VOICEPEAK review: natural speech, commercial use, and a one-time purchase

Note: This article is more than one year old.

• VoicePeak presets, phonetic notation, and the interface vary by version. Check the official updates and manual.
• Sample-voice availability and the product lineup may change. Follow the current purchase-page descriptions.
• Listening impressions depend on the playback setup, including headphones and DAC. This is a subjective account from the time.

About a year before writing this, I bought VOICEPEAK, a text-to-speech program.

It is AI speech-synthesis software.

AI voices come in many forms, from old-fashioned robotic readers to Vocaloid and Yukkuri-style voices.

After comparing several options, I chose VOICEPEAK for its balance and high quality.

The one-time purchase and commercial-use permission were also advantages.

I bought the VOICEPEAK Commercially Usable Six-Narrator Set, then added Furimomen and Miyamai Moka. Let's have them read the text above. The following clips were played without any particular tuning. What do you think? They sound rather good.

View VOICEPEAK Commercially Usable Six-Narrator Set (original, boxed edition) on Amazon (Ad)

View VOICEPEAK Furimomen (boxed edition) on Amazon (Ad)

View VOICEPEAK Miyamai Moka (boxed edition) on Amazon (Ad)

Furimomen reading sample

Miyamai Moka reading sample

Voice settings

Several settings are available. On the right are voice and emotion controls. The voice settings include speed, pitch, pause, and volume. Emotion options seem to vary slightly by character. I could not clearly tell the difference made by the pause setting. It is best to listen while adjusting them.

Accent settings

You can also adjust accent, intonation, and duration when something sounds unnatural. I understand the distinction between accent and intonation, but was not sure exactly how to tune them, so experimentation helps.

Pronunciation dictionary

Entering "wav" caused the program to read the letters W-A-V in Japanese. A dictionary entry lets it pronounce the word as "wavu" instead. Convenient.

Exporting audio

You can export the synthesized speech as WAV and other formats through Export. Options include saving separate blocks and setting filename rules.

Recent AI speech synthesis is impressive. Doesn't this sound usable? It is more natural than I expected. What a time to be alive. Licensing appears to have detailed categories, though, so check carefully before using it.

Scope of the individual commercial-use license:

https://www.ssw.co.jp/products/talk/voicepeak_una/VOICEPERAK_Personal_Commercial_License.pdf

VOICEPEAK

https://www.ah-soft.com/

VOICEPEAK Commercially Usable Six-Narrator Set (original, boxed edition)View details on Amazon Ad · Amazon Associates link

VOICEPEAK Furimomen (boxed edition)View details on Amazon Ad · Amazon Associates link

VOICEPEAK Miyamai Moka (boxed edition)View details on Amazon Ad · Amazon Associates link