Google Translate’s text-to-speech (TTS) function is terrible at establishing cultural context. Don’t believe me? Try typing the Korean last name “Jung” into either the English TTS or the Korean one. You’ll get the pronunciations “Young” and “Joong”, respectively. It’s basically common sense to English speakers that “Jung” is pronounced phonetically but Google Translate, as well as most TTS tools, arbitrarily apply pronunciation rules that shouldn’t be applied. In the case of “Jung”, Google thinks it is a Swiss name, like in the case of psychiatrist Carl Jung. And it's not just Google's TTS algorithms that do this. It's all of the major text-to-speech providers. Believe me, I tested them all before coming up with this software. Pronounce My manually programs accurate phonetic rules into a TTS algorithm so that you get a relevant pronunciation to the language you are generating pronunciations for. This was done with the consultation of native speakers and bilingual translators for every language we offer.
Is your pronunciation guide complete with every single name and location in your script? Is the pronunciation guide you received decipherable to the English-speaking ear? Do you currently feel empowered and informed enough to have a discussion with your client about how foreign versus localized the pronunciations should be? If you answered "no" to any of these questions, it's likely that Pronounce My would benefit you and your project.
Authentic Americanized Pronunciations are pronunciations that honor the phonetics of their source language while using sounds that American English speakers can easily say. A common misconception is that reading romanized names phonetically is the equivalent of “Americanizing” them. In Korean, this leads to all kinds of issues, such as in the case of the name “Seo.” English-speakers are inclined to read this as "See-oh" but it is actually pronounced “Suhh" which rhymes with "uhh". There are so many other examples in Korean and other Asian languages, where the romanization betrays what English speakers may phonetically be inclined to say. Authentic Americanized Pronunciations aim to be both culturally sensitive and straightforward in correcting these common misconceptions. Currently, Americanized pronunciations of Mandarin and Japanese terms are enunciated for clarity and accuracy.
Yes! Just right click the audio player of the pronunciation you want to keep. Click “Save audio as” and name the file to your liking. This will save an mp3 recording of the pronunciation to your computer, which you can then import into whatever software, folder, etc. that you prefer.
Pronounce My can help with any kind of localization project that involves the spoken word. This includes but isn't limited to films/tv, k-dramas, anime, video games, mobile games, audiobooks, podcasts, songs, etc.
Yes, that is certainly the plan. I’m hoping to include full support for Korean, Mandarin, Japanese, Tagalog, and hopefully Indonesian and Hindi. Additional languages will be added to all plans at no additional cost.
Yes, any device with Internet access can use Pronounce My because it's hosted on a website.
There aren’t plans for one because offering Pronounce My as a web application allows me to manage users and push out updates much more quickly.
On rare occasions when the input includes something too tricky for the English text-to-speech algorithm to read, only a foreign language pronunciation will generate. A written pronunciation guide will still generate for you in most cases.
While the text-to-speech algorithm on some of the voices does use neural networks to smooth out the sounds between syllables, Pronounce My does not use AI to determine pronunciations, nor does it store or use any of your inputs for machine learning. I understand that privacy is a must since many projects using our tools will be working under strict NDAs. I’ve created Pronounce My with security and safety in mind.
I wish I could make it free but it unfortunately costs quite a bit to run and maintain.