#cross-lingual synthesis

[dead-byte]

So, for those who don't know, I have a video series on YouTube that demonstrates the Vocal Modes and Cross-Lingual Synthesis capabilities of Synthesizer V AI libraries, primarily to help people make purchasing decisions.

I just uploaded videos for the new child vocals, Lin Lai, and Yun Quan. Check them out if you're interested!

[bmpmp3]

and dear god we need another irish vocal synth. avanna come back to us

[artifexvigen]

So I picked up the trial for SynthV 2 and...it's okay? Like, it's not a downgrade or anything, just not a huge improvement. SynthV was already the lead in vocal synthesizer technology so it's kinda hard to improve on that in the way that successive generations of VOCALOID did. I'm gonna go ahead and give my thoughts on it as a user of SynthV.

First of all, I'm not at ALL fond of the new UI. The sliders being replaced with knobs and graphs is a straight up downgrade. It turns a lot of the precision tuning into guesswork and I'm not a fan. Although, color-coding the tracks and notes is a big improvement.

Adding Korean as an option for cross-lingual synthesis is a huge plus, I'm always grateful for new language options.

The trial voice I used to test it, SF-A2 Miki AI 2 is...pretty good. About what I'd expect of a SynthV. (Though WOW that's a lot of vocal modes. Hello??? Miki???)

Ultimately, it's an improvement in some ways, and the UI is really my only hangup. I may end up upgrading within the year if I have the money because I have a discount as a returning user from SynthV.

I wouldn't say "don't get it", but I'm not exactly rushing to get it either as SynthV does more than a good job for my needs and I have other things and other vocal synths that take priority to me.

In any case, take my words with as much salt as you like, since I'm just one guy and my opinion on this ultimately doesn't mean much. I just wanted to say my piece.

[canadianlucifer]

I just wanna give my two cents on AI since everyone is talking about it again and honestly, AI has a lot of good uses!

AI is what powers characters in video games, it's able to analyze large amounts of data for healthcare or finances, it's able to improve translation of languages, and a lot more. When used properly, AI has plenty of benefits!

The problem is generative AI using content stolen from those who did not consent.

Generating visual art based off artist who work hard to produce their art, singers and voice actors who train their voices being made to sing or say something they didn't, and famous personalities as chatbots. That is what is terrible. It's not only stealing work away from them but trivializing it too.

AI is completely unregulated right now, anyone can use it to do just about anything, and it's completely unethical. What we need is regulations, rules against what can and cannot be made with AI. Ways to differentiate AI generated content from real work a real person put effort into.

I'll be honest and say it proudly: I use AI very often! Synthesizer V is a program I use to make music, like Vocaloid, and it makes use of AI. The difference between SynthV and some "make spongebob sing old town road!" program is:

Everyone who's voice is involved were paid to have it be used

Worked directly with the developers and consented to their voice being recorded for this purpose

And it requires significant effort and practice on the part of the end user to produce something.

I've been practising for ages and I'm nowhere near decent because the AI implemented in it isn't some crutch, it's something that allows for more human creativity! The AI helps make the voice sound more how I want it, but can also convert say, an english voice into spanish! Voice banks usually only come in one language, maybe two, and if you want a song where the vocalist switches languages for a few words or maybe you can't afford to buy another voice bank in a different language or maybe you just really like one's voice over another, it's as simple as flipping a switch! Cross lingual synthesis isn't perfect of course, but it's a lot better than struggling to twist phonemes from one language into another. And that feature is thanks to ethical AI!

AI isn't completely good or completely bad, it's a broad and complex topic. AI generating soulless "art"? Objectively bad. AI assisting artists reach new heights? Very good! Honestly I think the biggest problem is labelling everything that is AI as AI. It's vague and doesn't accurately describe what it does. Artificial intelligence, when I think of it, is a computer that is able to form it's own thoughts separate from any training data. That obviously doesn't exist yet, and what we currently call AI is nothing more than a complex algorithm.

I'm no expert in AI, I haven't got the slightest clue how it works, but I hesitate to denounce all AI when it does have many positives to it. Many negatives too, but everything is some shade of grey.

In short: get explicit consent from artists and compensate them.

[takoyakidonuts]

SynthV SAROS Release Day and Special Addition Revealed! + SeeU Final Sale Crowdfund Details!

Eclipsed Sounds released their newest demo for SAROS tonight featuring the Special Addition they’ve been working on for them being Official Spanish Cross-Lingual synthesis! SynthesizerV SAROS Official Demo song “Forever” by AkuP In Eclipsed Sounds’ blog post, they disclose that research and development of Spanish Cross-lingual synthesis have been primarily done by them at ES with guidance from…

[infiniteoreos]

synthv cross lingual synthesis is great when you're trying to make a japanese voicebank sing in english but god forbid you want an english voicebank to sing in japanese. asterian can't pronounce a る to save his life but mai can say tourniquet no problem

[yuraslefttoe]

asking this to every synthv prod i can contact;

when using a synthv vb for a song in english whose native language isn't english but has cross synthesis, do you use their english mode, or do you do it the good ol' way and tough it out through their native phonemes?

