No.15340
When will we get english equivalents to those Zundamon history lesson videos that japan has
No.15342
>>15340There are already some like Zundamon's culture guide
No.15348
ive always been partial to the ones built in to mac os. i dont feel strongly enough about it to get a mac though. maybe if its cheap secondhand.
im also charmed by lalavoice and dectalk. i really love older speech synthesizers so if anyone wants to talk about i would love that.
No.16471
For me it's Microsoft Sam
No.17405
>>15322Does /leek/ like CoeFont?
No.17477
>>17476Actual Mac synth or is it just FL Studio whisper?
No.17482
>>17476Do you have the GTA IV Journey radio voice? I think it's called Vicki
No.17488
>>17418Tell me more about her
No.17492
>>17491Now you mention it I guess that is why it sounds different.
No.17532
>>17528You can make Voicevox sing?
No.17541
>>17540It's called
ＡＭＡＺＵＮＤＡ
No.17545
>>17541The AMA here stands for AMAZING rather than AMAZON
No.18069
I have been bothered by the fact I can't get TT, Teto, or Zundamon to be my TTS on Android so it can be the voice I hear when using Organic Maps. Also I've actually never heard or found an example of TT speaking English. Currently I'm using SherpaTTS with en_GB-Cori-medium as my Android TTS. I don't know if this is even possible to port to SherpaTTS or just make a whole new TTS app for this purpose. SherpaTTS will stop working on Android though due to the developers not wanting to identify themselves to Google.
I also have been trying to figure out how difficult it would be to port in my own audio files into Bluetooth speaker devices so I can use whatever voice I want instead of these funny heavy Chinese accent girls which is fun but having a beloved talksynth is better. I'd be worried about risk of bricking whatever device I have if I tried to do this though but I might do a test run on some cheap speaker if I figured out how to do it.
>>17491>>17482Before I forget if you want to write up a script I could use Vicki's voice for an intermission for the Vocaloradio or something and I could upload it here then someone else can add the needed reverb and other touch ups. My old iMac is too old to be good editing on and it's only computer I have working right now and I've not gotten used to using Linux or BSD loaded from a flash drive or anything yet so I can't just get around the old OS limitations right now.
No.18087
>>18069The Bluetooth device is ready to pair *heavy chinese accent*
No.18088
>>18069I love the Chinese bluetooth lady
No.18646
>>18639cute outfit, reminds me of chiaki nanami
No.27032
>>17489That is definitely the Journey voice
No.27050
Is there anything else out there to use on Android for the TTS? All I can find is this Voicevox mobile but I assume it wont work because the VBs likely expect kanji or kana not roman letters. Was trying to figure if I can get TT-chan to use for my GPS navigation.
https://github.com/VOICEVOX/voicevox_mobile>>20125Is this you?
>>17071 if not then yes, if it is then seems not. I am confused why it requires a download if it is a web client, since I do not have a good enough working computer to try it on.
No.27065
>>27050There's an app called BOIMERO
No.27069
>>27065Thanks but seems it is not what I am looking for. That is apparently for singing and it uses a server side processing which would clearly not work as the speech program because it be slow and destroy privacy even if it worked somehow which I dont think it is built for that. I currently use SherpaTTS but the voices are uninteresting. I am using en_GB-Cori-medium which was most pleasant I could find from their available voices that worked but that also lacks Japanese so if I try to do audio playback on DeepL app for Japanese it says it is unsupported with the voice though that is not too big an issue but it would be nice if I could use a voice more entertaining to hear when I am out on long walks being navigated by my GPS guidance using the TTS set up on system wide use.
No.27359
Apparently CoeFont is cross lingual...sort of, it just converts English text to katakana and guesses the tonality. They have Voicevox voicebanks of Nurse Robot Type T. This is her (悲しみ) sad voice since the sample sounded clearer than her standard voice.
https://coefont.cloud/coefonts/437759d6-3c12-41b6-a1ef-89e34c7ff0d6Cant use this for what I wanted to do being cloud based but at least we can hear her accented English which was something I could no examples of on YouTube beyond the occasional loan word.
No.27360
>>27359I meant intonation, not tonality
No.27362
>>27359Need TT ASMR in Engrish right now
No.27364
>>27362You will have to try yourself. I already ran out of credits trying to reduce noise on her whisper voice by trying alternate spellings and don't want to give a card for free 7 day trial since I do not like subscription models. I guess their limit is more meant for Japanese users and they never considered English is not information dense like Japanese. Tip: if you want to do the full trial you can use virtual cards to hide your real card and prevent payments such as with Privacy is one I know you can do that with but I haven't set it up myself yet. Dialogue i started with is
"I see. So what would you like to eat today? Squishy squash? Prickly pears? Orange....oranges?" lol
No.27365
>>27364Ah don't tell me Coefont is one of those credit based things like Elevenlabs
No.27367
>>27365Yeah even top plan of $350 per month is capped at 1 million and cheapest is $20 a month. I am not sure how I went through 8,000 so fast. It either is based on rendering time or character count.
No.27373
Are there really no English TTS with cute mascots out there? I know that Voiceger has an English Zundamon but is that really it?.
No.27374
>>27373You know how people say there's an oversaturation in SVS? Well i think there's a lack of TTS aimed at Otaku in the west. So maybe someone should try developing that, a "Maghni AI" of text to speech.
No.27398
>>27373>>27374The main issue is that the TTS audience that’s related to Vocaloid in the west is significantly more niche than singing synths. There’s a reason that we haven’t seen beyond Zundamon, Maki, Soyogi, and the kotonahas get English tts. Heck, there’s a reason why Maki didn’t get English on Voicepeak. The bare minimum I hope for is Voicepeak to get some sort of cross-lang at least but it’s only ones that have no intention of gaining an otaku based market that do well in the west.
If AHS and DT at least released the English narrators they promised for Voicepeak in the past, maybe there would be some hope for more English.
No.27400
>>27398Ironically they would probably succeed if they marketed them to south americans considering they have their own version of the "avatar plays videogame with text to speech voice" genre
No.27402
>>27398>the TTS audience that’s related to Vocaloid in the west is significantly more niche than singing synthsI wonder why that is
No.27407
>>27402It almost never gets fansubbed, and now everyone in the west is into vtubers instead. The talkloid community would go nuts over a good TTS bank, though.
No.27546
>>27545i don't understand anything other than milk
No.27547
>>27546"Want some milky milk? Hey! Dont look at me like that! tehehe "
Yeah I was bored and had to shorten things a lot. The credits are per character count and that includes punctuation. Seems some versions she pronounces way better than others but like her whisper voice is so full of noise. Strange there is no examples of TT in English outside of what I have done. You would think she can pronounce it seeing she can do Mandarin Chinese which has more phonemes than Japanese does.
No.27549
>>27547So that's what the T stands for
No.27553
>>27545I can't be the only one who finds heavy Japanese accents like this adorable right?
No.27558
>>27553No, heavy accents are moe
No.27560
>>27558They are but there is a line between moe and too difficult to understand some words
No.27567
>>27560Two things can be true
No.27782
im surprised crypton have never made a miku tts