Raindrop like a tears on cheek
Shade is flowing in my heart
Nuck nuck candle my best friend
I’ll be going out today
My voice sound a lot unknown
Woah we will I go
Where we are living like this world
Can you give your ears and see
Let me tell ya, let me tell ya
People shouting urgently
Nothing crazy, nothing crazy
Nothing in their finery
Let me dub you in this spiderweb
No one makes you satisfied
When you’re alone then come and see your fantasy

So many lines that are hard to make out. Is it:

People shouting urgently

or is it:

People shout encouragingly

I think this is a fine demo for non-English speakers. But as an English speaker, I can’t tell if the reason I have trouble understanding the lyrics is because of the singer is hard to understand, or because the lyrics are basically nonsense.

The demo doesn’t make me want to buy the voicebank, because the demo is largely unintelligible even after several times listening through it.

And one of the things I want when people listen to my songs is to understand the lyrics.

The other demo song with bits of English is a bit easier to make out, but it helps immensely that the lyrics are printed in the video. But the lyrics are still pretty much random sentences thrown together:

I’ve never doubted the eternity
Your eyes are shining
I thought I could control our lives
Going to buy some bread for morning
So this story won’t end

The /ay/ phoneme is problematic here.

Fly through the sky

The /ay/ is actually quite good here on this lyric, but the /dh/ is bad. Which is… puzzling, honestly. It makes me wonder how consistent the voicebank is.

But all in all, it reinforces that idea that this voicebank is targeted at Japanese speakers.

Again, that’s fine - it’s an excellent business decision.

I’m not dumping on this voicebank at all. It’s an excellent product for it’s target market.

I’m just saying why listening to the demos don’t convince me to purchase it.

That’s correct, Maki English is a Japanese-accented English voice primarily targeted at the Japanese market. This is not uncommon for vocal synths, and is similar to how Crypton treats Megurine Luka. Another use may be to create American-accented Japanese, similar to how Kanaria used GUMI English for King.

International buyers can purchase her physical editions via Amazon: https://www.amazon.co.jp/-/en/gp/aw/d/B093LBJ4N6/

Keep in mind she is currently only available for preorder and additional purchasing options may be available post-release.

As for the “quality” of her English, I do agree it would have been nice to have English demos created by English-speaking producers.

I’d suggest trying out her lite version if you have doubts. Of course we can’t know for sure for a product that isnt released yet, but these things (awkward pronunciation and word choice) may both be things we can account for as producers. The best way to find out will be to wait for a lite version, as well as watching for English producers making songs with her shortly after release (for example, I know CircusP has a song prepared in advance for her).

「いいね!」 2

Oh, for sure! :smiley:

In the worst case, it would still be useful as an additional timbre on a background vocals.

The bank sounds good in general, but it wouldn’t be the one I’d reach for first for a lead English vocal.

「いいね!」 1