I don’t use synth v a lot but as a developer it really only makes sense for me to be using the cross lingual for the best results as there will be data from the base language model to assist in pronunciation. I also feel like synth v retains accents very well so it’s kinda a waste in my eyes to try and use the native phonemes if you have that other language model available

[exanuz]

however if i wanna get haiyi and use her, i think ill need the cross lingual synthesis, so ill need to buy the pro version of synthv

the pro version is 137$ cad, and haiyi is roughly 160$ (depending if i buy just her digital version or her version with the dvd box itll be a few dollars cheaper or more expensive) ... 300$ ... maybe i should jst get anri arcane instead so im only spending 160$ and dont need to spend that extra near 140$, but.... i want haiyi....... ive been excited............

[synthvnews]

GALENAIA - Vocal Mode & English Cross-Lingual Synthesis Demonstration (Synthesizer V Studio 2)

English Universal Demonstration Song by ‪‪unit.0‬

GALENAIA Showcase song by ‪vanelily‬

[bmpmp3]

i neeeeeed everyone to use chis-a more NOW she is free... she is free to use.... you can download voisona and her right now........ her husky androgynous vocal swag...... let her sing your tunes...

[music-dmt]

The Evolution of Lyricism in Modern Pop Music

Lyricism in pop music has undergone significant transformations over the years, reflecting changes in societal norms, technological advancements, and cross-cultural influences. At the heart of this evolution is the way songwriters and artists express emotions and narrate stories, resonating with audiences worldwide.

One prominent example of modern lyricism can be found in the song "Bad at Love," whose lyrics explore themes of vulnerability, failed relationships, and personal growth. The raw honesty in these lyrics resonates with many, reflecting a shift toward more authentic self-expression in pop music. The discussion of these lyrics can be further explored here.

Artists like Bella Poarch have brought a fresh perspective to pop music, not only through their innovative use of social media but also through their distinctive lyrical styles. Bella’s approach to music integrates catchy hooks with personal storytelling, creating a unique brand that engages listeners on multiple levels. Her impact on the industry and her approach to music and lyricism can be detailed here.

Meanwhile, Jeanna Friske, a revered figure in Eastern European pop music, contributed significantly to the genre with her ability to blend pop sensibilities with regional influences. Her music often carried a depth that spoke to the emotional and cultural narratives of her audience. More on her life and musical legacy can be found here.

In India, the lyrical content of pop music has seen an integration of traditional and contemporary elements. The top artists in India have played a pivotal role in this, using their lyrics to bridge the old and the new. This synthesis not only preserves cultural heritage but also makes it accessible to the younger generations, who are looking for modern interpretations of classic themes.

The therapeutic aspects of music, particularly its lyrical component, cannot be overlooked. Lyrics that address personal pain, societal issues, or universal truths can have a healing effect on listeners. This aspect of music is profoundly covered in discussions about how music has the power to heal. It explores the psychological impact of music and how it can restore and rejuvenate the human spirit.

Moreover, the digital age has brought about significant changes in how lyrics are crafted. Songwriters now have access to a global audience and are influenced by a wider array of cultural narratives and languages. This diversity is evident in the multi-lingual and cross-cultural collaborations that have become prevalent, allowing for a richer, more diverse lyrical tapestry in pop music.

Engagement with fans through social media has also changed the way lyrics are perceived and understood. Artists like Bella Poarch often use platforms like TikTok to share snippets of their music, creating a buzz around the lyrics even before the full song is released. This immediate feedback loop allows artists to gauge how their words resonate with the audience, often leading them to tweak their lyrics to better connect with their listeners.

The role of lyrics in storytelling has also expanded. Today’s artists are not just singers; they are storytellers whose songs are chapters in a larger narrative of their artistic persona and personal growth. Each song carries a piece of the story, whether it’s a tale of heartbreak, triumph, or self-discovery.

In conclusion, the evolution of lyricism in pop music reflects broader changes in our global society. As we move forward, the fusion of traditional and modern influences will likely continue to shape the lyrical landscapes of pop music, making it a powerful tool for cultural expression and personal healing. The contributions of artists from various backgrounds, such as Bella Poarch, Jeanna Friske, and those leading the charts in India, will continue to play a crucial role in this dynamic evolution. The ongoing discussion about the therapeutic powers of music adds another layer to our understanding of the impact of lyrics in our lives, underscoring music's enduring value as a source of comfort and connection.

[hua-mo-jin-is-a-cutie]

I decided to cover Humansongs by Porter Robinson using Hua Mo Jin :) 

[vocaloidfactoftheday]

In 2018, two separate Synthesizer V voicebanks for Xingchen were announced to be in the works, Xingchen Chinese and Xingchen English. (There was a possibility that a separate character called ZERO would be used if the voice was too different from Xingchen's VOCALOID4 voicebank.)

The current status of Xingchen English is unknown, but it likely canceled due to Synthesizer V Studio's cross lingual synthesis allowing Xingchen's SynthV voicebank, Xingchen Infinity, to sing in English. While it is not technically considered an English voicebank, this makes Xingchen the first Chinese Vocaloid to have had an official voicebank with English capabilities, albeit for a different editor.

(image source: Synthesizer V Wiki)

(source: Ddickky on Weibo, Synthesizer V Wiki)

[synthvbulletin]

Cantonese Singing + Rap Demo

Dreamtonics has released a demo for Cantonese cross-lingual synthesis. As an addendum, the end of the video shows a demo for a rapping SynthV voice, although it’s unclear how this will work within the actual program as of now. Stay tuned for updates!

[mirukutchi]

Aaa voca tumblr!

So gumi ai can sing english right? Like there isnt a separate english vb? Its just the gumi on the vocaloid website? Also how do u make her sing english, im curious abt the cross lingual synthesis!