Omnivocal’s STRONG German accent makes it impossible to use. I understand that Steinberg is a German company, but English is the most widely spoken language and music sung with an American accent has dominated global charts ever since Americans first created the record industry. So, you’d think English in an American accent would be the starting point, if the accent is necessary. At the very least they should have started with purely phonetic speech or at least the accent of a region capable of speaking phonetically.
This thing is ridiculous. The only way to make it say “Find” is to type “Fyend” Which requires 2 syllables, so you have to cut the note into 2 halves and then, somehow, it comes out as the monosyllabic word “Find”. “Match” and other words are impossible, apparently.
I was trying to use it to write a jingle for an ad, so that I can hear how it’ll sound while I’m working and the singer I hire will be able to use it to learn the jingle before she comes to track her vocals. I’ve gotten much of it close enough by typing in multisyllabic words one syllable at a time, but the German accent is still obvious and too much of it doesn’t sound right. I should be spending my limited time composing and coming up with lyrics, not going through each syllable/note one by one for hours trying to get it to speak correctly.
I have nothing against Germans. In fact, I’m proud of Germany for putting boots on the ground to defend Denmark from our fascist wannabe dictator (Its like we’re living in an alternate universe nowadays). However, the accent makes Omnivocal useless.
Come on, Steinberg. I know PreSonus stole a lot of your American team when they were first developing StudioOne because I did hardware & driver testing for PreSonus back then, but that was a long time ago now and that dud of a DAW StudioOne still isn’t much better than their terrible AD/DA converters. So, it shouldn’t be too difficult to poach them back over to Steinberg to fix this.
Ultimately, as I’m sure you’re aware, the ideal solution is to create versions with American, English, Japanese, Chinese, Cuban, Egyptian, Nigerian, French, Italian, etc. accents and have a drop down menu in the settings for users to select which one they need to use. I’m assuming that’s coming eventually, but in the meantime, can we just get it to be capable of reproducing phonetic input??
I agree with the OP - Omnivocal should be capable of correct English pronounciation. I don’t care if it’s AE or BE as long as it is more or less neutral.
True, there’s no German accent - it must be some kind of Japanese flavour. Anyhow, now I’m tempted to let Omnivocal handle this one:
“Ve Vill Ask Se Kwestions!”
Yeah. That’s why its important to voice our opinions in detail, so that they can work on resolving them. It also why I made sure to state it so strongly because, at the beta stage of development, how strongly users feel about the things we dislike and how many agree are the basis for where the issue fits into their priorities and ultimately dictates whether the issue gets resolved.
I use EmvoiceOne, and it is great at helping to define tempo variations due to the nuance it creates, as a sound or note transitions and ultimately peters out.
I am considering DreamTonics Synthesizer V, however ARA is not working in Nuendo at present, and while nuances can be more detailed, I am not yet ready to introduce, change or exchange vocals, when the 6 available with Emovice suffice.
I was very interested in OmniVocal, when it was released due to the fact that it is integrated into the DAW, but the difference between Emvoice and other contenders, is the fact that Emvoice has a very strict dictionary, so if a word isn’t available, or a pronunciation, you can use phonemes to create words or syllables instead, and this works quite well, whereas other systems primarily are made up of phonemes and therefore accents can come through very strongly.
Whether or not it improves before the official release depends heavily on how many of us tell them that its an issue for us and how strongly we feel about it. Aside from getting a large sample pool to see how software functions on an endless variety of PC hardware and any possible poor interactions with other software, its the whole point of beta testing. Being supportive and assuming something will just naturally get better is counterproductive no matter where you apply it, but its pure poison when it comes to beta testing. They’re not children singing badly in a Christmas play that’ll get discouraged and never try again if people are honest. The developers are professionals giving their product away free for the sole purpose of seeing what issues the market has with their product so that they can fix them before launch, in order to avoid negative reviews that will otherwise prevent consumers from buying it. When you offer undue support to a mediocre artist on social media, they tend to brush off honest feedback as “haters”. They confuse the uninvested likes, 100/thumbs-up/flame emojis, and incomplete sentences of praise that people add to make the commenter feel better about themselves for real support, and continue being mediocre rather than putting serious effort into improving themselves to actually become musicians/painters/illustrators worthy of praise. When you tell a company that their severely flawed software is “excellent” during a beta test, on the other hand, you water-down important feedback and those issues get pushed aside onto the list of improvements that can wait until after launch; where they may or may not be addressed later.
Omnivocal’s inability to pronounce certain phrases in English will result in it being a flop in countries where English is the primary language. The primary language in the U.S. is English. Whereas people throughout the planet speak English as a first or second language, the vast majority of Americans are unilingual. So, vocal synthesis software that can’t pronounce common English words will get poor reviews and ultimately flop in the U.S. That one issue alone will mean the product itself fails because The Unites States is the biggest consumer market in the world. If a product fails at launch, they usually discontinue rather than invest more money to fix the issues.
Calling a product that’s bound to fail in the country where most of its sales are expected to come from because its a vocal synthesizer that can’t pronounce words in one of the most widely spoken languages on Earth “excellent” during its beta testing phase, is ridiculous. They have to correct the pronunciation issue before we can even begin to complain about it only having 2 voice presets and a complete lack of voice design controls, unnatural timbre, lack of human character, etc.
The point is that being a beta test isn’t a reason to go easy on a product. Its a reason to be as critical as possible because that’s the feedback they need so badly that they’re willing to give it away for free. “Its a beta test” is only a valid reason for not judging the final release version prematurely.
I’m aware its branded as a “Yamaha” product and where Yamaha is based. Doesn’t change the fact that it sounded Eastern-European to me. With Yamaha’s premiere audio software company being Steinberg and having been falling asleep to WW2 documentaries for the past few weeks, making German accents more familiar to me than ever before, German seemed like the likely candidate. “German accent” is quicker to type than “some unidentified Eastern-European sounding or perhaps even Scandinavian accent” lol. It definitely doesn’t sound Japanese or Asian.
The difference between established software such as Emvoice and others, including OmniVocal, is the former is more about creating words as sung rather than making sounds resemble syllables or words, and OmniVocal is not currently a serious product in my view unfortunately.
I was very excited when it was released but I found out quickly that it was very hacky to use, for me to try and get it to sing common words, let alone more complex words with adequate diction.
The premise that you can use a virtual instrument, as a precursor to recording a real singing voice, is flawed in the case of OmniVocal because it isn’t about constructing words or phrases, but rather it is about trying to make sounds resemble syllables or words, and unfortunately it does a very poor job in my opinion but hopefully that will change because integration of vocal synthesis inside the DAW, would be very welcome.
No offense but I’m going to disagree with you. That’s what beta testers are for, I don’t think getting angry at a beta release is very constructive. Cheers.
Each to their own. I used it for guide backing vocals on a recent song production. We ended up using the Omnivocal guides in the final track because they sounded perfect for the song.
Yes, there is always room for improvement, but in this context they worked really, really well. So, in this case the Beta worked excellently.
